The Assessment and Treatment of Disordered Speech and Voice in Parkinson S Disease Using A PC-based Telerehabilitation System

The Assessment and Treatment of
Disordered Speech and Voice in

Parkinson‟s Disease Using a PC-based
Telerehabilitation System
Gabriella Amalia Constantinescu
B Sp Path (Hons)
A thesis submitted for the degree of Doctor of Philosophy at
The University of Queensland in July 2010
School of Health and Rehabilitation Sciences

ii
Declaration by author
This thesis is composed of my original work, and contains no material previously

published or written by another person except where due reference has been made in the text. I
have clearly stated the contribution by others to jointly-authored works that I have included in my
thesis.
I have clearly stated the contribution of others to my thesis as a whole, including statistical
assistance, survey design, data analysis, significant technical procedures, professional editorial
advice, and any other original research work used or reported in my thesis. The content of my
thesis is the result of work I have carried out since the commencement of my research higher degree
candidature and does not include a substantial part of work that has been submitted to qualify for
the award of any other degree or diploma in any university or other tertiary institution. I have
clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.
I acknowledge that an electronic copy of my thesis must be lodged with the University
Library and, subject to the General Award Rules of The University of Queensland, immediately
made available for research and study in accordance with the Copyright Act 1968.
I acknowledge that copyright of all material contained in my thesis resides with the
copyright holder(s) of that material.
Gabriella Amalia Constantinescu
iii
iv
Statement of Contributions to Jointly
Authored Works Contained in the
Thesis
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. Assessing
disordered speech and voice in Parkinson‟s disease: A telerehabiliation application.
International Journal of Language and Communication Disorders, Early Online Article,
1-15.
Constantinescu co-designed the software application for the study and was primarily responsible for
the concept and design of the study, data collection, statistical analysis and interpretation with
guidance from a statistician, drafting and writing the article; Russell was primarily responsible for
the technical development of the study including the software application, data capture techniques,
technical adjustments and acoustic analyses and was assisted by a computer programmer;
Theodoros, Russell and Ward contributed to the design of the study, assisted with data
interpretation and reviewed drafts of the article; Wilson was primarily responsible for the hardware
development of the acoustic speech processor and reviewed drafts of the article; Wootton reviewed
drafts of the article.
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. (in press).
Treating disordered speech and voice in Parkinson‟s disease online: A randomised
controlled trial. International Journal of Language and Communication Disorder.
v
Constantinescu, G.A., Theodoros, D.G., Russell, T.G., Ward, E.C., Wilson, S.J., & Wootton, R.
(2010). Home-based speech treatment for Parkinson‟s disease delivered remotely: A case
report. Journal of Telemedicine and Telecare, 16, 100-104.
vi
Statement of Contributions by Others
to the Thesis as a Whole
The PhD candidate was primarily responsible for the project design, participant recruitment, data
collection, analysis and interpretation of the research data as well as the preparation of the
manuscript. Significant input into the thesis has also been made by the following persons:
Prof Theodoros had substantial input into the concept and project design, gaining ethical approval,
data analysis, interpretation and revising the written work.
Dr. Russell was primarily responsible for the concept and design of the telerehabilitation system,
and had substantial input into the concept and design of the project, calibration of the
telerehabilitation system, data analysis, interpretation and revising the written work.
Prof. Ward has substantial input into the data analysis, interpretation and revising the written work.
Assoc/Prof Stephen Wilson was primarily responsible for the design of the telerehabilitation
system‟s acoustic speech processor and had significant input into the calibration trials.
Mr Roy Anderson has substantial input into the technical development of the telerehabilitation
system.
Dr. Anne Hill assisted in the development of the informal assessment of dysarthria used in the
assessment study in Chapter 3.
vii
viii
Statement of Parts of the Thesis
Submitted to Qualify for the Award of
Another Degree
None
ix
x
Published Works by the Author
Incorporated into the Thesis
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. Assessing
1-15.
Chapter 3 has been created and expanded from this manuscript (see Appendix).
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. (in press).
Constantinescu, G.A., Theodoros, D.G., Russell, T.G., Ward, E.C., Wilson, S.J., & Wootton, R.
(2010). Home-based speech treatment for Parkinson‟s disease delivered remotely: A case
xi
xii
Additional Published Works by the
Author Relevant to the Thesis but not
Forming Part of it
Theodoros, D., Constantinescu, G., Russell, T.G., Ward, E.C., Wilson, S.J., & Wootton, R. (2006).
Treating the speech disorder in Parkinson‟s disease online. Journal of Telemedicine and
Telecare, 12(Suppl. 3), 88-91.
xiii
xiv
List of Presentations
Theodoros, D., Constantinescu, G., Russell, T., Ward, E., Wilson, S., & Wootton, R. (2008). The
effectiveness of the LSVT online: A randomised control trial. The American Telemedicine
Association 13th Annual International Meeting and Exposition, Seattle: USA, April 2008.
Theodoros, D., Constantinescu, G., Russell, T., Ward, E., Wilson, S., & Wootton, R. (2008).
Validating the delivery of the LSVT online. Motor Speech Disorders Conference, Monterey:
USA, March 2008.
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. (2007).
Validating the online delivery of the LSVT for treating the speech disorder in Parkinson‟s
disease: A randomised control trial. 7th International Conference on Successes and Failures
in Telehealth, Brisbane: Australia, August 2007.
Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. (2006). A
telerehabilitation application for assessing the speech and voice difficulties in Parkinson‟s
disease. The American Telemedicine Association 11th Annual International Meeting and
Exposition, San Diego: USA, May 2006.
Theodoros, D., Constantinescu, G., Russell, T., Ward, E., & Wootton, R. (2006). A
telerehabilitation application for treating disorderd speech in Parkinson‟s disease. The
American Telemedicine Association 11th Annual International Meeting and Exposition, San
Diego: USA, May 2006.
Treating the speech disorder in Parkinson‟s disease online: Successess and challenges. 6th
International Conference on Successes and Failues in Telehealth, Brisbane: Australia,
August 2006.
xv
Theodoros, D., Russell, T., Hill, A., Constantinescu, G., Waite, M., Ward, L., & Cahill, L. (2006).
Telerehabilitation: An innovation in service delivery. Speech Pathology Australia
Conference, Freemantle: Australia, May 2006.
Assessing disordered speech in Parkinson‟s disease online: A telerehabilitation application.
Motor Speech Disorders Conference, Texas: USA, March 2006.
Constantinescu, G. (2004). A telehealth application for the assessment and treatment of disordered
speech in Parkionson‟s disease. School of Health and Rehabilitation Sciences Postgraduate
Day Conference, Brisbane: Australia, December, 2004.
xvi
Acknowledgements
This thesis is dedicated to Peggy and Dawn.
Throughout my PhD journey, there have been many individuals that have provided me
with assistance and support of which I am extremely grateful and wish to acknowledge here. In the
first instance, I wish to earnestly thank my doctoral supervisors, Professor Deborah Theodoros,
Professor Elizabeth Ward and Dr Trevor Russell. From the very beginning, you have provided me
with your expert guidance and have instilled in me extremely high research and writing standards
and encouraged independent thought. I have valued your continual support, commitment and
passion for the research throughout the years and I thank you for this.
I must acknowledge my colleagues in the Telerehabilitation Research Unit, Dr Anne Hill,

Monique Waite, Jasmine Cowles, Christina Iezzi and Roy Anderson for their assistance during the
data collection phases. Your enthusiasm, professionalism and support throughout the research were
gratefully appreciated. Anne and Monique, I would particularly like to thank you for your
friendship which I treasure. You have shared similar journeys to mine and I have always
appreciated your understanding and advice.
My deepest thanks are extended to all the participants in the studies. Your enthusiasm in
taking part in the research was very much appreciated and I was honoured to get to know each and
every one of you during the treatment phase and to be a part of your lives. Your dignity and
outlook on life was truly inspiring.
To my postdoctoral colleagues and friends, thank you for your friendship, light-
heartedness and support throughout the years which have made the PhD process an enjoyable one
for me. Thank you in particular to Liz Savina, Kieran Broome, Dr Emma Finch, Dr Jinny
Uthaikhup, Dr Andy Chien, Dr Rachel Wenke and Dr Erin Smith.
xvii
I would like to thank my friends and colleagues at the Hear and Say Centre for their
ongoing encouragement during the final stages of my research. I am especially indebted to Dimity
Dornan for her continual understanding and support. It has always been appreciated.
To all of my friends, particularly the high school, uni and Saturday night Church crowd,
thank you for your encouragement and patience throughout the years. I have always valued your
efforts in keeping me included in the events in your lives despite my random presence due to the
PhD commitments.
To the hundreds of children and their families from around the world that I met through
Operation Smile, thank you for making me a part of your lives. You have always inspired and
grounded me and given me the drive to complete the PhD. I hope that soon we can look at
improved access to services for you too.
Finally, I wish to acknowledge my family for their continual support and patience. I would
particularly like to express my sincerest thank you to my sister, Michelle. Without fail, you have
been there for me through everything and have always encouraged me to keep on going and to
never give up. You have always taken a keen interest in all aspects of my research and have
provided me with invaluable advice which I cherish. I am blessed to have you as my sister.
Doamne, buzele mele vei deschide si gura mea va vesti lauda Ta.
xviii
Abstract
Individuals with Parkinson‟s disease (PD) face numerous access barriers to speech-language
pathology services including their own physical incapacity, difficulties with transport, travel, cost of
travel and the large distances to healthcare facilities. Telerehabilitation is a possible solution for
these problems, whereby rehabilitation services may be delivered remotely to patients via
telecommunication and information technologies. A number of studies have demonstrated the
feasibility of telerehabilitation via personal computer (PC) based systems for assessment and
treatment of dysarthria, apraxia of speech, aphasia and voice disorders in adults with neurological
impairments. To date, however, no large-scale studies have specifically focused on the online
assessment and treatment of the speech and voice disorder associated with PD.
The aim of this thesis was to investigate the validity of telerehabilitation for the assessment
and treatment of PD and to provide a framework for telerehabilitation use in this area. For the
studies in this thesis, a custom made PC-based telerehabilitation system was developed. The
system incorporated a number of features to meet face-to-face requirements for the management of
PD including: (1) real-time videoconferencing which operated on a 128 kbit/s Internet connection;
(2) control of the participant web cameras remotely; (3) store-and-forward function; (4) the ability
to display printed materials for the participant remotely; (5) ease of operation; and (6) the ability to
objectively measure real-time calibrated average recordings of vocal sound pressure level (SPL),
fundamental frequency (F0) and duration.
The initial component of the research established the validity of the telerehabilitation system
as an acoustic measurement tool via a series of calibration and verification phases, where
comparisons were made to standard face-to-face reference measurement tools (Chapter 2). The
studies confirmed the accuracy of the telerehabilitation system with non-significant differences in
acoustic measures obtained between the telerehabilitation system and reference tools. All measures
also fell within predetermined clinical criteria.
xix
Following this, a series of four studies were conducted to achieve the aim of this thesis. The
first cohort study aimed to investigate the validity and reliability of online assessment of the speech
and voice disorder associated with PD (Chapter 3). A total of 61 participants with PD and mild to
severe hypokinetic dysarthria took part in the laboratory-based equivalence study. The participants
were assessed simultaneously in the online and face-to-face environments. The assessment protocol
included perceptual measures of voice and oromotor function, articulatory precision, speech
intelligibility, and acoustic measures of mean SPL, maximum F0 range and maximum duration of
sustained vowel phonation. The level of agreement between the online and face-to-face ratings was
determined using several different analyses, depending on the parameter. These included percent
close agreement, quadratic weighted Kappa and the Bland and Altman limits of agreement. The
results revealed that for all perceptual parameters, percent close agreement between the two
environments was within the predetermined clinical criterion of 80%. However, a number of these
parameters fell below the clinical criterion of good agreement based on the quadratic weighted
Kappa. The discrepancy in findings was considered to relate to rater variability commonly seen in
perceptual ratings and/or the nature of the statistic per se, rather than the online environment. For
the remaining speech intelligibility and acoustic parameters, the Bland and Altman limits of
agreement analysis revealed comparability between the two environments, with only the word
intelligibility values falling below the clinical criterion. It was suggested that speaker severity may
have influenced the results on this parameter. The intra- and inter-rater reliability scores were also
comparable between the online and face-to-face environments, achieving moderate to very good
agreement for all tasks.
Subsequent to the validation of the online assessment of hypokinetic dysarthria, three

treatment studies were conducted. The first was a laboratory-based randomised controlled non-
inferiority trial investigating the validity of online treatment delivery for the speech and voice
disorder associated with PD. In this study, 34 participants with idiopathic PD (IPD) and mild to
moderate hypokinetic dysarthria received the Lee Silverman Voice Treatment (LSVT®) for PD, in
either the online or face-to-face environment (n = 17 in each). The study findings were very
promising, with non-inferiority of the online LSVT® modality confirmed for the primary outcome
measure of mean change in SPL on a monologue task. Non-significant main effects of treatment
environment, dysarthria severity and interaction effects were also noted across all of the acoustic
and perceptual parameters (p > .05). Further, it was encouraging to note that significant gains
following the LSVT® were made on the majority of acoustic and perceptual parameters for
xx
participants in both treatment environments (p < .05 for SPL tasks; maximum F0 range; all voice
parameters and speech intelligibility).
In order to determine the feasibility of online LSVT® for complex cases of PD, the second
laboratory-based treatment study was conducted. Four participants with marked communication,
cognitive and physical difficulties took part in the study. The participants had been diagnosed with
either IPD and had undergone neurosurgical intervention or with progressive supranuclear palsy.
Online treatment delivery for these complex cases showed promise, with functional improvements
in communication reported for all cases. Although the treatment gains on the perceptual and
acoustic measures were generally variable, it was evident that the outcomes were influenced by the
complex participant factors that were all external to the treatment environment.
In the final study, a single participant with IPD and mild hypokinetic dysarthria received the
®
LSVT remotely from his home, in order to determine the feasibly of home telecare for PD.
Substantial improvements were achieved on the majority of acoustic parameters (SPL tasks and
maximum duration of sustained vowel phonation), and on the perceptual measure of vocal
breathiness. The results of the study demonstrated the feasibility of home-based treatment for PD in
the real-world setting, within an individual‟s most natural and preferred environment.
Overall, the findings presented in this thesis provide evidence for the validity of assessing
and treating the disordered speech and voice of PD via telerehabilitation. Although some
challenges unique to the online environment were encountered in the studies and were mainly due
to the audio and video quality during videoconferencing, the participants and speech-language
pathologists (SLPs) were able to appropriately manage these without substantial impact on the
online delivery overall. The high participant satisfaction with the online modality across all studies
further supported this, where the majority of participants (80% and above) were more than satisfied
or very satisfied with the online modality overall. The thesis provides a framework for the delivery
of telerehabilitation services for people with PD, with the potential to alleviate the current access
issues that exist for this population. Further research is needed to build on the findings in this thesis
and should include the validity of home-based assessment and treatment, in-depth investigations of
participant and SLP satisfaction and cost-benefit analyses of telerehabilitation for PD. Such
investigations are necessary in order to determine the complete benefits of telerehabilitation as an
xxi
additional or alternate mode of service delivery for this population, and the circumstances in which
this approach is most suitable.
xxii
Keywords
Parkinson‟s disease, telerehabilitation, assessment, treatment, Lee Silverman Voice Treatment,

speech, voice, speech-language pathology, PC-based.
Australian and New Zealand Standard

Research Classification (ANZSRC)
920201 Allied Health Therapies 50%, 100599 Communication Technologies not elsewhere
classified 50%.
xxiii
xxiv
Table of Contents
CHAPTER 1 INTRODUCTION TO PARKINSON’S DISEASE AND
TELEREHABILITATION ............................................................................................................... 1
1.1 INTRODUCTION ......................................................................................................................... 1

1.2 PARKINSON‟S DISEASE.............................................................................................................. 3
1.2.1 Economic Burden of Parkinson’s Disease ....................................................................... 3
1.3 IDIOPATHIC PARKINSON‟S DISEASE .......................................................................................... 4
1.3.1 Neuropathology and Clinical Presentation of Motor Symptoms ..................................... 4
1.3.2 Non-motor Symptoms ....................................................................................................... 6
1.3.3 Speech and Voice Disturbances ....................................................................................... 7
1.3.3.1 Respiratory Subsystem ............................................................................................. 7
1.3.3.2 Phonatory Subsystem ............................................................................................... 8
1.3.3.3 Articulatory Subsystem ............................................................................................ 8
1.3.3.4 Velopharyngeal Subsystem ...................................................................................... 9
1.3.3.5 Prosodic Features ..................................................................................................... 9
1.3.3.6 Overall Speech Intelligibility ................................................................................... 9
1.3.4 Occurrence of Hypokinetic Dysarthria .......................................................................... 10
1.3.5 Sensory Perception Difficulties ...................................................................................... 11
1.3.6 Impact of Hypokinetic Dysarthria on Communication .................................................. 11
1.4 PARKINSON-PLUS SYNDROMES ............................................................................................... 13
1.4.1 Progressive Supranuclear Palsy .................................................................................... 13
1.4.2 Multiple System Atrophy ................................................................................................ 14
1.5 MANAGEMENT OF PARKINSON‟S DISEASE .............................................................................. 15
1.5.1 Pharmacological Management ...................................................................................... 15
1.5.2 Neurosurgical Management ........................................................................................... 15
1.5.2.1 Speech and Voice Changes Associated with Neurosurgical Management ............ 17
1.5.3 The Lee Silverman Voice Treatment® and Idiopathic Parkinson’s Disease .................. 17
1.5.3.1 The Effects of the Lee Silverman Voice Treatment® at the Respiratory and
Phonatory Subsystems ........................................................................................................... 19
1.5.3.2 The Global Effects of the Lee Silverman Voice Treatment®................................. 21
1.5.3.3 The Lee Silverman Voice Treatment® and Surgical Idiopathic Parkinson‟s Disease
22
1.5.3.4 The Lee Silverman Voice Treatment ® and Parkinson-plus Syndromes ............... 24
xxv
1.6 OPTIMAL MANAGEMENT OF PD AND CURRENT CHALLENGES TO SPEECH-LANGUAGE
PATHOLOGY SERVICES .................................................................................................................... 25
1.7 TELEREHABILITATION NOMENCLATURE ................................................................................. 27
1.8 THE DEVELOPMENT OF TELEHEALTH ...................................................................................... 28
1.8.1 Early Telehealth Systems ............................................................................................... 29
1.8.2 The Birth of Modern Telehealth ..................................................................................... 29
1.8.3 The Rebirth of Modern Telehealth ................................................................................. 30
1.9 TELEHEALTH IN AUSTRALIA ................................................................................................... 31
1.10 SATISFACTION WITH TELEHEALTH ...................................................................................... 33
1.10.1 Professional Satisfaction............................................................................................ 33
1.10.1.1 Technology Concerns ............................................................................................ 34
1.10.1.2 Patient-Professional Dynamic and Service Delivery ............................................. 34
1.10.2 Patient Satisfaction .................................................................................................... 35
1.11 TELEREHABILITATION STUDIES IN SPEECH-LANGUAGE PATHOLOGY ................................. 36
1.11.1 Speech-Language Pathology Assessment Studies ...................................................... 37
1.11.2 Speech-Language Pathology Treatment Studies........................................................ 40
1.12 AIMS OF THE THESIS ........................................................................................................... 41
CHAPTER 2 TELEREHABILITATION SYSTEM DESIGN AND CALIBRATION ...... 45

2.1 INTRODUCTION TO THE TELEREHABILITATION SYSTEM .......................................................... 45
2.2 FEATURE SET OF THE TELEREHABILITATION SYSTEM ............................................................. 46
2.2.1 Videoconferencing ......................................................................................................... 46
2.2.2 Control of Remote Participant Web Cameras ............................................................... 49
2.2.3 Store-and-Forward Capabilities .................................................................................... 50
2.2.4 Remote Display of Materials.......................................................................................... 51
2.2.5 Ease of Operation .......................................................................................................... 52
2.2.6 Measures of Sound Pressure Level, Fundamental Frequency and Duration ................ 52
2.3 DESIGN OF THE TELEREHABILITATION SYSTEM AS AN ACOUSTIC MEASUREMENT TOOL ....... 55
2.4 CALIBRATION AND VERIFICATION OF THE TELEREHABILITATION SYSTEM AS AN ACOUSTIC
MEASUREMENT TOOL ..................................................................................................................... 56
2.4.1 Experiment 1: Calibration of Sound Pressure Level Using Pure Tones ....................... 57
2.4.1.1 Experiment 1: Method ........................................................................................... 57
2.4.1.2 Experiment 1: Results and Discussion ................................................................... 59
2.4.2 Experiment 2: Verification of SPL Using Pure Tones ................................................... 60
xxvi
2.4.3 Experiment 3: Verification of Sound Pressure Level Using Live Voice ........................ 62
2.4.4 Experiment 4: Verification of Fundamental Frequency Using Pure Tones .................. 67
2.5 CONCLUSION ........................................................................................................................... 70
CHAPTER 3 ASSESSING DISORDERED SPEECH AND VOICE IN PARKINSON’S

DISEASE ONLINE: A TELEREHABILITATION SYSTEM.................................................... 71
3.1 INTRODUCTION ....................................................................................................................... 71

3.2 METHOD ................................................................................................................................. 74
3.2.1 Sample Size Calculation................................................................................................. 74
3.2.2 Participants .................................................................................................................... 74
3.2.3 Assessors ........................................................................................................................ 80
3.2.4 Assessment Battery ......................................................................................................... 81
3.2.4.1 Acoustic Measures ................................................................................................. 82
3.2.4.1.1 Sound pressure level and maximum duration of sustained vowel phonation .... 82
3.2.4.1.2 Maximum fundamental frequency range ........................................................... 82
3.2.4.2 Perceptual Measures .............................................................................................. 83
3.2.4.2.1 Perceptual voice parameters............................................................................... 83
3.2.4.2.2 Oromotor function.............................................................................................. 83
3.2.4.2.3 Overall articulatory precision............................................................................. 83
3.2.4.2.4 Measures of speech intelligibility ...................................................................... 84
3.2.4.3 Participant Satisfaction Questionnaire ................................................................... 84
3.2.4.4 Speech-Language Pathologist Comments .............................................................. 85
3.2.5 Assessment Environment ................................................................................................ 85
3.2.5.1 Online Environment ............................................................................................... 85
3.2.5.2 Face-to-Face Environment ..................................................................................... 91
3.2.6 Statistical Analyses ........................................................................................................ 91
3.2.6.1 Bland and Altman Limits of Agreement ................................................................ 92
3.2.6.2 Percent Close Agreement ....................................................................................... 92
3.2.6.3 Quadratic Weighted Kappa Statistic ...................................................................... 93
3.2.6.4 Reliability ............................................................................................................... 93
xxvii
3.3 RESULTS ................................................................................................................................. 94
3.3.1 Acoustic Measures ......................................................................................................... 94
3.3.2 Perceptual Measures...................................................................................................... 96
3.3.2.1 Perceptual Voice Parameters ................................................................................. 96
3.3.2.2 Oromotor Function Parameters .............................................................................. 97
3.3.2.3 Overall Articulatory Precision ............................................................................... 99
3.3.2.4 Measures of Intelligibility ...................................................................................... 99
3.3.3 Reliability ..................................................................................................................... 101
3.3.4 Online Assessment Delivery ......................................................................................... 102
3.3.5 Participant Satisfaction Questionnaire ........................................................................ 103
3.4 DISCUSSION .......................................................................................................................... 104
3.4.1 Acoustic Measures ....................................................................................................... 105
3.4.2 Perceptual Measures.................................................................................................... 105
3.4.2.1 Perceptual Voice Parameters ............................................................................... 105
3.4.2.2 Oromotor Parameters ........................................................................................... 107
3.4.2.3 Overall Articulatory Precision ............................................................................. 109
3.4.2.4 Measures of Speech Intelligibility ....................................................................... 109
3.4.3 Benefits and Challenges of Online Assessment............................................................ 110
3.4.3.1 Audio Quality....................................................................................................... 110
3.4.3.2 Video Quality ....................................................................................................... 111
3.4.3.3 System Operation and Speech-Language Pathologist Challenges ....................... 112
3.4.4 Study Limitations and Future Directions ..................................................................... 113
3.5 CONCLUSION ......................................................................................................................... 114
CHAPTER 4 TREATING DISORDERED SPEECH AND VOICE IN PARKINSON’S

DISEASE ONLINE: A RANDOMISED CONTROLLED NON-INFERIORITY TRIAL .... 115
4.1 INTRODUCTION ..................................................................................................................... 115

4.2 METHOD ............................................................................................................................... 117
4.2.1 Study Design ................................................................................................................ 117
4.2.2 Sample Size Calculation............................................................................................... 117
4.2.3 Participants .................................................................................................................. 118
4.2.4 Speech-Language Pathologists .................................................................................... 123
4.2.5 The Lee Silverman Voice Treatment® Programme ...................................................... 123
4.2.6 Treatment Environment................................................................................................ 125
4.2.6.1 Online Lee Silverman Voice Treatment® Environment ...................................... 125
xxviii
4.2.6.2 Face-to-Face Lee Silverman Voice Treatment® Environment ............................ 127
4.2.7 Assessment of Outcomes .............................................................................................. 127
4.2.7.1 Acoustic Measures ............................................................................................... 128
4.2.7.2 Perceptual Speech and Voice Parameters ............................................................ 129
4.2.7.3 Participant Satisfaction Questionnaire ................................................................. 131
4.2.7.4 Speech-Language Pathologist Comments ............................................................ 131
4.2.8 Statistical Analyses ...................................................................................................... 132
4.3 RESULTS ............................................................................................................................... 132
4.3.3 Reliability of Direct Magnitude Estimation Scaling .................................................... 137
4.3.4 Online Lee Silverman Voice Treatment® Delivery ...................................................... 137
4.3.5 Participant Satisfaction Questionnaire ........................................................................ 137
4.4 DISCUSSION .......................................................................................................................... 139
4.4.1.1 Sound Pressure Level Tasks ................................................................................ 140
4.4.1.2 Maximum Duration of Sustained Vowel Phonation ............................................ 142
4.4.1.3 Maximum Fundamental Frequency Range .......................................................... 142
4.4.2.1 Perceptual Voice Parameters ............................................................................... 143
4.4.2.2 Overall Articulatory Precision ............................................................................. 144
4.4.2.3 Speech Intelligibility ............................................................................................ 144
4.4.3 Benefits and Challenges of Online Lee Silverman Voice Treatment® Delivery .......... 146
4.4.3.1 Audio Quality....................................................................................................... 146
4.4.3.2 Video Quality ....................................................................................................... 148
4.4.3.3 System Operation and Speech-Language Pathologist Challenges ....................... 150
4.5 STUDY LIMITATIONS AND FUTURE DIRECTIONS ................................................................... 151
4.6 CONCLUSION ......................................................................................................................... 153
CHAPTER 5 ONLINE LSVT FOR COMPLEX CASES OF PD: CASE STUDIES ....... 155
5.1 INTRODUCTION ..................................................................................................................... 155

5.2 METHOD ............................................................................................................................... 158
5.2.1 Participants .................................................................................................................. 158
5.2.2 Procedure ..................................................................................................................... 163
xxix
5.3 RESULTS AND DISCUSSION .................................................................................................... 164
5.3.1 Participant 55 .............................................................................................................. 164
5.3.1.1 Treatment Outcome.............................................................................................. 165
5.3.1.2 Participant Perspectives Regarding Online Lee Silverman Voice Treatment® ... 168
5.3.2 Participant 49 .............................................................................................................. 169
5.3.3 Participant 48 .............................................................................................................. 172
5.3.4 Participant 54 .............................................................................................................. 175
5.3.4.2 Participant Perspectives Regarding Online LSVT® ............................................. 177
5.3.5 Speech-Language Pathologist Challenges with Online Lee Silverman Voice Treatment®
Delivery .................................................................................................................................... 178
5.3.5.1 Participant factors ................................................................................................ 178
5.3.5.2 Audio Quality....................................................................................................... 180
5.3.5.3 Video Quality ....................................................................................................... 181
5.3.5.4 System Operation ................................................................................................. 182
5.3.5.5 Nature of the Online Treatment Modality............................................................ 183
5.5 CONCLUSION ......................................................................................................................... 184
CHAPTER 6 HOME-BASED SPEECH TREATMENT FOR PARKINSON’S DISEASE

DELIVERED REMOTELY: A CASE REPORT ....................................................................... 185
6.1 INTRODUCTION ..................................................................................................................... 185

6.2 METHOD ............................................................................................................................... 186
6.2.1 Participant ................................................................................................................... 186
6.2.2 Procedure ..................................................................................................................... 188
6.3 RESULTS ............................................................................................................................... 191
6.4 DISCUSSION .......................................................................................................................... 194
6.4.3 Perspectives Regarding Home-based Lee Silverman Voice Treatment® Delivery ...... 198
xxx
6.4.3.1 Participant Perspectives ....................................................................................... 198
6.4.3.2 Speech-Language Pathologist Perspectives ......................................................... 199
6.6 CONCLUSION ......................................................................................................................... 201
CHAPTER 7 CONCLUSIONS .............................................................................................. 203
REFERENCES .......................................................................................................................... 209
APPENDIX – PUBLISHED PAPERS...................................................................................... 233
xxxi
xxxii
List of Tables
Table 3.1 Descriptive Characteristics of Participants...................................................................... 76
Table 3.2 Assessment Procedure for Online and Face-to-Face Environment ................................. 88
Table 3.3 Perceptual Voice Parameters between Face-to-Face and Online Environments ............ 97
Table 3.4 Perceptual Oromotor Parameters between Face-to-Face and Online Environments ..... 98
Table 3.5 Intra-class Correlation Values for Intra-Rater and Inter-Rater Reliability for Online and
Face-to-Face Ratings ....................................................................................................................... 102
Table 4.1 Descriptive Characteristics of Participants who Received the LSVT® in the Online and
Face-to-Face Environments ............................................................................................................. 119
Table 4.2 Acoustic Parameters: Means, Standard Deviations (in parentheses), F and p values ... 133
Table 4.3 Perceptual Parameters: Means, Standard Deviations (in parentheses), F and p values
.......................................................................................................................................................... 135
Table 5.1 Descriptive Characteristics of Participants.................................................................... 160
Table 5.2 Pre- and Post-LSVT Values and Treatment Changes for the Acoustic Measures for
Participant 55 .................................................................................................................................. 166
Table 5.3 Pre- and Post-LSVT Values and Treatment Changes for the Perceptual Measures for
Participant 55 .................................................................................................................................. 167
Table 5.4 Pre- and Post-LSVT® Values and Treatment Changes for the Acoustic Measures for
Participant 49 .................................................................................................................................. 170
Table 5.5 Pre- and Post-LSVT® Values and Treatment Changes for the Perceptual Measures for
Participant 49 .................................................................................................................................. 171
Participant 48 .................................................................................................................................. 173
Participant 48 .................................................................................................................................. 174
xxxiii
Participant 54 .................................................................................................................................. 176
Participant 54 .................................................................................................................................. 177
Table 6.1 Pre- and Post-LSVT Values and Treatment Changes for the Acoustic Parameters ....... 192
Table 6.2 Pre- and Post-LSVT® DME Values and Treatment Changes for the Perceptual
Parameters ....................................................................................................................................... 193
xxxiv
List of Figures
Figure 1.1 Components of the basal ganglia in the forebrain and the substantia nigra in the
midbrain. .............................................................................................................................................. 5
Figure 1.2 Representation of the basal ganglia control circuit. ......................................................... 6
Figure 1.3 A schematic representation of the reduced sensory perception of effort, scaling and
amplitude of output in PD and the LSVT® programme targeting improvements in these areas. ...... 18
Figure 1.4 A schematic representation of the associated telehealth, telemedicine, telerehabilitation

and home telecare terms. .................................................................................................................... 28
Figure 2.1 Telerehabilitation system at participant site. ................................................................... 48
Figure 2.2 Telerehabilitation system at SLP site. . ........................................................................... 48
Figure 2.3 Web cameras on robotic arm at participant site. . ........................................................... 49
Figure 2.4 Schematic diagram of videoconferencing and the separate data channel used to transfer
store-and-forward audio and video files between the SLP and participant sites. .............................. 50
Figure 2.5 Telerehabilitation system‟s acoustic speech processor. ................................................. 54
Figure 2.6 Screen shot of the acoustic measurement software tool at the SLP site displaying SPL
and F0. . .............................................................................................................................................. 54
Figure 2.7 Schematic diagram of the telerehabilitation system‟s acoustic speech processor
hardware. ........................................................................................................................................... 56
Figure 2.8 Equipment and set-up of calibration phase for pure tone SPL. ....................................... 58
Figure 2.9 Comparative fit plot using the linear and power curve estimation regression models for
the non-calibrated telerehabilitation system and Visi-Pitch II measures of pure tone SPL. ............. 59
Figure 2.10 Comparative fit plot using the linear curve estimation regression model for the
calibrated telerehabilitation system and Visi-Pitch II measures of pure tone SPL. . ......................... 61
Figure 2.11 Bland and Altman LA for measures of pure tone SPL. . ............................................... 62
Figure 2.12 Equipment and set-up of verification phase of SPL using live voice. ........................... 63
xxxv
calibrated telerehabilitation system and Digital SLM for measures of SPL using a female voice. ... 65
calibrated telerehabilitation system and Digital SLM for measures of SPL using a male voice. ...... 65
Figure 2.15 Bland and Altman LA for measures of SPL using a female voice. .............................. 66
Figure 2.16 Bland and Altman LA for measures of SPL using a male voice. ................................. 66
telerehabilitation system and Visi-Pitch II measures of pure tone F0. ............................................... 69
Figure 2.18 Bland and Altman LA for measures of pure tone F0. ................................................... 69
Figure 3.1 Schematic flow chart of participants through the study. ................................................. 80
Figure 3.2 Online-led assessment by online SLP and equipment at site. . ....................................... 90
Figure 3.3 Online-led assessment at participant site with face-to-face SLP as the silent rater. . ..... 90
Figure 3.4 Bland and Altman LA for maximum sustained vowel phonation, readings of The
Rainbow and Grandfather Passages and monologue task. ................................................................ 95
Figure 3.5 Bland and Atlman LA for maximum duration of sustained vowel phonation. ............... 95
Figure 3.6 Bland and Altman LA for maximum F0 range. ............................................................... 96
Figure 3.7 Bland and Altman LA for word intelligibility. ............................................................... 99
Figure 3.8 Bland and Altman LA for sentence intelligibility. ........................................................ 100
Figure 3.9 Bland and Altman LA for communication efficiency ratio. ......................................... 100
Figure 3.10 Participant satisfaction with the online assessment sessions. ...................................... 103
Figure 3.11 Participant satisfaction with the online audio quality. ................................................. 103
Figure 3.12 Participant satisfaction with the online video quality. . ............................................... 104
Figure 3.13 Participant satisfaction overall with online modality. ................................................. 104
Figure 4.1 Schematic flow chart of participants through the study. ............................................... 122
Figure 4.2 Online LSVT® environment at participant site. ............................................................ 126
xxxvi
Figure 4.3 Online LSVT® environment at SLP site. ...................................................................... 126
Figure 4.4 Face-to-face LSVT® environment. ................................................................................ 127
Figure 4.5 Participant satisfaction with the individual online treatment sessions. ......................... 138
Figure 4.6 Participant satisfaction with the online audio quality. ................................................... 138
Figure 4.7 Participant satisfaction with the online video quality. . ................................................. 139
Figure 4.8 Participant satisfaction overall with online treatment. . ................................................ 139
Figure 4.9 Comparison of post-treatment mean SPLs for IPD participants in the online and face-to-
face laboratory LSVT® study with those reported in the efficacy studies of face-to-face LSVT® and
healthy-older adults. ........................................................................................................................ 141
Figure 6.1 SLP treatment site. . ....................................................................................................... 189
Figure 6.2 Participant treatment site. ............................................................................................. 189
Figure 6.3 Comparison of mean SPL treatment gains for Participant 23 with those reported in the
efficacy studies of face-to-face LSVT® and in the laboratory trials in Chapter 4. .......................... 195
Figure 6.4 Comparison of post-treatment mean SPL values for Participant 23 with those reported in
the efficacy studies of face-to-face LSVT®, healthy older adults and laboratory trials in Chapter 4.
.......................................................................................................................................................... 196
xxxvii
xxxviii
List of Abbreviations
ANZTC Australian and New Zealand Telehealth Committee

ASSIDS Assessment of Intelligibility of Dysarthric Speech
ATS Advanced Technology Communications Satellite
CC Clinical criterion
CER Communication efficiency ratio
CI Confidence interval
cm Centimetres
codec Compression and decompression
CVA Cerebrovascular accident
dB Decibel
dB-C Decibel C-weighting
DBS Deep brain stimulation
DDK Diadochokinetic
DME Direct Magnitude Estimation
EAI Equal appearing interval scaling
F0 Fundamental frequency
FDA Frenchay Dysarthria Assessment
frames/s Frames per second
FTF Face-to-face
GLM General linear model
Hz Hertz
ICC Intra-class correlation coefficient
IPD Idiopathic Parkinson‟s disease
kbit/s Kilobits per second
km Kilometre
Kw Quadratic weighted kappa statistic
LA Limits of Agreement
®
LSVT Lee Silverman Voice Treatment
m Minute
MAD Mean absolute difference
xxxix
Mbit/s Megabits per second
ms Milliseconds
MSA Multiple system atrophy
NASA National Aeronautics and Space Administration
OAP Overall articulatory precision
OIC Overall speech intelligibility in conversation
PC Personal computer
PCA Percent close agreement
PD Parkinson‟s disease
PSP Progressive supranuclear palsy
PPS Parkinson-plus Syndrome
RT Respiration treatment
s Second
SI Sentence intelligibility
SLM Sound level meter
SLP Speech-language pathologist
SPL Sound pressure level
ST Semitone
TBI Traumatic brain injury
UK United Kingdom
USA United States of America
WHO World Health Organisation
WI Word intelligibility
xl
Chapter 1
Introduction to Parkinson’s Disease and Telerehabilitation
1.1 Introduction
Healthcare constitutes one of the largest sectors of modern society and plays a significant
role for the individual and community at large (Roberts, Rigby, & Birch, 2000). According to
Benzeval, Judge and Whitehead (1999), it is essential that society endeavours to provide every
individual with adequate healthcare opportunities, in order to sustain healthy living. The World
Health Organisation (WHO) also has a similar vision for the 21st century through its health-for-all
strategy (WHO, 1997a). Consequently, the ever changing and increasing demands of modern day
healthcare have required proportional advancement in general healthcare services throughout the
years (Darkins & Cary, 2000). Improvements in the quality and efficiency of these services have
been fuelled by advancements in medical technologies, which have become integral to the provision
of high standard diagnoses and treatments, illness prevention and overall healthcare promotion
(Kuo, Delvecchio, Babayan, & Preminger, 2001; Leeder, 2000; Lovell & Celler, 1999; Mun &
Turner, 1999).
However, despite the advancements in general healthcare services, concomitant

improvements in universal healthcare for all members of modern society have not always been
attainable, even in first world countries (Stanberry, 1998a). A multitude of healthcare inequity
issues relate to the current “traditional” mode of service delivery, which is “doctor” and “hospital”
centred (Mun & Turner, 1999), where both the patient and healthcare professional must be present
in the same location at the same time (Craig, 1999). This requirement creates several geographic,
monetary and technological barriers to healthcare access for patients. Firstly, geographic barriers
may hinder access for persons residing in rural and remote areas of the country, as well as in
residential care facilities and prisons from accessing the major medical facilities situated largely in
metropolitan regions (Bashshur, 1997; Darkins & Cary, 2000; Mun & Turner, 1999). Secondly, the
time and cost of travelling to these major healthcare facilities is high for rural and remote patients
and may constitute an additional barrier to healthcare access (Gustke, Balch, West, & Rogers,
2000). Thirdly, the discrepancy in the availability of advanced medical equipment and technology
and professional expertise between major metropolitan hospitals and smaller rural ones may
correspond to an inequity in healthcare provision (Kuo, et al., 2001). The current mode of
healthcare delivery is also problematic for healthcare professionals and the healthcare system at
large. Difficulties encountered by rural and remote professionals may include: (1) increased clinical
load; (2) professional isolation; as well as (3) decreased access to professional education
development and networking opportunities (Guilfoyle et al., 2003; Qureshi & Kvedar, 2003). With
respect to the healthcare system, the rising health demands of an increasing and ageing population,
combined with the rising cost of healthcare pose the largest problems (Leeder, 2000; Mun &
Turner, 1999).
Difficulties accessing services are particularly evident for persons with Parkinson‟s disease
(PD) in Australia who encounter barriers such as the long distances to the healthcare facilities,
difficulties with transport, travel and cost of travel, and their own physical incapacity (Grimm, Paul,
& Wakeham, 2004). These barriers are further compounded by the currently low priority placed on
speech-language pathology services for people with PD in the public healthcare system and the
limited number of speech-language pathologists (SLPs) trained to administer the effective Lee
Silverman Voice Treatment (LSVT®) for PD.
The overall difficulties encountered by all members of the healthcare system with
traditional healthcare, together with the growing focus on “healthcare for all”, have become the
driving forces for change to the current mode of service delivery (Bashshur, 1997; Craig, 1999;
Yellowlees & Brooks, 1999). Consequently, it has become increasingly recognized that employing
communication and information technologies in the delivery of distant healthcare services or
telehealth, may help address the current healthcare inequity issues (Craig, 1999). Telehealth
enables patient access to specialist care and advanced medical technology at the site convenient to
the patient rather than the healthcare provider, where the impact of distance, patient travel time, cost
and physical incapacity is minimized (Kuo, et al., 2001; Mun & Turner, 1999). Telehealth can
therefore, be significant in improving the quality of and access to healthcare for all individuals, and
is a major driving force in shaping the healthcare system for the 21st century (Bauer & Ringel, 1999;
Zundel, 1996). Under the umbrella term of telehealth, the remote delivery of rehabilitation services
via information and communication technologies or telerehabilitation (Rosen, 1999), may
promisingly become an additional or alternate mode of service delivery for people with long-term
2
rehabilitative needs such as those with PD. It is therefore the overall aim of this thesis to
investigate the effectiveness of telerehabilitation for the management of the speech and voice
difficulties associated with PD.
1.2 Parkinson’s Disease
PD refers to a group of progressive neurodegenerative disorders that are associated with

significant motor disturbances and speech difficulties. The types of PD include idiopathic
Parkinson‟s disease (IPD), Parkinson-plus syndromes (PPS), symptomatic parkinsonism and
heredodegenerative diseases (Fahn & Przedborski, 2005; Waters, 2005). IPD is the predominant
disorder, occurring in approximately 80% of diagnosed patients (Fahn & Przedborski, 2005). PPS
are the second most common types of the disorder and present in up to 15% of diagnosed patients
(Jankovic, 1989; Waters, 2005). Throughout this thesis, specific references will be made to IPD
and PPS. Therefore, this chapter will review the aetiology, clinical presentation and management of
these disorders. Where necessary, the term PD will also be used in this thesis to refer collectively to
the entire group of disorders.
1.2.1 Economic Burden of Parkinson’s Disease
As a result of the debilitating and progressive nature of the disease, substantial healthcare
system costs are incurred in the management of PD. In Australia alone, the total cost to the
healthcare system in 2005 was estimated at AUD $343.9 million. The greatest expenditure was
aged care ($168.6 million), followed by pharmaceuticals ($44 million) and hospital services ($44.2
million). Allied health services contributed an additional $8.5 million to the total costs (Access
Economics, 2007). Similar healthcare system costs are incurred in other developed nations such as
the United States of America (USA) and the United Kingdom (UK), where the projected yearly
costs of US $23 billion (Huse et al., 2005), and £599.3 million have been reported respectively
(Findley et al., 2003). PD, therefore, can be seen as a significant economic burden on the healthcare
system, individuals and their families.
3
1.3 Idiopathic Parkinson’s Disease
The aetiology of IPD is currently unknown. However, researchers believe that genetic
predisposition, certain exposure to environmental toxins and aging may be contributing factors to
the disease (Przedborski, 2007). A global estimate of IPD occurrence is also difficult, due in part,
to the different methodologies and diagnostic criteria used in epidemiological studies (de Rijk et al.,
1997). However, the prevalence of IPD in industrialised nations has been estimated at 0.3% of the
general population, and 1% of persons over the age of 60 years (de Lau & Breteler, 2006; de Rijk,
et al., 1997). Further, age-adjusted prevalence rates have ranged from 18 to 234 cases per 100 000
population and incidence rates from 1.9 to 22.1 cases per 100 000 population per year (Zhang &
Roman, 1993). It is acknowledged that comparisons using age-adjusted estimates are difficult, due
to the different populations and age ranges utilised in the various studies (de Lau & Breteler, 2006;
Zhang & Roman, 1993). However, despite the current limitations, epidemiological studies have
suggested that IPD occurs in all races (Zhang & Roman, 1993). In Australia, the prevalence rate of
IPD in 2005 was estimated at 290 cases per 100 000 population aged 55 to 64 years, and with
approximately 8 900 of cases diagnosed that year (Access Economics, 2007). As a result of the
expected growth in the country‟s aging population, the prevalence of IPD is predicted to rise by
15% in the next five years (Access Economics, 2007). Both males and females are equally prone to
developing the disease, with possibly a three-to-two male-to-female predominance (Fahn &
Przedborski, 2005; Zhang & Roman, 1993). For the majority of patients, the age of onset of IPD is
typically between the fifth to the seventh decade (Harding, 1993). However, young-onset IPD can
occur in patients who are in their early 50s or younger, with initial symptoms typically appearing
between the ages of 21 and 40 years (Muthane et al., 1994; Waters, 2005). This smaller group
makes up 5 to 10% of the IPD population (Golbe, 1991). In general, the survival rate of IPD is
approximately 12 years from symptom onset to death, although it is not uncommon for patients to
live with the disease beyond 20 years (Access Economics, 2007).
1.3.1 Neuropathology and Clinical Presentation of Motor Symptoms
IPD results from a significant degeneration of the dopaminergic neurons in the substantia
nigra of the midbrain and the components of the basal ganglia in the forebrain including the globus
pallidus, putamen and caudate nucleus (see Figure 1.1) (LeWitt, 2000; Murdoch, 1998; Przedborski,
2007). The presence of intracellular inclusions or Lewy bodies in many of the remaining neurons is
4
viewed as the pathological hallmark of the disease (Fahn & Przedborski, 2005; LeWitt, 2000). By
the time the clinical motor symptoms of IPD become apparent, the substantia nigra has already lost
approximately 60% of its dopamine neurons, and the resulting dopamine neurotransmitter levels in
the basal ganglia are already 80% below normal (Fahn & Przedborski, 2005). Consequently, the
reduced dopamine levels cause a disturbance in the basal ganglia control circuit (see Figure 1.2),
leading to the presentation of clinical motor symptoms (Duffy, 2005). With time, the dopamine
levels continue to decline, leading to the progressive worsening of the symptoms (Waters, 2005),
although this progression can vary between patients (Becker et al., 2002). Individuals with IPD
typically present with cardinal motor signs including resting tremor, cogwheel rigidity (increased
tone in the upper limbs during range of motion tasks), akinesia (inability to initiate movement),
bradykinesia (reduced speed of movement), hypokinesia (reduction of movement), and postural
disturbances (decline in postural reflexes resulting in significant freezing, shuffling and falling)
(Hurting, 2000; Marsden, 1989). A diagnosis of IPD is usually made when at least two of the
cardinal signs are present, one of which being tremor or bradykinesia (Fahn & Przedborski, 2005).
With the disease progression, the hypokinesia may also manifest as masked facial expression,
where the patient is seemingly unemotional (Duffy, 2005).
Figure 1.1 Components of the basal ganglia in the forebrain and the substantia nigra in the
midbrain.
5
Figure 1.2 Representation of the basal ganglia control circuit. Note. Blue lines represent excitatory
pathway; Red lines represent inhibitory pathway; Dashes represent indirect pathway.
1.3.2 Non-motor Symptoms
In addition to the motor symptoms associated with IPD, patients may experience a number
of non-motor symptoms during the course of the disease. These symptoms may include
disturbances in mood, cognition, sleep, autonomic function and sensation (Dewey, 2000; Tolosa &
Katzenschlager, 2007). Depression is a common mood disturbance that may develop in 40 to 60%
of people with IPD. The symptom is thought to relate to the neural loss and/or the patient‟s
psychological response to their disability (Dewey, 2000). Depression can be managed
pharmacologically. Cognitive disturbances may also be evident to varying degrees in IPD. These
changes may include decreased executive functioning, attention, emotional processing, visuospatial
ability and memory loss (Lewis, Dove, Robbins, Baker, & Owen, 2003; Owen, 2004; Zgaljardic,
Borod, Foldi, & Mattis, 2003). Although less common, dementia may be present in the later stages
of the disease in approximately 10 to 30% of cases (Dewey, 2000).
6
1.3.3 Speech and Voice Disturbances
Along with the clinical motor symptoms of IPD, patients may experience specific speech
and voice deficits that are the flow on effects of the disturbed basal ganglia control circuit serving
the cortical motor areas (Duffy, 2005; Murdoch, 2001). The motor speech impairment associated
with IPD is known as hypokinetic dysarthria and is commonly characterized by: (1) reduced
loudness; (2) monopitch and monoloudness; (3) a rough (harsh) and breathy vocal quality; (4)
imprecise articulation; (5) reduced stress; (6) variable rates of speech including accelerated or short
rushes of speech; (7) inappropriate silences; (8) reduced breath support for speech; (9) reduced
phonation duration; and (10) difficulties initiating speech (Chenery, Murdoch, & Ingham, 1988;
Critchley, 1981; Darley, Aronson, & Brown, 1969a, 1969b, 1975; Logemann, Fisher, Boshes, &
Blonsky, 1978; Ludlow & Bassich, 1983, 1984; Scott, Caird, & Williams, 1985). Hypokinetic
dysarthria can manifest in any or all of the respiratory, phonatory, articulatory and resonatory
subsystems of speech production, with the interplay of these subsystem impairments resulting in
further disturbances in prosody and overall speech intelligibility (Duffy, 2005). The effects of
rigidity, bradykinesia and hypokinesia are thought to be the main causes of the presenting speech
disturbances (Duffy, 2000; Ramig, Bonitati, Lemke, & Horii, 1994). The following sections will
further outline the contributions of the subsystem impairments to the perceptual features of
hypokinetic dysarthria.
1.3.3.1 Respiratory Subsystem
At the respiratory level, hypofunctioning of the respiratory muscles has been associated
perceptually with reductions in vocal loudness, phonation length and duration, and short rushes of
speech. Physiologically, some of these changes have been measured as: (1) reduced vital capacity;
(2) lower ribcage volume than controls; (3) irregular breath patterns; (4) reduced strength and
endurance of respiratory muscles; and (5) asynchrony of speech and respiration with in-
coordination of chest wall movement during speech (Duffy, 2005; Huber, Stathopoulos, Ramig, &
Lancaster, 2003; Murdoch, Chenery, Bowler, & Ingram, 1989; Solomon & Hixon, 1993; Theodoros
& Murdoch, 1998).
7
1.3.3.2 Phonatory Subsystem
Features of phonatory hypofunction such as: (1) reduced vocal fold adduction; (2) bowing;
(3) glottal chink; (4) laryngeal phase asymmetry; and (5) tremor, have been identified in patients
with IPD using telescopic cinelaryngoscopy and videostroboscopy (D. Hanson, Gerratt, & Ward,
1984; Perez, Ramig, Smith, & Dromey, 1996; M. E. Smith, Ramig, Dromey, Perez, & Samandari,
1995). Glottal chink has been associated perceptually with breathiness and reduced vocal loudness
(D. Hanson, et al., 1984), while laryngeal phase asymmetry with hoarse vocal quality (Murdoch,
Manning, Theodoros, & Thompson, 1997; Perez, et al., 1996). Acoustically, the lower fundamental
frequency (F0) ranges and F0 variability levels reported for IPD (King, Ramig, Lemke, & Horii,
1994; Ludlow & Bassich, 1984) have been associated with cricothyroid muscle rigidity and
perceptually, with monopitch speech (Aronson, 1990; Ramig, Pawlas, & Countryman, 1995).
Additionally, the combination of respiratory and phonatory difficulties has been associated with
lower mean sound pressure levels (SPLs) for individuals with IPD than age-matched controls on
speech and voice tasks (between 2 to 4 dB lower) (Fox & Ramig, 1997).
1.3.3.3 Articulatory Subsystem
Impairments at the articulatory level have included reductions in: (1) range; (2) accuracy;
(3) strength; (4) endurance of articulatory movements; as well as (5) tremor in the orofacial
muscles. These factors contribute to imprecise articulation predominantly affecting consonant
targets (Chenery, et al., 1988; Duffy, 2005; Forrest, Weismer, & Turner, 1989; Logemann & Fisher,
1981; Solomon, Robin, & Luschei, 2000; Theodoros & Murdoch, 1998). The disordered speech
production of plosives, fricatives and affricates that are commonly perceived as this level have been
closely associated with difficulties achieving complete closure of the articulators and vocal tract
constriction respectively (Canter, 1965; Logemann & Fisher, 1981). Inappropriate consonant
voicing has also been noted to occur in some individuals with IPD as a result of impaired
neuromuscular articulatory function or poor motor drive to the posterior cricoarytenoid muscle.
Although not predominant features of impairment at this subsystem level, vowel imprecision,
phoneme prolongations and voicing errors have also been identified in IPD (Theodoros & Murdoch,
1998).
8
1.3.3.4 Velopharyngeal Subsystem
Velopharyngeal function may also be affected by muscle rigidity, bradykinesia and

hypokinesia, and dysfunction can result in insufficient opening, closure or control of velar
movements (Hoodin & Gilbert, 1989b; Theodoros & Murdoch, 1998; Theodoros, Murdoch, &
Thompson, 1995). These features have been suggested to result in perceived hypernasality.
However, the extent to which hypernasality is typically present in IPD remains unclear. To date,
there have been mixed reports in the literature that have ranged in opinion from hypernasality being
the most predominant feature of hypokinetic dysarthria, to it being only a minor feature (Darley, et
al., 1975; Hoodin & Gilbert, 1989a, 1989b; Logemann, et al., 1978; Theodoros, et al., 1995). Due
to the variability in findings, hypernasality is currently not viewed as a perceptual hallmark of
hypokinetic dysarthria (Schulz & Grant, 2000).
1.3.3.5 Prosodic Features
In addition to the subsystem impairments, prosodic aspects of speech production can also
be disturbed in this population as an end result of impairment at the level of one or more subsystems
(Theodoros & Murdoch, 1998). Some of the specific impairments may include: (1) monopitch and
monoloudness; (2) reduced stress patterning; (3) difficulties initiating speech; (4) inappropriate
silences; (5) variable rates of speech including short rushes of speech or accelerated speech; (6)
decreased variation in syllable duration; (7) indistinct boundaries between syllables; (8) phoneme
repetition and palilalia (the repetition of utterances compulsively); as well as (9) increased pause
frequency and duration (Chenery, et al., 1988; Duffy, 2005; Kent & Rosenbek, 1982; Ludlow &
Bassich, 1983, 1984; Schulz & Grant, 2000; Theodoros & Murdoch, 1998; Zwirner & Barnes,
1992).
1.3.3.6 Overall Speech Intelligibility
The combined effects of subsystem and prosodic disturbances can lead to further deficits in
speech intelligibility and functional communication. Few studies have directly investigated speech
intelligibility in IPD. Coates and Bakheit (1997) found that 64.6% of their 48 IPD participants (all
9
on a consistent drug regimen for IPD) demonstrated reduced intelligibility on word and sentence
tasks, which were predominantly mild reductions in speech intelligibility (in 50% of the group).
Despite the mild reductions, 71% of the participants reported difficulties being understood by
unfamiliar listeners, and further, 61% of them found it hard to communicate with family and
friends. Recently, N. Miller et al. (2007) compared the speech intelligibility of people with IPD
(n = 125; medicated) with age-matched controls (n = 97). Similarly, 69.6% of their participants
with IPD were less intelligible than controls on word, sentence and conversational tasks.
Additionally, 76% of all participants with IPD in the study rated their voices as worse than pre-IPD,
and 38% also reported their speech intelligibility to be one of their top four concerns relating to the
disease (N. Miller, et al., 2007). Overall, it would appear that there seems to be a general trend to
suggest that people with IPD demonstrate greater reductions in overall speech intelligibility than
their healthy peers.
1.3.4 Occurrence of Hypokinetic Dysarthria
The speech and voice disturbances associated with hypokinetic dysarthria and IPD are
common causes of disability, with as many as 50 to 90% of individuals affected during the course
of their disease (Hartelius & Svensson, 1994; Ho, Iansek, Marigliani, Bradshaw, & Gates, 1998;
Logemann, et al., 1978; Ramig, Fox, & Sapir, 2004). The severity of the hypokinetic dysarthria has
also been suggested to increase with disease progression (Hartelius & Svensson, 1994). The
general presentation of hypokinetic dysarthria has been reported to begin with impairments at the
respiratory and phonatory subsystems in the early stages of IPD, with perceived changes in vocal
quality. This is followed by deficits in articulation, fluency, prosody and speech intelligibility with
disease progression (Ho, et al., 1998; Logemann, et al., 1978; Sapir et al., 2001). Logemann and
colleagues (1978) specified the rate of occurrence of hypokinetic symptoms in their participants
with idiopathic and postencephalitic PD (N = 200; non-medicated). Eighty-nine percent of
participants presented with voice disorders of roughness, breathiness, hoarseness and tremor.
Impairments in articulation (45% of participants) and speech rate (20% of participants) were less
frequent. Additionally, 45% of participants in the cohort presented solely with voice impairments,
while for another 45% of participants, articulation difficulties accompanied voice difficulties. A
similar presentation of speech and voice difficulties was also reported by Ho and colleagues (1998).
10
1.3.5 Sensory Perception Difficulties
In addition to the speech and voice presentation of hypokinetic dysarthria, people with IPD
may show signs of sensory perception difficulties that can affect their ability to recognize and
appropriately monitor their disordered speech production. Clinically, it is not uncommon for
patients to perceive their vocal loudness and speech intelligibility levels as adequate for
communication, when in fact, these levels are reduced and family members and friends often ask
the individuals to repeat themselves (Ramig, Pawlas, et al., 1995). A study by Ho and colleagues
(2000) confirmed that participants with IPD and moderate to severe hypokinetic dysarthria (n = 15)
had greater reductions in mean vocal SPLs during reading and conversational tasks than the age-
matched controls (n = 15). Yet the IPD participants themselves believed to be louder on these tasks
than they actually were (the actual mean differences in SPL between task performance and
perception of loudness on these tasks were not specified). It was also interesting to note that the
participants with IPD were able to increase their vocal SPL when explicitly cued to speak loudly
during reading and conversation. Consequently, it was suggested that people with IPD are able to
increase or regulate their vocal output appropriately when consciously focused on the task (Ho,
Bradshaw, Iansek, & Alfredson, 1999). This mismatch in performance between external and
internal cueing in IPD has been related to some extent to the effects of bradykinesia and
hypokinesia (Berardelli, Rothwell, Thompson, & Hallett, 2001; Ramig, Countryman, Thompson, &
Horii, 1995; Spielman, Ramig, Mahler, Halpern, & Gavin, 2007). The sensory perception
difficulties can have a great impact on patient motivation and treatment progress, as patients may be
resistant to increasing vocal loudness and generalizing treatment effects to everyday speech (Fox,
Morrison, Ramig, & Sapir, 2002; Fox & Ramig, 1997). In fact, some people with IPD may believe
that their spouses are hard of hearing or non-attentive rather than being aware of the extent of their
own communication difficulties (DePippo, 1994; Stewart, 2000).
1.3.6 Impact of Hypokinetic Dysarthria on Communication
The speech and voice difficulties associated with IPD can have a direct impact on the
person‟s ability to communicate effectively (DePippo, 1994). In the early stages of the
communication impairment, people with IPD may occasionally notice difficulties in being heard on
the telephone. With time, and as the speech and voice difficulties become more pronounced,
individuals may experience greater vocal fatigue with prolonged use, and increased difficulties
11
being understood in everyday circumstances, especially if not within close proximity to the listener
(DePippo, 1994). Consequently, individuals may be asked to repeat themselves more frequently,
which in turn, makes the interaction frustrating for them and their communicative partners
(DePippo, 1994; Stewart, 2000).
Along with hypokinetic dysarthria, the effects of reduced facial expression can further
impact on the quality of the communicative interaction. People with IPD may exhibit masked facial
features due to hypokinesia that reduce their ability to convey emotion and appropriately signal to
their communicative partners that they are attentive and engaged (DePippo, 1994; Stewart, 2000).
Moreover, individuals may have more limited ability to gesture as a result of muscle rigidity, and
eye contact may also be more difficult to maintain due to stooped forward posture. Together, these
factors may also impact on communication (Robertson, 1988).
The effects of the overall verbal and non-verbal communication difficulties can, over time,
make routine activities such as engaging in a conversation laborious and frustrating for individuals
with IPD, their families and friends (DePippo, 1994; Stewart, 2000). This may lead to a shift in the
interaction dynamics for people with IPD, from active participation to reduced talking and the
adoption of a more passive listening role (Stewart, 2000). As a result, individuals may lose
confidence in their ability to contribute to the conversation (Argue, 2000). With the gradual
worsening of symptoms, people with IPD may become more socially withdrawn and increasingly
dependent on their spouses to communicate their needs (N. Miller, Noble, Jones, Allcock, & Burn,
2008; Scott, 1991; Stewart, 2000). This negative cycle can bring about a sense of hopelessness and
loneliness that can have a negative impact on the person‟s quality of life (Argue, 2000; Ramig, et
al., 2004). Schrag et al. (2000) found that all IPD participants in their study (N = 97) demonstrated
impairments on quality of life measures to a greater extent than the general population.
Impairments were noted regardless of sex and gender and increased with disease severity.
Furthermore, physical and social functioning were the areas that impacted the greatest on quality of
life (Schrag, et al., 2000). In other studies, impaired communication has been viewed by
individuals with IPD as one of the most concerning factors of the disease (Argue, 2000; N. Miller,
et al., 2007). Additionally, it is not uncommon for younger people with IPD, who rely more heavily
on their communication for daily living, to notice the impact of the disorder earlier on than older
individuals with IPD (Duvoisin & Sage, 1996; Stewart, 2000). In many cases, younger people with
12
IPD may find it difficult to maintain employment as a result of their communication impairment,
thus creating significant distress for them and a negative impact on quality of life (Stewart, 2000).
1.4 Parkinson-plus Syndromes
PPS are the second most common group of PD after IPD. The syndromes encompass a
group of idiopathic disorders that include progressive supranuclear palsy (PSP), multiple system
atrophy (MSA), dementia with Lewy bodies and corticobasal degeneration (Dewey, 2000; Waters,
2005). Patients usually present with symptoms in addition to the classical signs of IPD (Stacy,
2000; Waters, 2005). However, they are often misdiagnosed as having IPD in the early stages of
the disease, when the distinguishing symptoms are not evident (Dewey, 2000; Jankovic, 1989).
Individuals with PPS may experience similar communication difficulties to those with IPD as
mentioned above. As PSP and MSA are the only two subgroups of the PPS to be specifically
referred to in this thesis, they are the only ones to be reviewed in this Chapter.
1.4.1 Progressive Supranuclear Palsy
The neuropathology of PSP is generally more diffuse than in IPD, and involves additional
components of the basal ganglia, other subcortical structures and occasionally the cortex (Stacy,
2000). The globose (round) neurofibrillary tangles found in the subcortex are viewed as the
pathological hallmark of PSP (Stacy, 2000). As a result of the diffuse pathology, the additional
clinical symptoms of PSP usually include: (1) a down-gaze paresis that is viewed as the hallmark of
the disease; (2) severe gait disturbances; (3) tremor to a lesser extent than IPD; and (4) axial rigidity
of the head and neck (Stacy, 2000; Waters, 2005). Patients with PSP may also be more emotionally
labile than those with IPD, and display sudden emotional outbursts such as crying or laughing
inappropriately to the context (Stacy, 2000). The symptoms of PSP generally present in patients
around 60 to 70 years of age, and the prognosis is often worse than IPD, with motor and cognitive
difficulties appearing earlier and declining more rapidly. Most patients become wheelchair bound
within nine years after the initial presentation of symptoms and unfortunately, death soon follows
(Stacy, 2000).
13
As PSP is a less common form of PD, few studies have documented the associated speech
and voice impairments that occur in this population. Consequently, it is difficult to describe the
typical presentation of the dysarthria, although it is generally accepted that patients demonstrate
more severe speech and voice deficits than those with IPD (Countryman, Ramig, & Pawlas, 1994).
Kluin and colleagues (1993) found that all 44 participants with PSP in their study exhibited mixed
dysarthria, with combinations of hypokinetic, spastic and ataxic elements. Spastic quality was the
most prominent (evident in 50% of participants), followed by hypokinetic (34%), ataxic (14%) and
equal involvement of spastic, hypokinetic and ataxic aspects (2%). Other reported speech and voice
features of PSP have included reduced vocal loudness, hoarseness, monopitch, articulatory
imprecision, stuttering and palilalia (Greene, 2005; W. Hanson & Metter, 1980; Kluin, et al., 1993;
Metter & Hanson, 1991).
1.4.2 Multiple System Atrophy
MSA represents a number of disorders of similar neuropathology, with extensive neural

loss or gliosis predominantly in the basal ganglia, other subcortical areas, and the cerebellum
(Bower, 2000). The pathological hallmarks of MSA are glial cytoplasmic inclusions that stain
positive with α-synuclein (Waters, 2005). As a result of the diverse pathology, MSA can present as
different contributions of PD features, spasticity, cerebellar ataxia, and autonomic disturbances
(Bower, 2000). The clinical symptoms of MSA typically appear in patients around 50 to 60 years
of age (Bower, 2000; Waters, 2005). Cognitive difficulties may also present in around 22% of
cases and typically as mild to moderate impairment (Bower, 2000). Similarly to PSP, the prognosis
for MSA is worse than IPD, with faster deterioration, and death generally occurring around nine
years after initial symptom presentation (Waters, 2005). The speech and voice difficulties
associated with MSA are also typically more severe than those with IPD (Quinn, 1989). Although
hypokinetic dysarthria remains the predominant disorder, it can also be perceived in combination
with ataxic and/or spastic dysarthria (Duffy, 2005). Other reported speech and voice characteristics
of MSA have included reduced vocal loudness, monopitch, imprecise articulation and respiratory
stridor, particularly during sleep (Quinn, 1989).
14
1.5 Management of Parkinson’s Disease
Despite medical advances, no current treatment is able to stem the disease progression in
people with PD (Becker, et al., 2002). As such, the various treatments available aim to alleviate the
symptoms associated with PD and maintain the patient‟s functional independence (Waters, 2005).
The following sections outline the current pharmacological, surgical and speech and voice
treatments for IPD and PPS.
1.5.1 Pharmacological Management
Pharmacotherapy remains the predominant treatment for IPD, although the initial benefits
generally decrease with chronic use (Waters, 2005). Levodopa is the most effective treatment in
managing the motor symptoms associated with IPD (rigidity, bradykinesia and tremor), and is
metabolised to dopamine by the brain (Tolosa & Katzenschlager, 2007). Used chronically,
however, levodopa may have substantial side effects such as: (1) “wearing-off effects” of
medication; (2) “on-off” fluctuations (spontaneous motor fluctuations due to reduced action of
dose); (3) levodopa-induced dyskinesias (involuntary movements); and (4) freezing during the “on”
or “off” states (Waters, 2005). Unfortunately, the benefits of levodopa therapy for patients with
PPS are greatly limited, due to the more extensive degeneration associated with the syndromes
(Stacy, 2000). Typically, only around one third of patients show any improvements with levodopa,
and these improvements are generally short-lived (Bower, 2000; Stacy, 2000). Overall,
pharmacological treatments for patients with IPD and PPS have not shown substantial and
consistent improvements in the associated hypokinetic dysarthria, and consequently, the
accompanying speech and voice difficulties continue to increase with disease progression (Ramig,
1998; Schulz & Grant, 2000).
1.5.2 Neurosurgical Management
Until the late 1960s, neurosurgical intervention for IPD such as pallidotomy and
thalamotomy was viewed as the treatment of choice in relieving motor symptoms (Waters, 2005).
With the advent of levodopa therapy in the 1960s, however, the use of such procedures rapidly
15
decreased. Recently, there has been renewed interest in neurosurgical procedures, spurred on by the
persistent motor difficulties associated with chronic levodopa use, as well as the advances in these
procedures (Danish & Baltuch, 2007; Waters, 2005). Pallidotomy, thalamotomy and deep brain
stimulation (DBS) procedures attempt to modulate the neural over-activity in the control circuits of
the thalamus and basal ganglia (Marsden & Obeso, 1994). These procedures are performed via
stereotactic surgery which is a form of surgery that utilises three dimensional coordinate systems to
enhance the accuracy of locating the target brain areas for intervention. Despite the success of these
procedures, pre-morbid motor function is not attainable and the majority of patients continue to
manage their condition with medication even after surgery (Goetz, Poewe, Rascol, & Sampaio,
2005; Shannon, 2000; Walter & Vitek, 2004; Waters, 2005).
During a thalamotomy, a lesion is made in the ventrolateral nucleus of the thalamus, which
has been found to successfully alleviate tremor in up to 90% of IPD patients (Shannon, 2000), and
to also provide some additional relief from rigidity and dyskinesias (Krauss & Grossman, 2007).
There have been, however, minimal improvements reported in bradykinesia, gait and balance
following this procedure (Krauss & Grossman, 2007; Shannon, 2000). Pallidotomy involves
lesioning the posteroventral portion of the globus pallidus of the basal ganglia (Shannon, 2000).
Improvements have also been found in rigidity, tremor and bradykinesia (in up to 70% of IPD
patients), as well as additional relief from dyskinesia (in more than 80% of IPD patients) (Shannon,
2000; Waters, 2005). However, improvements in balance and gait are variable and often short-lived
(Waters, 2005).
Unlike the permanent effects of pallidotomy and thalamotomy, DBS is a reversible

procedure, where an electrode is implanted in the target area. This is most often the subthalamic
nucleus, globus pallidus or ventrolateral nucleus of the thalamus (Shannon, 2000; Waters, 2005). A
neurostimulator is positioned beneath the skin and sends impulses to the electrode which modulates
the neural activity in the area. The frequency settings can be programmed and then activated by the
patient using a handheld magnet (Shannon, 2000). DBS works best for IPD patients who continue
to benefit from pharmacological management, but who suffer from drug-related dyskinesias and
motor fluctuations (Halpern & Hurtig, 2007). In their systematic review of the literature, Hamani
and colleagues (2005) found that IPD patients typically improved in rigidity (in 63% of cases),
tremor (81%), bradykinesia (52%), gait (64%) and postural stability (69%) following subthalamic
16
nucleus stimulation. Unfortunately, patients with PPS have shown little benefits from neurosurgical
intervention (Bower, 2000; Stacy, 2000).
1.5.2.1 Speech and Voice Changes Associated with Neurosurgical Management
Unlike the relief in motor symptoms obtained with neurosurgery, the associated effects on
hypokinetic dysarthria have been inconsistent and less than promising. Pallidotomy studies have
documented: (1) minimal or no overall improvements in functional speech (Baron et al., 1996;
Farrell, Theodoros, Ward, Hall, & Silburn, 2005; Mourao, Aguiar, Ferraz, Behlau, & Ferraz, 2005;
Schulz, Peterson, Sapienza, Greer, & Friedman, 1999); (2) some gains in labial control and
movement (S. Barlow, Iacono, Paseman, Biswas, & D'Antonio, 1998); as well as (3) some
improvements in mean SPLs which were predominantly noted for IPD participants with mild
dysarthria (Schulz, Greer, & Friedman, 2000). Additionally, Farrell and colleagues (2005) revealed
that on the whole, the 22 participants with IPD in their study who received pallidotomy,
thalamotomy, or DBS (surgical IPD), did not show substantial gains in perceptual speech, voice or
overall intelligibility. Although the abovementioned studies were small-scale, the findings
suggested that overall improvements in hypokinetic dysarthria following neurosurgical intervention
were limited. Together with the limited improvements with pharmacological management, it is the
case that behavioural management of hypokinetic dysarthria remains the most effective approach
for PD.
1.5.3 The Lee Silverman Voice Treatment® and Idiopathic Parkinson’s Disease
To date, the most effective behavioural treatment for improving the perceptual speech and
voice characteristics associated with mild to moderate hypokinetic dysarthria and IPD is the LSVT®
(Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ramig & Fox, 2004; Ramig, Pawlas, et al.,
1995). The LSVT® is an intensive programme that is delivered over four weeks (one-hour sessions,
four days a week), with a direct focus on improving vocal loudness by performing all treatment
tasks (daily variables and hierarchical speech loudness drills) at high intensity and with maximum
effort. Please refer to Chapter 4, section 4.2.5 for a detailed description of the treatment tasks and
the LSVT® delivery. Training loud phonation helps increase respiratory and phonatory muscle
drive, thereby reducing the effects of hypokinesia, bradykinesia and muscle rigidity (Fox, et al.,
17
2002; Fox et al., 2006; Ramig, et al., 1994; Ramig, Countryman, et al., 1995). The additional focus
on sensory awareness training or “calibration” with the LSVT® assists individuals to overcome the
sensory perception difficulties associated with IPD, through improved self-monitoring and
consistent use of the louder voice in daily communication (see Figure 1.3) (Fox, et al., 2006;
Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ramig & Fox, 2004; Ramig, Pawlas, et al.,
1995). The LSVT® design is also in keeping with the principles that encourage neural plasticity
(the reorganisation and change of neural functioning as a response to new conditions), such as
intensive treatment delivery, complexity of movements and saliency of therapy tasks (Fox, et al.,
2006).
Figure 1.3 A schematic representation of the reduced sensory perception of effort, scaling and
amplitude of output in PD and the LSVT® programme targeting improvements in these areas.
Note. From “Treatment of Speech and Voice Problems Associated with Parkinson‟s Disease,” by L.
Ramig, 1998, Topics in Geriatric Rehabilitation, 14(2), p. 34. Copyright 1998 by Aspen Publishers,
Inc. Adapted with permission.
18
1.5.3.1 The Effects of the Lee Silverman Voice Treatment® at the Respiratory and Phonatory
Subsystems
The efficacy of the LSVT® programme has been extensively documented in face-to-face
literature, for participants with IPD predominantly with mild to moderate hypokinetic dysarthria.
Acoustically, significant improvements in mean SPL have been reported post-LSVT® on a variety
of speaking tasks (maximum sustained vowel phonation, reading and monologue) (El Sharkawi et
al., 2002; Liotti et al., 2003; Ramig, Countryman, et al., 1995; Ramig et al., 2001; Ramig, Sapir,
Fox, & Countryman, 2001; Sapir, Spielman, Ramig, Story, & Fox, 2007; M. E. Smith, et al., 1995;
Ward, Theodoros, Murdoch, & Silburn, 2000). Of particular interest are the studies comparing the
results of the participants who received the LSVT® to those who received an alternate intensive
respiration treatment (RT). The LSVT® groups consistently achieved greater post-treatment gains
in mean SPL on more tasks than the RT groups (Ramig, Countryman, O'Brien, Hoehn, &
Thompson, 1996; Ramig, Countryman, et al., 1995; Ramig, Sapir, Countryman, et al., 2001). The
differences in results between the two approaches indicated that intensive treatment targeting both
respiratory and phonatory drive (LSVT®) was needed to generate the greater changes in mean SPL
(Fox, et al., 2002; Ramig, Countryman, et al., 1995; Ramig, Sapir, Countryman, et al., 2001). The
additional focus on calibration of the loud phonation with the LSVT® was further crucial in the
long-term maintenance of SPL, with post-treatment gains maintained above baseline for up to 24
months without additional treatment (Ramig, et al., 1996; Ramig, Countryman, et al., 1995; Ramig,
Sapir, Countryman, et al., 2001). Ramig, Sapir, Fox, et al. (2001) further documented treatment
specific improvements with the LSVT®, where only the LSVT® participants in their study (n = 14)
showed significant post-treatment changes in mean SPL on all tasks (maximum sustained vowel
phonation, reading, monologue and picture description), which were maintained above baseline at
six months follow-up. In contrast, the control groups consisting of untreated IPD participants (n =
15) and age-matched controls (n = 14) did not demonstrate significant changes in mean SPL on the
same tasks at assessment. Overall, the improvements in mean SPL have been attributed to greater
respiratory muscle drive and more efficient vocal fold adduction with the LSVT® (Ramig, et al.,
1994; Ramig, Countryman, et al., 1995; Ramig, Pawlas, et al., 1995).
Together with changes in mean SPL, significant post-LSVT® improvements have been
measured on the additional acoustic parameters of maximum duration of sustained vowel
phonation, maximum F0 range on vocal glides, and mean F0 and F0 variability during reading and
19
monologue tasks (El Sharkawi, et al., 2002; Liotti, et al., 2003; Ramig, et al., 1994; Ramig, et al.,
1996; Ramig, Countryman, et al., 1995; Ramig, Sapir, Countryman, et al., 2001). The treatment
gains on these parameters were also greater for the LSVT® compared to the RT groups, and post-
treatment improvements with the LSVT® were successfully maintained above baseline for up to 24
months without additional treatment (Ramig, et al., 1994; Ramig, et al., 1996; Ramig, Countryman,
et al., 1995; Ramig, Sapir, Countryman, et al., 2001). Improvements in sustained phonation with
the LSVT® have been attributed to greater respiratory drive and vocal fold adduction, while
increases in F0 range and F0 variability relate to the more efficient cricothyroid and thyroarytenoid
muscle movements (Ramig, et al., 1994).
Aerodynamic changes with the LSVT® have been reported for participants with IPD.
These changes have included increased subglottal air pressure and maximum flow declination rate
for the 17 participants in Ramig and Dromey‟s (1996) study, as well as the additional increased
laryngeal airway resistance and decreased open quotient for the single participant in Dromey et al.‟s
(1995) study. In this latter study, the treatment specific changes were also appropriately maintained
at 12 months follow-up (Dromey, et al., 1995). Greater vocal fold efficiency with the LSVT® has
also been confirmed with videostroboscopy (Dromey, et al., 1995; M. E. Smith, et al., 1995). In
particular, Smith and colleagues (1995) found that only the LSVT® participants in their study
(n = 13) demonstrated improved vocal fold adduction post-treatment in the direction of normal, as
well as non-significant changes in supraglottic hyperfunction. In contrast, the participants treated
with the alternate RT (n = 9) showed no post-treatment changes in laryngeal hypofunction (M. E.
Smith, et al., 1995).
Perceptually, changes in speech and voice with the LSVT® have also been documented at
the respiratory and phonatory levels. Improvements on parameters of vocal loudness, breathiness,
hoarseness and pitch variability during reading and monologue tasks have been identified in a
number of studies by assessing SLPs, participants and their families (Baumgartner, Sapir, & Ramig,
2001; Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Sapir et al., 2002). In line with the
acoustic measures above, post-treatment improvements in vocal breathiness and hoarseness,
loudness level and overall quality were also perceived to be greater for the participants who
received the LSVT® than those who received the RT (Baumgartner, et al., 2001; Sapir, et al., 2002).
Moreover, it was reported that vocal loudness and overall quality were maintained above baseline
by the LSVT® group only at 12 months follow-up (Sapir, et al., 2002).
20
1.5.3.2 The Global Effects of the Lee Silverman Voice Treatment®
In addition to the aforementioned findings, the benefits of the LSVT® have been
documented beyond the direct targets of respiration and phonation, to articulation, tongue function,
swallowing, prosody, speech intelligibility, facial expression and subcortical brain activity
(Dromey, et al., 1995; El Sharkawi, et al., 2002; Liotti, et al., 2003; Ramig, et al., 1994; Ramig, et
al., 1996; Ramig, Countryman, et al., 1995; Ramig, Sapir, Countryman, et al., 2001; Sapir, et al.,
2007; Spielman, Borod, & Ramig, 2003; Spielman, Ramig, & Borod, 2001; Ward, et al., 2000).
Specific improvements in articulation were noted on acoustic parameters of vowel and whole word
duration, extent and rate of movement, transition duration, rise time, frication duration and vowel
formants (Dromey, et al., 1995; Sapir, et al., 2007). Additionally, improvements in vowel formants
and the accompanying perceptual gains in vowel quality on specific targets were only identified for
the LSVT® participants (n = 14) and not the control groups (n = 15 untreated IPD participants;
n = 14 age-matched controls) in the study by Sapir et al. (2007). The associated benefits on the
articulatory subsystem have been suggested to result from improved drive to the orofacial structures
and greater sensory perception with loud phonation (Ramig, Sapir, Countryman, et al., 2001). The
hypothesis of improved muscle drive is supported by additional evidence of enhanced tongue
strength and endurance (Ward, et al., 2000), as well as tongue-base swallowing function post-
LSVT® (El Sharkawi, et al., 2002).
The effects of loud phonation with the LSVT® have also been noted beyond the speech
subsystems to prosody, where increased pause length and decreased utterance duration have
contributed to overall decreased speech rate, towards normal levels (Ramig, Countryman, et al.,
1995). Overarching improvements in speech intelligibility have also been perceived by assessing
SLPs, participants and their families, and attributed to the follow-on effects of loud phonation
(El Sharkawi, et al., 2002; Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ward, et al.,
2000). Moreover, it has been reported that the combined effects of the speech and voice
improvements with the LSVT® have had substantial impact on participants and their approach to
communication. Individuals with IPD have reported subsequent increased confidence in their
communicative abilities, where they were more likely to initiate conversation, have a more positive
attitude to interacting with others and overall, perceived their communication skills as improved
following treatment (Ramig, et al., 1994; Ramig, et al., 1996; Ramig, Countryman, et al., 1995;
Ramig, Sapir, Countryman, et al., 2001). Participants have also been reported to show greater facial
21
expressivity, mobility and engagement during conversational speech following the LSVT® than the
RT (Spielman, et al., 2003; Spielman, et al., 2001). Collectively, these post-treatment findings of
greater ease of communication for individuals with IPD are promising and may suggest a likely
positive impact on quality of life as well.
On the whole, the findings of improved performance with the LSVT® across all levels of
speech production have been directly attributed to the “global effects” of training loud phonation.
Specifically, the LSVT® is thought to activate neural mechanisms across the speech system, thereby
enhancing the drive and coordination of all related structures (Dromey & Ramig, 1998; Fox, et al.,
2002; Fox, et al., 2006; Ramig, et al., 1994; Sapir, et al., 2007; Spielman, et al., 2003). Further
evidence of the global effects of loud phonation has emerged, connecting the benefits of the LSVT®
to concomitant neurological changes in subcortical function at the level of the basal ganglia.
Through the use of positron emission tomography, Liotti and colleagues (2003) identified a
decrease in abnormal regional cerebral blood flow to cortical motor-premotor areas with the
LSVT®, and an increase in subcortical flow to areas such as the basal ganglia and anterior insula of
the cerebral cortex. These post-treatment changes were similar to the pattern of presentation of
healthy controls, thus indicating the possible normalisation of activity with the LSVT ® (Liotti, et
al., 2003). Additionally, the findings have suggested changes in speech-motor activity and
functional reorganisation from effortful cortical control of speech to a more automatic regulation by
the basal ganglia and anterior insula (Liotti, et al., 2003; Spielman, et al., 2003). There is
recognised need for additional investigations using larger participant sizes, severity levels and
instrumental examinations to further determine the global effects of the LSVT®. However, the
general findings suggest that overall, the intensive voice treatment is beneficial in targeting the
underlying deficits associated with IPD and promoting functional communication and long-term
independence for this population.
1.5.3.3 The Lee Silverman Voice Treatment® and Surgical Idiopathic Parkinson’s Disease
The LSVT® has also been used to treat the dysarthric speech and voice in persons with IPD
who had undergone neurosurgical intervention for the management of their motor symptoms
(surgical IPD). However, the treatment gains have been more variable for this more complex
population (detailed further in Chapter 5). In a study by Ward and colleagues (2000), the surgical
22
IPD participants with mild to severe hypokinetic dysarthria who had undergone pallidotomy and/or
thalamotomy (n = 12), demonstrated significant post-treatment improvements in mean SPL during
maximum sustained vowel phonation and reading tasks, as well as improvements in sentence
intelligibility. Despite these improvements, the treatment gains were lower than those of the non-
surgical IPD group in the same study (n = 18, with mild to severe hypokinetic dysarthria).
Furthermore, only the non-surgical IPD group showed substantial gains in tongue strength and
endurance, possibly indicating that generalisation of the LSVT® effects to the articulatory
subsystem was more difficult for the surgical IPD group (Ward, et al., 2000).
Two additional case studies have documented post-LSVT® improvements for surgical IPD
participants (detailed further in Chapter 5). In one of the studies, the participant had undergone
thalamotomy and pallidotomy and presented with severe hypokinetic dysarthria (Theodoros,
Thompson-Ward, Murdoch, Lethlean, & Silburn, 1999). In the other study, the participant had
undergone bilateral thalamotomy and exhibited mixed hypokinetic-spastic dysarthria of moderate
impairment (Countryman & Ramig, 1993). Both studies reported post-treatment improvements on
SPL tasks (maximum sustained vowel phonation and reading) and maximum duration of sustained
vowel phonation. However, it was generally noted that these gains were lower than those typically
expected for non-surgical IPD participants with mild to moderate hypokinetic dysarthria as reported
in the LSVT® efficacy studies (Ramig, et al., 1994; Ramig, Countryman, et al., 1995).
Perceptually, improvements were noted to some extent on a number of parameters including speech
intelligibility, loudness level, pitch variability, and articulatory precision. In general, however, the
initial post-LSVT® gains were not maintained by the participants and only some of the parameters
remained above baseline at six and 12 months follow-up (Countryman & Ramig, 1993; Theodoros,
et al., 1999).
Certain factors may have contributed to the poorer performance and long-term LSVT®
outcomes for these complex cases of PD mentioned here. Firstly, the participants were overall at
more advanced stages of PD and in some instances also experienced greater involvement of
hypokinetic dysarthria than the IPD participants with mild to moderate hypokinetic dysarthria
described previously in the efficacy studies. Secondly, mild to moderate cognitive difficulties were
also reported in one study as was noted disease deterioration in participants over the course of the
study (Countryman & Ramig, 1993). Collectively, such factors may have directly impacted on the
participants‟ ability to maintain the initial LSVT® gains (Countryman & Ramig, 1993; Theodoros,
23
et al., 1999; Ward, et al., 2000). Finally, the direct interference to the basal ganglia control circuit
(see Figure 1.2) with neurosurgery may have also impacted on the participants‟ ability to reach the
target loudness levels and then to self-monitor and maintain the initial gains long-term (Countryman
& Ramig, 1993; Theodoros, et al., 1999). The combination of such factors may have also
contributed to the lack of generalisation of the LSVT® effects to the articulatory subsystem as noted
previously (Ward, et al., 2000). Overall, in acknowledgement of the additional difficulties in long-
term maintenance incurred by participants following neurosurgical intervention, an extended
LSVT® programme of one to four weeks and follow-up therapy has been proposed for this complex
population to maximise treatment benefits (Countryman & Ramig, 1993; Theodoros, et al., 1999).
Further large-scale studies are warranted in order to standardise the number of additional treatment
sessions required.
1.5.3.4 The Lee Silverman Voice Treatment ® and Parkinson-plus Syndromes
The LSVT® programme has also been employed in the treatment of two participants with
PSP and MSA, presenting with moderate to severe hypokinetic dysarthria (Countryman, et al.,
1994). Both participants showed post-treatment improvements in mean SPL (maximum sustained
vowel phonation, reading and monologue) and perceived improvements in loudness level, slurring,
pitch variability and overall speech intelligibility as reported by the participants and their families.
However, similar to the surgical IPD participants mentioned above, the post-treatment gains for
these individuals with PPS were lower than those reported in the IPD efficacy studies.
Additionally, only some of the treatment gains were maintained above baseline at 6 months follow-
up and only for the participant with PSP (Countryman, et al., 1994). Further details on this
participant‟s performance are presented in Chapter 5. The more significant speech and voice
difficulties associated with PPS and the more rapid deterioration of physical function that were
reported during the study compared to IPD may have impacted on the treatment success and long-
term maintenance for these individuals. As recommended for the surgical IPD participants, an
extended LSVT® programme of at least a week, along with follow-up therapy would potentially
benefit patients with PPS (Countryman, et al., 1994). Further large-scale studies are needed in this
area.
24
1.6 Optimal Management of PD and Current Challenges to Speech-Language Pathology
Services
Four factors relating to speech-language pathology service delivery have been identified
are integral to the successful management of hypokinetic dysarthria associated with PD. Firstly,
early assessment of PD is essential. Assessments protocols specific to PD help quantify the speech
and voice difficulties associated with hypokinetic dysarthria and provide an indication of suitability
for the LSVT®. Secondly, the benefits of the LSVT® are maximised when the treatment is provided
as early intervention for people with PD who present with mild to moderate hypokinetic dysarthria
and before the speech and voice difficulties begin to impact on their quality of life (Countryman, et
al., 1994; Fox, et al., 2006; N. Miller, Noble, Jones, & Burn, 2006). Timely delivery of the LSVT®
is also particularly relevant for people with PD who rely on their communication skills to maintain
employment (Ramig, Pawlas, et al., 1995). Thirdly, the LSVT® must be delivered intensively, in
order to maximise generalisation of treatment effects to everyday speech (Spielman, et al., 2007).
Finally, the one-to-one participant-SLP interaction is ideal, as it allows the participant to maintain
high phonatory effort throughout the session, which is an important treatment requirement (Fox, et
al., 2006). However, despite these specific requirements for the optimal management of PD, to
date, patient access to speech-language pathology services for appropriate assessment and treatment
remains relatively limited.
Studies conducted in the UK and Sweden have identified reduced services for people with
PD, where relatively few participants surveyed (as little as 2 to 14.2%) had been able to access
speech-language pathology services for either assessment or treatment (Hartelius & Svensson,
1994; Mutch, Dingwall-Fordyce, Downie, Paterson, & Roy, 1986; Oxtoby, 1982; Peto, Fitzpatrick,
& Jenkinson, 1997). Despite the low access, it was identified that 49% of people with PD in
Oxtoby‟s (1982) study reported having speech difficulties and would have benefited from speech-
language pathology services. Not surprisingly, in Peto‟s (1997) study, 33% of people with PD who
presented with speech difficulties were also dissatisfied with their level of service access. Of
further relevance is Australian data where only 36.7% of the people with PD surveyed (N = 250)
reported having access to speech-language pathology services, and of these, only 24.1% had
actually been treated using the LSVT® (Grimm, et al., 2004). Certain barriers were identified by
Grimm and colleagues (2004) to impact on service assess for this population. Firstly, there was a
low priority placed on speech-language pathology services for PD in the public health system as
25
well as inadequate knowledge of services and their benefits by both the individual with PD and
referring healthcare professionals. Secondly, only a limited number of SLPs were currently trained
to administer the LSVT®. To date, only 148 SLPs have been trained in Australia (LSVT Global,
2010), a ratio of approximately one SLP for every 488 individuals with PD nationwide. With
population aging, this discrepancy will only continue to increase in future years, making it even
harder for individuals to access appropriate and timely speech-language pathology services.
Thirdly, the debilitating physical symptoms experienced by individuals were identified as additional
barriers to attending services. Lastly, transport and travel difficulties were also seen as access
barriers for 15.6 and 32.2% of participants surveyed respectively. These difficulties may be further
magnified for individuals with PD who no longer drive and rely on their spouses who may also be
elderly, for assistance with transportation. Cost of travel may also factor in as a barrier for some
individuals. Moreover, the combined effects of such barriers may have a profound impact on
people with PD residing outside metropolitan areas, as access to services may be even more
difficult. Grimm et al. (2004) reported that 50.2% of participants in their study lived in regional
areas and 11.6% in rural areas, with a greater number of these individuals also finding it harder to
access services than those in metropolitan areas. The geographic distribution of people with PD is
also similar to that of Australia‟s population, where approximately 36% of Australians live in
regional and remote areas of the country (Australian Bureau of Statistics, 2009a). In general, access
difficulties due to distance in Australia are significant.
Having acknowledged the aforementioned barriers, it is still imperative that all patients
receive the most effective speech-language pathology interventions, regardless of their geographic
location (Wilson, Lincoln, & Onslow, 2002). Furthermore, it has been proposed that investment in
early diagnosis and access to health services may help to lessen the burden of PD in Australia
(Access Economics, 2007). For this to be achieved, it has been recommended that access to allied
health services needs to increase to approximately 70% for people with PD, including greater access
to services for individuals residing in regional and remote areas (Access Economics, 2007). To
date, the disparity between the supply and demand of speech-language pathology services for
people with PD suggests the need for an additional or alternate mode of service delivery for this
population. One possible solution is the use of telerehabilitation whereby telecommunication and
information technologies are used in the delivery of healthcare at a distance.
26
1.7 Telerehabilitation Nomenclature
The term telerehabilitation describes the remote delivery of rehabilitation services via
telecommunication and information technologies (Rosen, 1999). Telerehabilitation services may
include: (1) assessment; (2) therapeutic intervention; (3) remote monitoring of treatment progress;
(4) the provision of educational support, training and networking opportunities to individuals with
disabilities and their families; as well as (5) professional development and networking for
healthcare professionals (Rosen, 1999). In the last decade, telerehabilitation has been added to the
nomenclature of distant healthcare delivery, to best represent the various healthcare disciplines such
as speech-language pathology, occupational therapy, physiotherapy and audiology that have
adopted the use of telecommunication and information technologies in their service delivery (Bauer
& Ringel, 1999; Maheu, Whitten, & Allen, 2001). Telerehabilitation stems from the terms
telemedicine and telehealth that are commonly utilized in the healthcare domain. Telemedicine was
first introduced in the 1970s and largely describes the distant delivery of medical care. Telehealth
was introduced later, in the 1990s as a more inclusive term to describe distant healthcare services
for persons who are not necessarily ill (Craig & Patterson, 2006; Maheu, et al., 2001; Stanberry,
1998a). The WHO (1997b) makes the distinction between the terms telemedicine and telehealth:
If telehealth is understood to mean the integration of telecommunication systems into the

practice of protecting and promoting health, while telemedicine is the incorporation of
these systems into curative medicine, then it must be acknowledged that telehealth
corresponds more closely to the international activities of the WHO in the field of public
health. It covers education for health, public and community health, health systems
development and epidemiology, whereas telemedicine is oriented more towards the clinical
aspect (WHO, 1997b).
Telemedicine, along with telerehabilitation and home telecare, or the distant delivery of
services directly to the individual‟s home, can therefore be seen as divisions of telehealth (Darkins
& Cary, 2000). Please see Figure 1.4. Craig and Patterson (2006) clarified further aspects relating
to telemedicine and its role within the healthcare system that also apply to telerehabilitation.
Telemedicine should not be viewed as a technology or considered a new division of medicine. It is
also not intended to substitute healthcare professionals or cure all health problems. Rather, a wide
range of technologies can be utilised by healthcare professionals to deliver distant heathcare (Craig
& Patterson, 2006). As a result, telemedicine can be provided as an additional or alternate form of
27
healthcare delivery, with a primary focus on improving access to and effectiveness of the current
health services (ANZTC, 2002; Craig, 1999; Craig & Patterson, 2006; Mun & Turner, 1999;
Perednia & Allen, 1995; Zundel, 1996). The following sections highlight the significant telehealth
applications that have helped to establish this mode of service delivery. A brief history of telehealth
use in Australia is also provided and a number of universal barriers to the integration of telehealth
into mainstream clinical practice are discussed. Following these sections, the term telerehabilitation
will be used throughout this thesis to denote rehabilitation services in speech-language pathology
delivered at a distance and in particular, the studies in this thesis.
Figure 1.4 A schematic representation of the associated telehealth, telemedicine, telerehabilitation

and home telecare terms.
1.8 The Development of Telehealth
Telehealth has undergone two significant phases of development after initial pioneering
efforts in distant healthcare communication. These phases have included the “birth” of modern
telehealth, with the introduction of the television and satellite communication, and the “rebirth” in
the 1990s, which has paved the way for current telehealth services. The driving forces behind the
evolution and development of telehealth services have been the communication and information
technologies available at the time (Bashshur, Reardon, & Shannon, 2000).
28
1.8.1 Early Telehealth Systems
The use of communication devices for transmitting healthcare information at a distance is

not new, rather it is the technology that has largely evolved (Darkins & Cary, 2000; Kolitsi &
Iakovidis, 2000). The early use of bonfires and flags to communicate medical information
regarding the bubonic plague in the Middle Ages was followed by the employment of the postal
service and the telegraph in the mid 19th century for patient management and physician diagnosis
(Craig & Patterson, 2006; Darkins & Cary, 2000; Zundel, 1996). Later, the telephone became a
milestone for modern telehealth with its invention in the late 19th century and still remains a main
communication tool today. In addition to standard verbal communication, the telephone network
was used in 1910 to transmit amplified stethoscope sounds and later, to transmit electrocardiograms
and electroencephalograms. At the end of the 19th century, radio communication became the next
landmark in telehealth development. This form of communication was a valuable source of medical
contact for seafarers and is still used extensively today (Craig & Patterson, 2006).
1.8.2 The Birth of Modern Telehealth
Two major events contributed to the foundation of modern telehealth: (1) the introduction
of the television and advances in computer technology; and (2) the National Aeronautics and Space
Administration‟s (NASA) interest in monitoring the health of astronauts in space via satellite-
communication (Bashshur, et al., 2000; Craig & Patterson, 2006; Zundel, 1996). Healthcare
delivery via the television was made possible by the use of a two-way closed-circuit microwave
television system. This application was first utilised in 1964 to provide direct communication
between the Nebraska Psychiatric Institute in Omaha and the Norfolk State Mental Hospital,
180 km away (Benschoter, Wittson, & Ingham, 1965). Later, in 1967, an additional two-way
audiovisual microwave circuit was used in Boston between Massachusetts General Hospital and
Logan International Airport Medical Station. This provided 24-hours-a-day medical care for airport
passengers and staff, as well as facilitated communication and training between professionals
(Murphy & Bird, 1974).
NASA‟s success in the space monitoring of astronauts via satellite technology led to the
development of telehealth programs for rural and remote populations, and disaster relief projects in
29
the USA. During the late 1950s, NASA, in collaboration with Lockheed and the USA Public
Health Service, initiated the STARPAHC program (Space Technology Applied to Rural Papago
Advanced Health Care). The program provided general medical services via satellite-based
communications to the people of the Papago Indian Reservation in Arizona (Fuchs, 1979; Maheu, et
al., 2001; Stanberry, 1998a). A subsequent project undertaken between 1971 and 1975 utilized the
Advanced Technology-6 communications satellite (ATS-6), and proved the viability of technology
use to deliver remote healthcare in Alaska (Foote, 1977). Disaster relief and coordination of rescue
operations using satellite technology soon followed both nationally and internationally. In 1985,
the ATS-3 satellite facilitated communication between the American Red Cross and the Pan
American Health Organization in earthquake stricken Mexico City (Garshnek & Burkle, 1999a,
1999b). Following this, NASA‟s “Space Bridge to Russia” project in 1988 enabled international
medical relief via satellite to earthquake stricken Armenia (Garshnek, 1991; Llewellyn, 1995).
Around the same time, the SatelLife/HealthNet program in 1985 facilitated the delivery of medical
services via satellite from metropolitan to rural areas within the Philippines, nine African counties
and three countries in the Americas (Ferguson, Doarn, & Scott, 1995).
The successful delivery of telehealth services at the time showed the potential of this mode
of service delivery for overcoming cultural, political, social and economic hurdles (Maheu, et al.,
2001). However, despite the promising results from early telehealth studies and the advances in
communication and information technologies, the great majority of projects initiated prior to 1986
were not able to be sustained long-term (Kuo, et al., 2001; Maheu, et al., 2001). Their failure was
greatly attributable to the excessive costs of the telecommunication infrastructure at the time as well
as the lack of sustained government funding (Bashshur, et al., 2000; Maheu, et al., 2001).
1.8.3 The Rebirth of Modern Telehealth
The rebirth of telehealth occurred during the 1990s and was driven by the social and
political push for telehealth services as well as the increasing affordability of telecommunication
infrastructure (Maheu, et al., 2001; Whitten, Frances, & Collins, 1997). In relation to the social
influence, the increasing concerns about unequal access to healthcare for rural and remote
populations and the rising costs of healthcare placed pressure in favour of using telehealth as an
alternative mode of service delivery (Maheu, et al., 2001). Political interest in telehealth arose from
30
the social push for healthcare reform. Consequently, a significant increase in government funding
occurred in the USA for telehealth projects and telecommunication infrastructure in rural and
remote communities (Maheu, et al., 2001; Weismer, Jeng, Laures, Kent, & Kent, 2001). The
success of these projects facilitated further interest in telehealth services and led to additional
project funding (Maheu, et al., 2001).
The advances in digital communications and the Internet were also significant to the
rebirth of telehealth (Kuo, et al., 2001). These advances permitted interactive PC-based
videoconferencing over lower-bandwidth Internet connections as cheaper alternatives to the costly
satellite options (Maheu, et al., 2001). The increased popularity of videoconferencing use has made
telehealth a practical option for modern healthcare (Kuo, et al., 2001; Mun & Turner, 1999). In
1993, with the rebirth of telehealth, 10 telehealth videoconferencing programs were active in the
USA and this number has roughly doubled each year (Maheu, et al., 2001). Of all the disciplines
within telehealth, teleradiology has been incorporated best into clinical practice, and so much so,
that it is now impossible to estimate the level of activity in this area. For all other disciplines, a
survey of telehealth activity in the USA in 2002 revealed that nearly 85 000 teleconsultations had
been undertaken in more than 30 disciplines. In this survey, the most active disciplines were
paediatrics, mental health, cardiology, dermatology and orthopaedics, with approximately half of
the applications using interactive videoconferencing (Grigsby, 2004). Worldwide, the most
proactive countries to conduct telehealth research to date have included Australia, Canada, Finland,
France, Germany, Greece, Hong Kong, Italy, Japan, Malaysia, the Netherlands, New Zealand,
Norway, Sweden, Switzerland, the USA and the UK (Darkins & Cary, 2000; Garshnek & Hassell,
1997).
1.9 Telehealth in Australia
Australia has been actively involved in telehealth projects since the 1960s (Garshnek &
Hassell, 1997). This continent has an area of approximately 7.7 million square kilometres and a
small population of 22.1 million inhabitants, distributed widely in many geographic locations
including major cities (64%), and regional and remote areas (36%) (Australian Bureau of Statistics,
2009a, 2009b, 2009c). Given its size and population distribution, the use of telecommunication and
information technology for distant healthcare services in this country is ideal. It has already been
31
identified that the greater the remoteness of hospitals in Australia, the higher the levels of telehealth
use by healthcare professionals (O'Shannessy, 2000; Wootton, Blignault, & Cignoli, 2003).
Like many countries including Canada, the USA and the UK, Australia faces comparable
challenges relating to current healthcare delivery. Consequently, the driving forces for change in
healthcare services and uptake of telehealth in Australia have included: (1) the rising costs of
healthcare and an ageing population; (2) increased consumer demands on healthcare; (3) pressures
to improve the quality of and access to healthcare services; (4) a shift from treatment to prevention
and care; (5) improved communication and information technologies; and (6) the need to export
quality healthcare services to South East Asia (Alexander, 1995; Crowe & McDonald, 1997). The
uptake of telehealth in Australia has been comparable to that of the rest of the world and largely
influenced by the communication and information technologies available at the time. Early systems
including the telegraph and Flying Doctor services progressed to the telephone, facsimile, radio and
satellite communication, to the current use of videoconferencing technologies which are now most
commonly PC-based (ANZTC, 2002; Blignault, 2000; Crowe & McDonald, 1997 Dillon &
Loermans, 2003; Mitchell, 1999). To date, all states and territories in Australia have employed
telehealth systems in a wide range of specialty areas. Teleradiology and telepsychiatry have
dominated the field since 1994 (ANZTC, 2002; Mitchell, 1999).
Although telehealth projects in Australia have demonstrated the potential to improve

healthcare access for all (Crowe & McDonald, 1997), current barriers to the successful integration
of this modality into mainstream healthcare do exist. The high costs of telecommunication
infrastructure (including satellite coverage and Integrated Services Digital Network) in areas of
rural and remote Australia is currently making it difficult to access the bandwidth capacity required
for videoconferencing communication and high quality audio and video transfer. This creates a
major setback in achieving the overall telehealth objective. Obtaining State and Federal
Government funding for telehealth services is an additional barrier to the development and long-
term sustainability of these services. Funding may improve once telehealth is viewed as a mode of
healthcare delivery that obtains funding from current available sources rather than from separate
funding (ANZTC, 2002).
32
Additional and universal barriers to the successful integration of telehealth into mainstream
healthcare have been identified to include: (1) reimbursement issues; (2) the cost-effectiveness of
the service; (3) ethical and legal aspects relating to medical malpractice; (4) patient privacy,
confidentiality and consent; (5) professional portability; as well as (6) patient and professional
satisfaction with the modality (Craig, Russell, Patterson, & Wootton, 1999; Crowe & McDonald,
1997; Jennett et al., 2003; Stanberry, 1998b; Theodoros & Russell, 2008; Williams, May, & Esmail,
2001). As the former five universal barriers are beyond the scope of the present thesis, they are not
reviewed in this chapter. However, professional and patient satisfaction with telehealth is reported
and discussed in this thesis, and as such, these areas are briefly reviewed below.
1.10 Satisfaction with Telehealth
It has been suggested that the level of professional and patient satisfaction with telehealth
may directly impact on their willingness to adopt this practice (Craig, et al., 1999). Further,
professional and patient satisfaction can also influence the uptake of telehealth at the level of the
healthcare service, as granting agencies and research institutes utilize satisfaction as an indicator of
patient and provider experience and quality of healthcare delivery (Collins & O'Cathain, 2003;
Gustke, et al., 2000; Owens & Batchelor, 1996). In order to facilitate the integration of telehealth
into mainstream healthcare, it is therefore necessary to acknowledge and address the possible issues
relating to professional and patient satisfaction with this modality (Darkins, 1996).
1.10.1 Professional Satisfaction
Across a variety of disciplines, numerous professional-related benefits of telehealth use

have been closely linked to satisfaction. Such benefits have included improved: (1) access to
patients, especially those residing in rural and remote areas; (2) time-effectiveness and efficiency of
service delivery; and (3) professional inclusion and development opportunities that may also lead to
the retention of more professionals in rural and remote areas (Brennan, Georgeadis, & Baron, 2002;
Cohn & Goodenough, 2002; Hicks et al., 2000). However, despite these benefits, telehealth is still
not readily embraced by all professionals (Kolitsi & Iakovidis, 2000). Some of the major concerns
voiced include difficulties in using the technology and potential changes to the patient-professional
dynamic and delivery of the consultation with telehealth.
33
1.10.1.1 Technology Concerns
Professionals may be concerned about the use of telehealth in healthcare due to their
limited experiences with such systems, as well as the currently few published studies on the validity
and reliability of this modality (Theodoros & Russell, 2008). The seemingly distant interaction
between the patient and professional with telehealth may concern some professionals, as there is a
real need for them to successfully integrate the use of the technology in clinical practice, to best
complement their senses and guide clinical decisions (Bashshur, et al., 2000). To successfully
incorporate telehealth into healthcare delivery, additional demands may be posed on professionals.
These may include the need to acquire new skills and expertise, use the equipment in a safe manner
and effectively troubleshoot where necessary, all while maintaining the highest quality services
(Kolitsi & Iakovidis, 2000). If not integrated successfully into clinical practice, professionals may
feel a level of technology anxiety with telehealth, which may lead to a further decrease in
confidence with the service (Hjelm, 2006). Therefore, to minimise some of the technology
concerns of professionals, certain factors relating to the development of telehealth systems may
need to be considered. For example, user friendly systems that require minimum troubleshooting
will be favoured by professionals above more complicated technologies, as these systems will best
assist professionals to effectively deliver online consultations with minimal distractions (Craig, et
al., 1999; Ghosh, McLaren, & Watson, 1997; Montani et al., 1996). Involving professionals in the
system design and testing will also help to promote the user friendliness of the technology
(Salvemini, 1999). With the effective delivery of online consultations, increased professional
confidence and acceptance of telehealth will continue, which will aid the adoption of this mode of
service delivery into routine practice (Yellowlees, 2006).
1.10.1.2 Patient-Professional Dynamic and Service Delivery
Additional concerns for professionals have been reported to relate to the potential changes
to the patient-professional dynamic and patient satisfaction with telehealth. In relation to the
former, there is some apprehension from professionals about their ability to effectively
communicate and engage with patients during the session, and to maintain an appropriate level of
rapport online. Of particular concern to professionals is their ability to portray and detect essential
verbal (e.g. sarcasm and intonation) and non-verbal cues (e.g. facial expression and mood) via
telehealth that are considered important for communication, rapport building, diagnosis and
34
management (Ghosh, et al., 1997; Hicks, et al., 2000; Malagodi & Smith, 1999). Additionally,
professionals may be concerned with the level of patient acceptance of telehealth use and the effects
of the perceived physical distance between themselves and their patients (Mashima et al., 2003).
Professionals may also be apprehensive about potential changes to the service provision with the
adoption of telehealth, when traditional face-to-face consultations are still considered to be the
“gold standard” (Darkins, 1996). Acknowledging these concerns, it is important that online systems
are developed with the capabilities to closely resemble the face-to-face consultation and to provide
professionals with the necessary information required for successful patient management (Bashshur,
et al., 2000). For example, the use of real-time videoconferencing with minimal audio and video
delays would promote the ease of communication between patients and professionals and assist
with the delivery of assessment and treatment, in keeping with the face-to-face modality (Malagodi
& Smith, 1999; Zarate et al., 1997). Combined, these factors would further promote professional
satisfaction with telehealth.
1.10.2 Patient Satisfaction
Various patient benefits with telehealth have contributed to growing patient support for the
uptake of this modality. These benefits have included: (1) improved service access; (2) shorter
waiting periods; as well as (3) reduced cost and travel time to services (Gustke, et al., 2000;
Whitehill, Lee, & Chun, 2002). Despite the potential benefits to patients, some individuals still
prefer the traditional mode of service delivery to telehealth (Mair, Whitten, May, & Doolittle,
2000). It has been identified that patients, like professionals, may feel apprehension regarding the
use of technology for their clinical management (Werner & Karnieli, 2003). Further, patient
satisfaction may be influenced by the perceived quality of the interaction between themselves and
the professional (Werner & Karnieli, 2003). Where telehealth alters the nature of the interaction,
for example, where the professional is distracted during the session by operating the technology,
patients may feel more aware of the physical distance between themselves and the professional.
This may lead to decreased confidence in being able to speak freely with the professional, and an
overall sense of dissatisfaction with the professional and telehealth service (Bashshur, et al., 2000;
Mair, et al., 2000; Mechanic, 1998; Werner & Karnieli, 2003). Therefore, to maximise patient
satisfaction with telehealth, it is important for professionals to be aware of the potential impact of
the technology use on the consultation. As mentioned previously, telehealth systems that are user-
friendly and allow for optimal patient-professional interactions in keeping with the face-to-face
35
modality would help to improve satisfaction, as would familiarising patients with the online
applications (Hill et al., 2006; Montani, et al., 1996; M. Waite, Cahill, Theodoros, Busuttin, &
Russell, 2006; Zarate, et al., 1997). It has further been suggested that patient confidence with the
service is likely to increase in circumstances where professionals are able to display appropriate
empathy, warmth, confidence, interest and eye contact through their manner of consultation
delivery (Hughes, 2001; Mechanic, 1998; E. Miller, 2001).
1.11 Telerehabilitation Studies in Speech-Language Pathology
Telerehabilitation as a mode of service delivery is presently in its infancy compared to

telehealth. Relatively few published studies have validated this approach within the disciplines of
speech-language pathology, physiotherapy, audiology and occupational therapy. Specifically,
speech-language pathology research has primarily involved small-scale proof-of-concept studies
that have touched on the benefits of online assessment and treatment of adult and paediatric cases.
Adult research in particular has been conducted in:
(1) motor speech disorders (Duffy, Werven, & Aronson, 1997; Hill, et al., 2006; Hill,
Theodoros, Russell, & Ward, 2009a, 2009b; Palsbo, 2007; Theodoros, Russell, Hill, Cahill,
& Clark, 2003; Vaughn, 1976);
(2) language disorders (Brennan, Georgeadis, Baron, & Barker, 2004; Duffy, et al., 1997;
Georgeadis, Brennan, Barker, & Baron, 2004; Helm-Estabrooks & Ramsberger, 1986; Hill,
Theodoros, Russell, Ward, & Wootton, 2008; Palsbo, 2007; Theodoros, Hill, Russell,
Ward, & Wootton, 2008; Vaughn, 1976; Vestal, Smith-Olinde, Hicks, Hutton, & Hart,
2006; Wertz et al., 1987; Wertz et al., 1992);
(3) voice disorders (Burgess et al., 1999; Howell, Tripoliti, & Pring, 2009; Mashima,
Birkmire-Peters, Holtel, & Syms, 1999; Mashima, et al., 2003; Tindall, Huebner, Stemple,
& Kleinert, 2008);
(4) fluency disorders (Kully, 2000; O'Brian, Packman, & Onslow, 2008); and
(5) dysphagia and the management of head and neck oncology cases (Lalor, Brown, &
Cranfield, 2000; Myers, 2005; Perlman & Witthawaskul, 2002; Ward et al., 2009; Ward et
al., 2007).
36
In the management of adult neurological impairments, the telerehabilitation research has
been influenced by the communication and information technologies available at the time. The
earlier use of the telephone (Helm-Estabrooks & Ramsberger, 1986; Vaughn, 1976) and closed-
circuit television systems (Mashima, et al., 2003; Wertz, et al., 1987; Wertz, et al., 1992), initially
demonstrated the potential of telerehabilitation as an additional or alternate form of service
provision in adult rehabilitation. Today, videoconferencing is the preferred method as it is able to
resemble the face-to-face modality more closely. Videoconferencing systems can simultaneously
transmit audio and video information, which allows for optimal interaction between the SLP and
patient online, as well as obtaining salient information on aspects of speech that are necessary for
diagnostic and treatment purposes (Duffy, et al., 1997; Kully, 2000). Within the area of
videoconferencing, satellite-based systems are gradually being replaced by PC-based ones. The
latter form is proving to be more cost-effective, convenient and likely to assist with the long-term
sustainability of telerehabilitation projects, as PCs are readily available and relatively inexpensive
to install and maintain (Hill, et al., 2006). As the studies in this thesis utilised a PC-based
telerehabilitation system with videoconferencing features, it was considered important to highlight
in the following sections the existing telerehabilitation assessment and treatment studies that also
utilised PC-based videoconferencing in the management of adult neurological impairments.
1.11.1 Speech-Language Pathology Assessment Studies
The few assessment studies in adult telerehabilitation have investigated the ability to assess
dysarthria, apraxia of speech, and aphasia online. Hill and colleagues (2006; 2008; 2009a, 2009b),
utilised a custom made PC-based telerehabilitation system developed at The University of
Queensland, Australia for the above assessments. Nineteen participants with mild to moderate-
severe dysarthria (including six participants with hypokinetic dysarthria and PD), 11 participants
with apraxia of speech (severity levels not stated), and 32 participants with mild to severe aphasia
were assessed respectively in the laboratory setting. A re-designed dysarthria study was also
undertaken later with 24 participants (severity levels not stated). The features of the
telerehabilitation system used in the studies included: (1) videoconferencing over 128 kbit/s Internet
connection; (2) store-and-forward capabilities which allowed for the capture of high resolution and
high quality audio and video footage independent of videoconferencing for later review of task
performance; (3) the ability to display reading material as well as audio and video task
demonstrations for the participant; (4) control of the web cameras at the participant site by the
37
online SLP for optimal viewing of the participant during the assessment (all studies with the
exception of the earlier pilot dysarthria study); and (5) touch screen facility to enable timely
participant interaction as per face-to-face procedure (aphasia study only). An improvement in the
methodology from online and face-to-face assessments conducted separately on two occasions such
as in the pilot dysarthria study (Hill, et al., 2006), to simultaneously (apraxia, aphasia and re-
designed dysarthria studies), ensured that participant test-retest variability was eliminated in the
latter studies (Hill, et al., 2008; Hill et al., 2009a, 2009b; Theodoros, et al., 2008). By utilising the
telerehabilitation system, the authors reported: (1) initially an acceptable level of agreement to
traditional face-to-face ratings for the majority of motor speech parameters in the dysarthria pilot
study (clinical criteria consisting of ± 1 scale point and 80% agreement), and moderate to high
inter- and intra-rater reliability between the online and face-to-face ratings (detailed further in
Chapter 3); (2) further improved outcomes in the re-designed dysarthria study including a good
strength of agreement between all ratings in the online and face-to-face environments, high inter-
and intra-rater reliability for all parameters as well as high participant satisfaction with the online-
led assessments (detailed further in Chapter 3); (3) non-significant differences and moderate to very
good agreement between the online and face-to-face ratings of apraxia of speech, along with
reasonable intra- and inter-rater reliability (exact levels were unable to be calculated due to the
small sample size), and high participant satisfaction with the online-led assessments (detailed
further in Chapter 3); and (4) non-significant differences between the online and face-to-face ratings
on the aphasia assessment, as well as moderate to good agreement between ratings, good to very
good intra- and inter-rater reliability on the majority of the online ratings, and very high participant
satisfaction with the online-led assessments. However, it was acknowledged that certain aspects of
the online environment did make it more difficult at times to rate some assessment tasks. For
instance, even with the improved background lighting and contrast in the re-designed dysarthria
study that helped to resolve several difficulties in viewing participants in the pilot study, the fixed
web camera positioning and focus still made it more difficult at times to view some of the
participants‟ facial features in the later study (Hill, et al., 2006; Hill et al., 2009b). Additionally, the
occasional audio break-up and difficulties clearly viewing the participants made it more challenging
at times to assess aspects of aphasia such as naming and paraphasia, as well as apraxia of speech
(parameters not specified) (Hill, et al., 2008; Hill, et al., 2009a; Theodoros, et al., 2008). It was
suggested that for more severe individuals with apraxia of speech, evaluations conducted face-to-
face may be better suited as they may provide a greater understanding of the participant‟s severity
level and performance (Hill, et al., 2009a). Overall, despite the aforementioned technical
difficulties and generally small sample sizes of the studies, the telerehabilitation modality was
found to be feasible for the assessment of dysarthria, apraxia of speech and aphasia.
38
The assessment of aphasia and motor speech function have also been explored in two
laboratory studies conducted at the National Rehabilitation Hospital in Washington, the USA using
custom-built PC-based telerehabilitation systems (Brennan, et al., 2004; Georgeadis, et al., 2004;
Palsbo, 2007). The first system was used to assess language comprehension and expression using
story retell for 40 participants with acquired brain-injury (traumatic brain injury, TBI; and
cerebrovascular accident, CVA), and speech and language diagnoses which were a combination of
mild to moderate dysarthria, cognitive-communication impairment, aphasia and/or apraxia
(Brennan, et al., 2004; Georgeadis, et al., 2004). The participants were assessed in the online and
face-to-face environments on two separate occasions. The telerehabilitation system operated on a
10 Mbit/s Local Area Network videoconferencing connection with additional features including the
ability to display printed materials remotely at the participant site and to audio record the sessions.
The authors reported that the telerehabilitation system was effective for assessment, as evidenced by
non-significant differences in task performance between the face-to-face and online environments
(Brennan, et al., 2004; Georgeadis, et al., 2004). Further, non-significant differences in
performance between environments were noted for participant factors relating to age, gender,
technology experience and educational level. These findings suggested that participants with
acquired neurological disorders in general could be accurately assessed online on an aphasia battery
(Brennan, et al., 2004). Interestingly, when participant performance was evaluated in relation to the
type of neurological impairment, it was found that the participants with CVA performed equally or
better online than face-to-face. In contrast, the participants with TBI showed worse performance
online. Although these differences in performance (between online vs face-to-face performance for
TBI participants and between the TBI and CVA group) were statistically non-significant
(Georgeadis, et al., 2004), the study showed that attention difficulties can potentially impact on
participant performance online. Further investigations are needed in this area to determine the
extent to which attention difficulties impact on performance. The final study finding of high
participant satisfaction overall with the online assessments has added further promise for the
assessment of language disorders using telerehabilitation (Brennan, et al., 2004; Georgeadis, et al.,
2004).
The second study conducted at the National Rehabilitation Hospital investigated the online
assessment of language comprehension, expression and motor speech performance for 24
participants with acquired brain injury post-CVA (the specific speech and language diagnoses and
severity levels were not reported) (Palsbo, 2007). The telerehabilitation system operated on a
384 kbit/s Internet videoconferencing connection and was described as appropriate for such
39
assessments, with a high level of agreement reported between the online and face-to-face ratings.
However, as no additional information relating to the system or study design was presented, it is
difficult to further interpret the findings or replicate the study in the future.
On the whole, the aforementioned studies have touched on the feasibility of online
assessment of dysarthria, apraxia of speech and aphasia for adults with acquired neurological
disorders. Further large-scale studies are needed in these areas to validate online assessment and
diagnosis using a wide range of disorders, severity levels, assessment parameters, as well as
standardized and non-standardized assessments. As previous studies have demonstrated the
potential of telerehabilitation, it was proposed that providing speech-language pathology assessment
services to people with PD via a PC-based telerehabilitation system might lessen the access issues
that currently exist for this population.
1.11.2 Speech-Language Pathology Treatment Studies
In terms of speech-language pathology treatment for adults with neurological impairments,

principally two proof-of-concept studies have been documented using PC-based telerehabilitation.
An earlier study conducted at the Tripler Army Medical Centre in Hawaii, the USA, focused on
remote voice treatment for military personnel, their families and veterans in the community,
including those with neurological voice disorders. In this study, Burgess et al. (1999) investigated
the online treatment of voice disorders using PC-based videoconferencing (bandwidth and specific
features of the telerehabilitation system were not specified) in the laboratory setting. Ten
participants with voice disorders were randomly assigned and received treatment in either the face-
to-face (six participants) or online treatment environments (four participants). The preliminary
findings showed that all voice samples post-treatment were rated as better than baseline for the
participants in the online environment, while only four of the six participants in the face-to-face
environment showed similar improvements. Participant feedback regarding online treatment was
also positive (Burgess, et al., 1999). Although these findings suggested that online voice treatment
was as effective as face-to-face, certain study limitations make it difficult to generalise the results.
These include the lack of specification of the: (1) online application; (2) types and severity levels of
the voice disorders studied; (3) voice therapy program used; (4) treatment duration; (6) assessment
parameters used; and (6) the small sample size.
40
A later proof-of-concept study by Howell and colleagues (2009) actually investigated the
home-based delivery of the LSVT® via a PC-based telerehabilitation system. Three participants
with IPD and mild to moderate hypokinetic dysarthria were treated online within their homes via a
broadband connection over Skype. Post-treatment improvements were noted in mean SPL for
maximum sustained vowel phonation, reading and monologue tasks (detailed further in Chapter 4).
Although this study touched on the feasibility of online LSVT® delivery, a number of study
limitations make it difficult to determine the full potential of online treatment specifically for
individuals with PD. These limitations include the: (1) small sample size; (2) lack of a comparison
face-to-face treatment group; (3) inability to objectively measure SPL and F0 data in real-time
during the sessions via the online application, as required for treatment; and (4) mixed treatment
modality where four of the 16 sessions were conducted face-to-face in order to monitor SPL,
provide treatment material and build rapport (see Chapter 4 for further details on these specific
limitations). Further studies are therefore needed to establish the feasibility of telerehabilitation for
the delivery of treatment to adults with neurological impairments and more specifically, to those
with PD. To best achieve this, large-scale treatment studies are needed that include participants
with a wide range of neurological disorders and severity levels, valid speech and language treatment
programs, comparison groups, and a wide range of performance measures including SPL, F0,
functional communication outcomes, and participant satisfaction with the online modality. Further,
to ensure best practice, there is a need for online technology to closely replicate face-to-face
assessment and treatment delivery.
1.12 Aims of the Thesis
Telerehabilitation assessment and treatment research in speech-language pathology using

PC-based telerehabilitation has highlighted the potential of this mode of service delivery extending
into mainstream practice for adults with neurological impairments. However, telerehabilitation is
still in its infancy and it is only with further large-scale validation and reliability studies that the full
potential of this approach can be realised and accepted by patients, professionals and policy makers
as an additional or alternate mode of service delivery. Specifically, for individuals with PD, it is
important that future telerehabilitation studies focus on protocols for the online assessment of the
speech and voice difficulties associated with hypokinetic dysarthria, and that treatment studies
investigate the effectiveness of online LSVT® delivery. For this population, where the known
debilitating physical symptoms, distance to health services, cost, travel and transport difficulties
41
currently limit access to speech-language pathology services, telerehabilitation may prove to be a
necessary modality, thereby lessening the existing access issues. The studies in this thesis were
undertaken to add further knowledge about the effectiveness of telerehabilitation for the
management of the speech and voice difficulties associated with PD and to provide a framework for
telerehabilitation use in this area.
Specifically, the studies in this thesis set out to achieve the following aims:
(1) To investigate the validity of the PC-based telerehabilitation system that was custom-built
for the studies in this thesis as an acoustic measurement tool.
(2) To investigate the validity and reliability of online assessment of PD using the PC-based
telerehabilitation system in a laboratory setting, by comparison to face-to-face assessment
for participants with PD and mild to severe hypokinetic dysarthria.
(3) To investigate the validity of online LSVT® delivery using the PC-based telerehabilitation
system in a laboratory setting compared to face-to-face LSVT® for participants with IPD
and mild to moderate hypokinetic dysarthria.
(4) To investigate the feasibility of online LSVT® delivery in a laboratory setting using the
PC-based telerehabilitation system for a small number of complex cases of PD.
(5) To investigate the feasibility of remote home-based LSVT® delivery using the PC-based
telerehabilitation system for a single case of IPD and mild hypokinetic dysarthria.
From these aims, the following hypotheses were established:
(1) The PC-based telerehabilitation system will be an accurate acoustic measurement tool for
SPL and F0 in line with the face-to-face reference measurement tools.
(2) The online assessment of the speech and voice disturbances in PD on the acoustic and
perceptual measures can be achieved to a level comparable to standard face-to-face
assessment.
(3) The LSVT® outcomes obtained online on the acoustic and perceptual measures will not be
inferior to those obtained face-to-face.
42
(4) Online LSVT® can be effectively delivered for individuals with complex PD, resulting in
functional gains.
(5) Home-based LSVT® can be effectively delivered, with treatment outcomes consistent with
those obtained face-to-face.
43
44
Chapter 2
Telerehabilitation System Design and Calibration
2.1 Introduction to the Telerehabilitation System
In order to achieve the study aims set out in this thesis (section 1.12) the PC-based
telerehabilitation system needed to meet certain requirements for the successful management of the
speech and voice disorder associated with hypokinetic dysarthria and PD. For the online SLP, these
demands included the ability to: (1) deliver assessment and treatment in real-time; (2) perceptually
rate speech, voice and oromotor parameters during assessment in real-time and where necessary, to
be able to record and store video and audio assessment data for review off-line at a later date; (3)
provide the participant with appropriate reading materials during assessment and treatment; (4) be
easy to operate and follow a similar approach to face-to-face management; and (5) sample real-
time, calibrated average measures of vocal SPL (dB-C), F0 (Hz) and duration (s) during assessment
and treatment. Additionally, for both the SLP and participant, it was important to establish
appropriate rapport and be able to adequately view and hear each other during the online assessment
and treatment sessions. As outlined previously (section 1.11), a number of custom made PC-based
videoconferencing systems have been successfully utilized to assess dysarthria, aphasia and apraxia
of speech, and to treat neurological voice disorders, which to some extent has also included
hypokinetic dysarthria associated with PD (Brennan, et al., 2004; Burgess, et al., 1999; Georgeadis,
et al., 2004; Hill, et al., 2006; Hill, et al., 2008; Hill, et al., 2009a, 2009b; Howell, et al., 2009;
Palsbo, 2007; Theodoros, et al., 2008). The telerehabilitation systems utilized in these studies
incorporated a number of key features relevant to the present study. However, as no system
available at the time of the study was sufficiently complex to meet all demands of the assessment
and treatment studies in this thesis, a custom made PC-based videoconferencing system was
developed specifically for the studies in this thesis. The telerehabilitation system was developed
within the Telerehabilitation Research Unit at The University of Queensland and with similar
features to those described elsewhere (Hill, et al., 2006; Russell, 2004; Russell, Buttrum, Wootton,
& Jull, 2003). The purpose of this chapter was to firstly outline the rationale and design of the
feature set of the telerehabilitation system used in this thesis. Secondly, this chapter provides
45
experimental evidence for the validity of the telerehabilitation system as an acoustic measurement
tool for SPL and F0, prior to its use in the studies outlined in the thesis.
2.2 Feature Set of the Telerehabilitation System
To meet the clinical demands listed above, the specific features of the telerehabilitation
system needed to include: (1) videoconferencing; (2) the ability to control the participant web
cameras remotely; (3) store-and-forward capabilities; (4) the capacity to display printed materials
on the participant‟s remote computer; (5) ease of operation; and (6) the ability to view and
objectively sample duration and calibrated measures of SPL and F0.
2.2.1 Videoconferencing
In order to administer the comprehensive speech and voice assessment battery, build
appropriate participant-SLP rapport, and deliver the LSVT® within the treatment guidelines, it was
imperative that all sessions be delivered in real-time. This requirement was in keeping with
traditional face-to-face practice. Consequently, an important feature of the telerehabilitation system
needed to be real-time videoconferencing between the participant and SLP site. Videoconferencing
was achieved using a 128 kbit/s Internet connection over the PC. Although it is acknowledged that
there is greater potential for image pixelation and frame rate at this lower bandwidth that can
compromise visual clarity, previous studies assessing dysarthria, aphasia and apraxia of speech at
this bandwidth have demonstrated that administration is comparable to traditional face-to-face
standards and so are the results (Hill, et al., 2006; Hill, et al., 2008; Hill, et al., 2009a, 2009b;
Theodoros, et al., 2008). Furthermore, at the time of the study, the 128 kbit/s Internet connection
was the minimum speed used in Queensland‟s public health systems. It was therefore considered
important to develop a telerehabilitation system with real applicability to the clinical setting for
which it was intended. For videoconferencing, the H.323 standard was used which consisted of
high compression, low bit-rate codecs including the H.263 video codec and the G.723 (5.3 kbit/s)
audio codec. These codes enabled the highest quality service over the low bandwidth connection at
the time of development (Russell, 2004).
46
During the online assessment and treatment sessions, the SLP and participant were able to
view each other via web cameras which facilitated image resolution of 320 x 240 pixels. At the
time of the study, this resolution was common in videoconferencing systems (Russell, 2004). At
the participant site, two web cameras were utilised. One web camera was required for
videoconferencing and the other was used to capture audio and video data independent of
videoconferencing for later review (see section 2.2.3 for further information on the store-and-
forward function). A single screen was utilized at the participant site for assessment and treatment
which allowed viewing of the SLP or where necessary, the display of printed task material (Figure
2.1). At the SLP site, a single web camera was required for videoconferencing purposes. A dual
screen was utilized which permitted continual viewing of the participant on one screen and the
display of printed material or objective measures of SPL and F0 on the other screen (Figure 2.2).
The dual screen option allowed for ease of operation of the telerehabilitation system, maximum
efficiency and a close resemblance to face-to-face practice where the SLP is typically able to
perform a task and also view the participant.
Verbal communication between the participant and SLP over videoconferencing was
achieved via headset microphones (Altec Lansing, Model number AHS202i) that were connected to
each system (Figure 2.1). For the participant site in particular, the use of a headset microphone was
chosen above a web camera or desktop microphone as it helped to ensure a consistent placement of
the microphone from the participant‟s mouth throughout the sessions. This factor helped to control
for any influences relating to possible changes in the participant‟s positioning during the sessions or
excessive movements due to dyskinesia. It also allowed for consistent measures of SPL and F0
throughout the sessions, reduced sound distortion and improved speech clarity.
47
Figure 2.1 Telerehabilitation system at participant site. Note. (1) the videoconferencing system
displaying the SLP; (2) web cameras; and (3) headset microphone.
Figure 2.2 Telerehabilitation system at SLP site. Note. (1) the videoconferencing system
displaying the participant; (2) web camera; (3) display of printed material as seen on the
participant‟s screen.
48
2.2.2 Control of Remote Participant Web Cameras
For both the assessment and treatment sessions, it was important to replicate the SLP‟s
ability to view both the participant‟s face and upper torso as is possible in the face-to-face
environment. This was necessary to ensure that the: (1) participant was able to phonate effectively
at the top of the breath with an open-mouth position during specific assessment and treatment tasks;
(2) SLP was able to detect the presence of muscle tension in the head and neck (contraindicated for
treatment); and (3) participant was positioned appropriately during the assessment and treatment
sessions. The ability to view the participant‟s face and upper torso was therefore considered an
important feature requirement of the telerehabilitation system. To help achieve this, a robotic arm
(Trackerpod, Eagletron) was attached to the participant‟s web cameras that were located on top of
the monitor. The robotic arm allowed for 150 degree left and right pan and 50 degree up and down
pan. Despite the lack of web camera zoom capabilities, the SLP was able to control the remote
participant cameras with the use of a robotic arm and adjust their alignment when necessary, for
optimal viewing of the participant at all times during the online sessions (Figure 2.3).
Figure 2.3 Web cameras on robotic arm at participant site. Note. (1) web camera used for
recording audio and video data; (2) web camera used for videoconferencing; and (3) robotic arm.
49
2.2.3 Store-and-Forward Capabilities
For assessment and treatment, it was necessary to document the sessions for later review
by the assessing and treating SLPs for clinical purposes. Later review of assessment data was also
necessary for reliability testing (intra- and inter-rater reliability), as well as for rating of specific
tasks by independent SLPs who were blind to the intent of the studies. During traditional face-to-
face sessions, the SLP is able to record audio and video information using commercially available
equipment such as minidisk recorders and video cameras respectively. In order to replicate these
capabilities via the online environment, the telerehabilitation system was designed with a store-and-
forward function. This feature enabled the online SLP to capture high quality video (640 x 480
pixel resolution compressed with the windows media video codec version eight at 384 kbit/s) and
audio recordings (windows media audio codec version eight at 368kbit/s) at the participant end via
the first web camera on top of monitor (Figure 2.3). Video was recorded at 25 frames/s. These files
were stored on the participant computer and then forwarded back through a data channel to the SLP
computer. Using separate data channels for the videoconferencing and store-and-forward file
transfer ensured that the video and audio quality of the real-time conference was not affected by the
presence of a concurrent store-and-forward transfer (Figure 2.4).
Figure 2.4 Schematic diagram of videoconferencing and the separate data channel used to transfer
store-and-forward audio and video files between the SLP and participant sites.
An additional benefit of the store-and-forward feature was the improved audio and video
quality compared to real-time videoconferencing. Therefore, on the occasions where the 128 kbit/s
bandwidth made it more difficult to judge some specific parameters such as: (1) subtle speech
production features; (2) fine facial movements and precision; and (3) the presence of the participant
factors mentioned previously (section 2.2.2), the store-and-forward feature could be used to
50
facilitate these ratings. The store-and-forward feature has been incorporated successfully into
previous online systems to record assessment sessions for later review, as well as to assist in rating
certain motor speech parameters requiring a high level of visual clarity (Brennan, et al., 2004;
Georgeadis, et al., 2004; Hill, et al., 2006; Hill, et al., 2008; Hill, et al., 2009a, 2009b; Palsbo, 2007;
Theodoros, et al., 2008).
2.2.4 Remote Display of Materials
Two important aspects that facilitate the assessment and treatment sessions in the
traditional face-to-face setting are the ability to provide the participant with reading material during
the sessions and to demonstrate tasks as required. These elements are readily achieved during
traditional sessions via paper-based reading materials and SLP demonstrations respectively. In
order to replicate these capabilities during the online sessions, the remote display of text and pre-
recorded audio and video demonstrations at the participant site were incorporated in the design of
the telerehabilitation system. This was also in keeping with a number of online assessment studies
that had successfully incorporated the use of display functions in their telerehabilitation systems in
order to pre-empt any audio and video difficulties with videoconferencing, and to ensure that the
telerehabilitation systems closely resembled the features of the face-to-face sessions (Brennan, et
al., 2004; Georgeadis, et al., 2004; Hill, et al., 2006; Hill, et al., 2008; Hill, et al., 2009a, 2009b;
Theodoros, et al., 2008). In these studies, scanned test materials and video recordings of task
demonstrations were incorporated into the telerehabilitation systems and displayed by the online
SLPs on the participants‟ screens as necessary.
By utilising the remote display feature of the telerehabilitation system designed for this
thesis, the online SLP was similarly able to present printed reading material on the participant‟s PC
during the assessment and treatment sessions. This feature allowed the online SLP to independently
and efficiently deliver all aspects of the assessments and treatments that required reading material.
This function eliminated the need to provide material in hard copy to the participant in advance of
the sessions, and also, the potential level of confusion in then having the participant try to locate the
correct materials. The SLP was also able to simultaneously display the reading materials on her
screen, allowing close monitoring of tasks and the ability to provide prompt feedback, in keeping
with traditional face-to-face management (Figure 2.2). In order to ensure an exact replication of the
51
reading material, the text was either scanned or created in Microsoft Office Word 2003 using the
same font face and size, prior to its display online. Moreover, to improve the clarity of the text for
participants with visual difficulties, the SLP was able to increase the font size during the session
and select a desired section of text to display on their screen. In addition to reading material, the
telerehabilitation system allowed for the display of pre-recorded speech and oromotor task
demonstrations during the assessment sessions. These demonstrations facilitated the participant‟s
ability to follow tasks on the occasions where audio-visual difficulties with videoconferencing
compromised the instructions delivered in real-time by the SLP.
2.2.5 Ease of Operation
To ensure efficient assessment and treatment in line with face-to-face delivery, an

important requirement of the telerehabilitation system was its ease of operation (Yellowlees, 1997).
In order to meet this requirement, a certain consideration was made in the design of the
telerehabilitation system. This was the ability for the SLP to use the telerehabilitation system
intuitively, without technical knowledge or extensive training. Further, to simplify operations from
the perspective of the participant, the telerehabilitation system enabled the online SLP to remotely
control all aspects of online assessment and treatment. This ensured that the participant was not
required to operate any aspects of the telerehabilitation system, making the system very user
friendly for them. Participants with no prior computer skills or knowledge could therefore be
recruited to the study. Overall, the ease of operation for the SLP and the user friendliness of the
telerehabilitation system for both the SLP and participant also met the recommendations for
maximising participant and SLP satisfaction with telehealth, as outlined in Chapter 1 (section 1.10).
2.2.6 Measures of Sound Pressure Level, Fundamental Frequency and Duration
A key aspect of both assessment and treatment delivery was the ability to objectively
measure SPL, F0 and duration in real-time. During a traditional speech and voice assessment for
PD, it is necessary to measure these acoustic parameters in order to determine an overall dysarthria
severity level and the suitability of the LSVT® for the patient. Measurements of mean SPL and
duration of phonation are generally obtained on assessment tasks from speech samples such as
maximum sustained vowel phonations, reading passages and monologues, while vocal glides are
52
used to obtain measures of maximum F0 range (see section 3.2.4.1 for specific details relating to the
assessment protocol for PD used in this thesis). Likewise, during traditional face-to-face delivery of
the LSVT®, the SLP must be able to objectively measure and monitor the participant‟s vocalisations
in real-time on a number of SPL tasks including maximum sustained vowel phonations, speech
loudness drills of reading and conversation, and to measure maximum F0 ranges and length of
phonations (section 4.2.5 further outlines the LSVT® tasks). Timely feedback on task performances
must also be provided during the treatment sessions to assist with achieving the required loudness
levels and calibration, or the participant‟s ability to self-monitor and maintain their louder voice in
everyday communication (Ramig, Pawlas, et al., 1995).
During conventional assessment and treatment (LSVT®) sessions for PD, quantification of
SPL, F0 and duration are made possible using commercially available instruments such as a Digital
Sound Level Meter (SLM), chromatic tuner and stopwatch respectively (Ramig, Countryman, et al.,
1995; Ramig, Pawlas, et al., 1995). To enable quantification of these acoustic parameters in the
online environment, an acoustic speech processor was specifically developed for the studies in this
thesis as a plug in device to the telerehabilitation system at the participant site (Figure 2.5). An
acoustic measurement software tool was then utilized to quantify the input from the
telerehabilitation system‟s acoustic speech processor. This in turn, enabled the online SLP to view
and sample real-time calibrated average recordings of SPL (dB-C) and F0 (Hz) data and to obtain
measures of duration (s) (Figure 2.6).
53
Figure 2.5 Telerehabilitation system‟s acoustic speech processor (lid removed) at the participant
site (1) and headset microphone (2).
Figure 2.6 Screen shot of the acoustic measurement software tool at the SLP site displaying SPL
and F0. Note. (1) An array of numbers measuring SPL in real-time; (2) peak SPL; (3) averaging
function of SPL; Similarly, (4) measures of F0; (5) peak F0 level; (6) averaging function of F0; and
(7) a historical representation of the sample where 2 data points can be selected and a mean SPL or
F0 can be obtained for the range bounded by the two points.
54
2.3 Design of the Telerehabilitation System as an Acoustic Measurement Tool
The purpose-built external acoustic speech processor was developed for this
telerehabilitation system to ensure that the variability in gain and frequency characteristics of the
on-board sound card of the PC did not contribute to an error in measured SPL or F0. This aspect of
the telerehabilitation system also ensured that the acoustic speech processor could function across
any PC, making it widely applicable to the clinical setting. The final design of the telerehabilitation
system‟s acoustic speech processor consisted of a headset microphone, amplifier, filter and
microcontroller that delivered serial data to the dedicated acoustic measurement software tool of the
telerehabilitation system. As indicated in Figure 2.7, the signal from the microphone was buffered
and amplified uniformly across the audio range initially. A “C” weighted filter was then applied
prior to digitisation within the microcontroller. The same signal was amplified to a nonlinear range
and applied to a fixed threshold above ambient noise level to produce a digital transistor-transistor
logic (TTL) signal at the frequency of the maximum amplitude F0 applied to the microphone. This
digital signal was applied to a frequency counting routine within the microcontroller based on a
time gate of 125 ms. A frequency reading was thereby generated at a maximum rate of 8/s. The
digital communication to the telerehabilitation system‟s acoustic measurement software tool was
based on a conventional RS-232 protocol and the data (SPL and F0) were transmitted in the
American Standard Code for Information Interchange (ASCII) text with embedded flags to
delineate F0 from SPL readings every 125 ms. An interrupt driven driver within the acoustic
measurement software tool then scaled (according to the calibration procedure described in the
following sections) and displayed these two data sources for feedback to the SLP remotely.
Duration was measured by the system clock.
55
Figure 2.7 Schematic diagram of the telerehabilitation system‟s acoustic speech processor
hardware, whereby Path A processed SPL (dB) data and Path B processed F0 (Hz) data.
Note. TTL = transistor-transistor logic.
2.4 Calibration and Verification of the Telerehabilitation System as an Acoustic

Measurement Tool
In order to achieve real-time calibrated average measures of SPL and F0 and to determine
the validity of the telerehabilitation system as an acoustic measurement tool, a number of
calibration and verification phases were performed. These involved a series of four experiments
that comprised the following stages of system validation:
(1) Calibration of SPL using pure tones
(2) Verification of SPL using pure tones
(3) Verification of SPL using live voice
(4) Verification of F0 using pure tones
The overall hypothesis for this series of verification investigations was that the measures of
SPL and F0 obtained using the telerehabilitation system would be as accurate as those obtained
using the face-to-face reference measurement tools.
56
2.4.1 Experiment 1: Calibration of Sound Pressure Level Using Pure Tones
The calibration of SPL was initially performed using pure tones. This was a necessary
phase in order to overcome some of the challenges associated with live voice including: (1) the
different F0s present in voice; (2) the fact that the amplitude of the sound varies with different sound
frequencies (Wagner, 1994); as well as (3) challenges relating to the production of standing waves
in the sampling room. Therefore, the single F0 associated with pure tones (Wagner, 1994) made it
easier to isolate the target frequency and to then perform the calibration process at different SPLs.
Prior to the calibration process, the mean SPL values delivered by the telerehabilitation system‟s
acoustic speech processor were arbitrarily. This is because the SPL output was influenced by a
number of factors including the: (1) distance of the microphone from the speaker‟s mouth;
(2) number of amplification stages and the amount of gain applied to the signal by the pre-
amplifiers in the telerehabilitation system‟s acoustic processor; and (3) resistance of the processor‟s
circuitry. Therefore, it was hypothesised that the average measures obtained using the non-
calibrated telerehabilitation system for pure tone SPL would vary as a function of the reference
instrument other than a one-to-one linear relationship. Consequently, the telerehabilitation system
would require calibration.
2.4.1.1 Experiment 1: Method
The arrangement of the apparatus in this calibration phase is illustrated in Figure 2.8. For
this phase, a commercially available Function Generator (Topward Electronic Instruments, Model
TFG-462) was used to produce a pure tone at varying SPLs. The output from the Function
Generator was connected to the input channel of a commercially available speaker (Behringer,
MS16 Monitor Speaker), which produced the audio signal. The telerehabilitation system‟s acoustic
speech processor captured this signal via a commercially available headset microphone (Altec
Lansing, Model number AHS202i) that was connected to the processor. The headphones were
placed so that the distance from the microphone to the dust cap centre of the speaker was 5 cm.
This distance represented the optimal microphone distance from the corner of a person‟s mouth that
allowed for reduced sound distortion as well as maximum visibility of the person‟s face. This set
distance was maintained throughout all studies in this thesis. For the calibration phase, the Visi-
Pitch II (Kay Elemetrics, Model 3300) was chosen as the reference measurement tool. The Visi-
Pitch instruments are widely used in traditional speech-language pathology practice as clinical and
57
research tools. To ensure that the same acoustic signal was obtained by both the telerehabilitation
system and Visi-Pitch II in the present study, the output from the speaker was also directly
connected to the input channel of the Visi-Pitch II (see Figure 2.8).
Figure 2.8 Equipment and set-up of calibration phase for pure tone SPL.
Average measures of SPL were sampled within the range of 55 dB to 85 dB which

incorporated conversational speech. The lower limit represented room ambient noise level and the
upper limit was within the maximum loudness range that is recommended with the LSVT® training
to ensure that the behaviour is not vocally abusive (Ramig, Pawlas, et al., 1995). For this phase, a
consistent 500 Hz pure tone level was used as it represented a F0 level that is attainable by both
males and females of various ages when performing maximum F0 range tasks (Colton, Casper, &
Leonard, 2006). To reach the desired SPL, the speaker volume was manually adjusted using the
volume dial and cross checked with the output of the Visi-Pitch II. The sampling levels were
increased by approximately 2 dB increments (M = 1.85 dB, SD = 1.21), which was the minimum
level that could be reliably achieved via manual adjustment. When the closest level to the desired
SPL was registered using the Visi-Pitch II, the acoustic signal was sampled simultaneously using
the telerehabilitation system and Visi-Pitch II for an average of 10 s. Two researchers from the
Telerehabilitation Research Unit took part in the study and were responsible for sampling the output
58
using either system. Following the calibration procedure, the raw values obtained using the
telerehabilitation system and the calibrated measures of the Visi-Pitch II were averaged at each SPL
level sampled. A scatter plot of the paired data points was used to determine the curve trend
between the two instruments.
2.4.1.2 Experiment 1: Results and Discussion
A scatter plot of the 17 paired data points revealed a non-linear distribution between the
telerehabilitation system and Visi-Pitch II (Figure 2.9). This result supported the study hypothesis
of differences in measures between the two instruments and confirmed that a calibration procedure
needed to be performed. Consequently, a curve fitting operation using the power model equation
y b0 ( x b1 ) was computed, producing the regression coefficients b0 = 46.78 and b1 = 0.11. This
calibration model was subsequently incorporated into the telerehabilitation system‟s acoustic
measurement software tool to transform the raw SPL values sampled via the telerehabilitation
system into calibrated measures.
Figure 2.9 Comparative fit plot using the linear and power curve estimation regression models for
the non-calibrated telerehabilitation system and Visi-Pitch II measures of pure tone SPL.
59
2.4.2 Experiment 2: Verification of SPL Using Pure Tones
Experiment 1 was repeated with the incorporation of the calibration equation into the
acoustic measurement software tool of the telerehabilitation system to verify the accuracy of the
calibration model in correcting raw SPL values using pure tones. It was hypothesised that with the
inclusion of the calibration model, the SPL measures obtained using the calibrated telerehabilitation
system would be as accurate as those using the Visi-Pitch II reference tool for pure tone.
For this experiment, the testing procedure described in section 2.4.1.1 was repeated. A
sampling SPL range of 60 dB to 93 dB was used to again incorporate the conversational speech
range and sampling at approximately 2 dB increments (M = 1.84 dB, SD = 1.00) was performed.
The curve trend of the two instruments was determined using a scatter plot. Furthermore, statistical
analyses were performed to determine the accuracy of the calibrated telerehabilitation system‟s
average measures of pure tone SPL compared to the Visi-Pitch II. For this, paired samples t-tests
(level of significance set at p < .05) and the Bland and Altman (1986) “limits of agreement” (LA)
were used. The Bland and Altman statistic establishes the LA within which 95% of differences
between the two environments are predicted to lie. If the LA are found to be within a
predetermined clinical criterion, the new method can be considered an acceptable measurement tool
and the two methods can be used interchangeably (Bland & Altman, 1986). For investigations of
SPL in this study, the clinical criterion was set at ± 2.25 dB difference between the two
measurement tools which was half of the 4.5 dB mean level of improvement in SPL reported during
a monologue task in the LSVT® efficacy study for IPD participants with mild to moderate
hypokinetic dysarthria following treatment (Ramig, Countryman, et al., 1995). Monologue
loudness was chosen in this study as it is commonly used as a primary outcome measure for the
LSVT® as it provides a good indication of the level of carryover of the louder voice to daily life
(Ramig, Pawlas, et al., 1995). Additionally, the criterion level for this study was set below the
minimum improvement level in monologue loudness in order to determine whether the
telerehabilitation system would be able to discriminate a relevant treatment change when used
clinically.
60
The scatter plot of the 19 paired data points revealed an almost one-to-one linear trend
between the telerehabilitation system and Visi-Pitch II, producing regression coefficients
b0 = -0.789 and b1= 1.007 (Figure 2.10). Minimal differences were noted in the mean SPL values
(mean absolute difference, MAD = 0.36, SD = 0.19) between the two instruments (M = 77.95 dB,
SD = 9.70 for the telerehabilitation system; M = 77.71 dB, SD = 9.77 for the Visi-Pitch II).
Statistical analyses using the paired samples t-test also revealed non-significant differences
(t = 0.08, p = .939) between the telerehabilitation system and Visi-Pitch II in measuring pure tone
SPL. Furthermore, the Bland and Altman LA (-0.91 to 0.44 dB) were within the predetermined
clinical criterion of ± 2.25 dB (Figure 2.11).
calibrated telerehabilitation system and Visi-Pitch II measures of pure tone SPL.
61
Figure 2.11 Bland and Altman LA for measures of pure tone SPL. Note. Clinical criterion (CC) =
± 2.25 dB; LA (-0.91 to 0.44 dB) within the clinical criterion.
The results obtained on the verification trial using pure tone SPL have demonstrated that
the calibrated telerehabilitation system was able to accurately measure pure tone SPL in line with
the Visi-Pitch II reference measurement tool. This was evident by an almost one-to-one linear trend
and non-significant differences (p > .05) between the two instruments. Furthermore, the Bland and
Altman LA were within the predetermined clinical criterion of ± 2.25 dB. These findings supported
the study hypothesis that for pure tone, measures of SPL were as accurate using the calibrated
telerehabilitation system as they were using the reference instrument.
2.4.3 Experiment 3: Verification of Sound Pressure Level Using Live Voice
The accuracy of the calibrated telerehabilitation system was further verified using live
voice. This additional step was important in determining the validity of using the telerehabilitation
system as a clinical measurement tool for online assessment and treatment delivery in this thesis. It
was hypothesised that the SPL measures obtained for live voice using the calibrated
telerehabilitation system would be as accurate as the reference measurement tool.
62
For this phase, a Digital SLM (Radio Shack®, Model No. 23-553) was chosen as the
reference measurement tool. This instrument is used clinically in the traditional face-to-face
delivery of the LSVT® to measure mean SPL (Ramig, Pawlas, et al., 1995), and was therefore also
used for this verification phase instead of the Visi-Pitch II. The set-up of the verification phase is
illustrated in Figure 2.12. The participants in this study were two researchers within the
Telerehabilitation Research Unit (female aged 23 years; male aged 30 years). Each participant was
assessed individually and was seated comfortably at a table, facing the Digital SLM which was
positioned at a distance of 30 cm from their mouth. This set-up was in keeping with the face-to-
face LSVT® protocol (Ramig, Pawlas, et al., 1995). At this distance, the Digital SLM produced
values consistent with the VisiPitch II, allowing it to be used as the reference tool in this phase.
Each participant wore the headset microphone that had been used in the previous calibration and
verification phases (sections 2.4.1 and 2.4.2). The microphone distance was measured at 5 cm from
the corner of their mouth, in keeping with the previous calibration set-up. Also consistent with the
previous phases was the headset microphone that was connected to the telerehabilitation system‟s
acoustic speech processor.
Figure 2.12 Equipment and set-up of verification phase of SPL using live voice.
63
For this verification phase, each participant was instructed to sustain a vowel phonation of
/a/ for approximately 5 s at the desired loudness level. The SPL was monitored using the Digital
SLM by the researchers involved in the previous calibration and verification phases. The
participants were instructed verbally by one of the researchers who was monitoring the Digital SLM
to increase or decrease their loudness until the SPL registered as close as possible to the desired
level on the meter. Once the level was achieved, the output was sampled simultaneously by the two
researchers using the telerehabilitation system, and manually using the Digital SLM. For the latter,
the peak SPLs were recorded by hand as they were displayed on the meter. For the female
participant, the SPLs were sampled within the range of 66 dB to 91dB at approximately 2 dB
increments (M = 2.27 dB, SD = 0.90). For the male participant, the SPLs were sampled within the
range of 62 dB to 91dB at approximately 2 dB increments (M = 2.80 dB, SD = 1.03). At each SPL
sampled, the measures obtained using the two instruments were averaged and compared using
scatter plots, paired samples t-tests and the Bland and Altman LA to determine any differences
between the two instruments. Consistent with section 2.4.2.1, the clinical criterion using the Bland
and Altman LA was set at ± 2.25 dB.
Figures 2.13 and 2.14 display the scatter plots of the 12 paired data points for the female
speaker and 11 paired data points for the male speaker respectively. Both scatter plots revealed an
almost one-to-one linear trend between the calibrated telerehabilitation system and Digital SLM
(regression coefficients b0 = 1.104 and b1= 0.987 for the female participant; b0 = 3.268 and
b1= 1.042 for the male participant). There were also minimal differences in the mean SPL values
between the calibrated telerehabilitation system and the Digital SLM for both the female participant
(MAD = 0.08, SD = 0.67; M = 79.42 dB, SD = 8.16 for the telerehabilitation system; M = 79.50 dB,
SD = 8.08 for the SLM). A similar trend was also identified for the male participant (MAD = 0.09,
SD = 0.70; M = 76.18 dB, SD = 9.44 for the telerehabilitation system; M = 76.10 dB, SD = 9.85 for
the Digital SLM). For both participants, the paired samples t-test further revealed non-significant
differences between the two instruments (t = -0.03, p = .980 for the female; t = -0.43, p = .676 for
male). Furthermore, the Bland and Altman LA for both the female (-1.46 to 1.28 dB) and male
participants (-1.23 to 1.40 dB) were within the predetermined clinical criterion of ± 2.25 dB (see
Figures 2.15 and 2.16 respectively).
64
calibrated telerehabilitation system and Digital SLM for measures of SPL using a female voice.
calibrated telerehabilitation system and Digital SLM for measures of SPL using a male voice.
65
Figure 2.15 Bland and Altman LA for measures of SPL using a female voice. Note. Clinical
criterion (CC) = ± 2.25 dB; LA (-1.46 to 1.28 dB) within the clinical criterion.
Figure 2.16 Bland and Altman LA for measures of SPL using a male voice. Note. Clinical criterion
(CC) = ± 2.25 dB; LA (-1.23 to 1.40 dB) within the clinical criterion.
66
Overall, the results obtained on the verification phases for SPL using live voice supported
the study hypothesis that the calibrated telerehabilitation system could accurately measure SPL in
line with the reference measurement tool, the Digital SLM. The results revealed an almost one-to-
one linear trend between the calibrated telerehabilitation system and Digital SLM, non-significant
differences (p > .05) between the instruments, and Bland and Altman LA within the predetermined
clinical criterion of ± 2.25 dB. Together, these findings provided strong support for the use of the
telerehabilitation system as the acoustic measurement tool for SPL in the assessment and treatment
studies described in this thesis.
2.4.4 Experiment 4: Verification of Fundamental Frequency Using Pure Tones
A verification trial of F0 using pure tones was also performed in order to determine the
accuracy of the telerehabilitation system for this measure. It was hypothesised that F0 using pure
tones could be accurately measured using the telerehabilitation system, in keeping with the
reference tool. This is because frequency is not influenced by factors such as microphone distance
or pre-amplification as SPL, and should therefore not require calibration.
The equipment and set-up of this phase were consistent with the calibration phase
described for pure tone SPL (section 2.4.1.1). Measures of F0 were sampled within the range of
100 Hz to 975 Hz to incorporate the maximum F0 ranges reported in the literature for males and
females of various ages (Colton, et al., 2006). This was an important consideration of the
telerehabilitation system as the LSVT® in particular requires the measurement of maximum F0
range for each patient. In this phase, sampling was performed at the set SPL of 80 dB to
incorporate the upper end of conversational speech loudness. Within the sampling range, the F0
produced by the Function Generator was varied in approximately 25 Hz increments (M = 24.20,
SD = 5.53), which was the minimal level that could be reliably achieved via manual adjustment of
the dial. At each desired interval, the output was sampled simultaneously for an average of 10 s
using the Visi-Pitch II and telerehabilitation system. The two researchers involved in the earlier
SPL phases sampled the output on either instrument. Once completed, the values obtained from the
Visi-Pitch II and telerehabilitation system were averaged across the 10 s sample at each frequency
67
level. In keeping with the previous experiments (sections 2.4.2.1 and 2.4.3.1) scatter plots, paired
samples t-tests and the Bland and Altman LA were used to determine any differences between the
two instruments.
For measures of F0, the clinical criterion was more difficult to determine in relation to
previous LSVT® efficacy studies reporting improvements in maximum F0 range, which is a
measure of treatment success. This is because the minimum 4 semitones (ST) benchmark for
improvement on this task (Ramig, et al., 1994), was measured using a different scale to the present
study (Hz), making conversion difficult. As a result, the clinical criterion was based on the minimal
change in F0 with the LSVT® during a monologue task as this has previously been expressed in
hertz in the literature. The clinical criterion was set at ± 3 Hz difference between the two
measurement tools which was at least half of the 6.31 Hz and 10.88 Hz mean improvement levels
with the LSVT® reported in an efficacy study for female and male IPD participants respectively
(Ramig, Countryman, et al., 1995). The use of this more stringent criterion was also considered to
provide a clearer indication of the sensitivity of the telerehabilitation system as an acoustic
measurement tool for F0.
The scatter plot of the 35 paired data points revealed a one-to-one linear trend between the
telerehabilitation system and Visi-Pitch II with regression coefficients b0 = 0.462 and b1= 1.000 (see
Figure 2.17). There were also minimal differences in the mean values (MAD = 1.46, SD = 3.66)
between the telerehabilitation system and Visi-Pitch II (M = 535.94 dB, SD = 263.32 for the
telerehabilitation system; M = 536.85 dB, SD = 263.86 for the Visi-Pitch II). Furthermore, a paired
samples t-test showed a non-significant difference (t = -.02, p = .988, between the telerehabilitation
system and the Visi-Pitch II measures. The Bland and Altman LA (-1.82 to 0.88 Hz) were also
within the predetermined clinical criterion of ± 3 Hz (Figure 2.18).
68
telerehabilitation system and Visi-Pitch II measures of pure tone F0.
Figure 2.18 Bland and Altman LA for measures of pure tone F0. Note. Clinical criterion (CC) =
± 3 Hz; LA (-1.82 to 0.88 Hz) within the clinical criterion.
69
The present findings have demonstrated that the telerehabilitation system provided
accurate measures of F0 using pure tones, which were consistent with the Visi-Pitch II reference
measurement tool. The validity of the telerehabilitation system was demonstrated by the one-to-one
linear relationship and non-significant differences (p > .05) between the two instruments, as well as
the Bland and Altman LA which were within the predetermined clinical criterion of ± 3 Hz. The
findings verified the study hypothesis that F0 using pure tones could be accurately measured using
the telerehabilitation system, in line with the reference measurement tool. As a strong one-to-one
linear relationship was obtained between the telerehabilitation system and the Visi-Pitch II, no
further verification phases were necessary for F0.
2.5 Conclusion
The telerehabilitation system described in this thesis was specifically designed to

incorporate the features of videoconferencing, control of remote web cameras, store-and-forward
function, remote display of materials, ease of operation and the objective measurement of SPL, F0
and duration. The validity of the telerehabilitation system for measuring SPL and F0 was
demonstrated in the various calibration and verification phases. The study findings supported the
use of the telerehabilitation system as an acoustic measurement tool in the online assessment and
treatment studies in this thesis, in order to replicate measurements obtained by a SLP in a typical
face-to-face environment.
70
Chapter 3
Assessing Disordered Speech and Voice in Parkinson’s Disease
Online: A Telerehabilitation System
3.1 Introduction
As outlined in Chapter 1 (section 1.11.1), few speech-language pathology studies have

investigated online assessment of motor speech function using PC-based telerehabilitation.
Although none of these studies specifically examined online assessment of hypokinetic dysarthria
associated with PD, their findings have demonstrated the feasibility of this type of service delivery
for adults with neurological disorders in general. Hill and Colleagues (2006; 2009a, 2009b) utilised
a similar telerehabilitation system to that described in this thesis to assess dysarthria and apraxia of
speech. The features of the telerehabilitation system included: (1) videoconferencing between the
participant and SLP sites operating on a 128 kbit/s Internet connection; (2) store-and-forward
function; (3) the display of printed material and demonstrational information on the participant‟s
screen to resemble face-to-face assessment instruction; and (4) control of the participant‟s web
cameras at the SLP site for optimal viewing of the participant during the assessment (only the later
studies). In their earlier dysarthria pilot study, 19 participants with mild to moderate-severe
dysarthria resulting from TBI, CVA, hypoxic brain injury, neurosurgery for tumour removal, and
PD (6 participants) were assessed (Hill, et al., 2006). For each participant, an online and face-to-
face assessment was conducted on two separate occasions. The outcome measures included: (1) an
overall dysarthria severity rating obtained from a conversational sample (using a seven-point scale);
(2) perceptual analysis of speech production using a standard reading passage; (3) a 19-item version
of the Frenchay Dysarthria Assessment (FDA) (Enderby, 1983); and (4) sections of the Assessment
of Intelligibility of Dysarthric Speech (ASSIDS) (Yorkston & Beukelman, 1981a). Results showed
that the online ratings were within predetermined clinically acceptable levels of agreement to face-
to-face on the majority of parameters (dysarthria severity, percent intelligibility in sentences and
perceptual ratings, using clinical criteria of ± 1 scale point and 80% agreement). Moderate to high
intra- and inter-judge reliability were also reported on the majority of parameters. On the FDA,
however, several online ratings were not comparable to face-to-face, particularly the palatal
movement in speech, laryngeal volume, tongue elevation and lateral tongue movements parameters.
71
The authors suggested that general assessment limitations such as poor parameter and rating
definitions and possible participant test-retest variability may have contributed to the lower
agreement levels for these parameters. Additionally, the specific technical issues encountered in the
online environment such as difficulties clearly viewing the participants‟ facial features due to the
fixed positioning and focus of the web camera and lack of background lighting, may have
influenced the ratings (Hill, et al., 2006). Although no specific references were made to the
performance of the PD participants overall, it was noted in this study that the fixed camera
positioning made it more difficult to view those individuals with PD with head or body dyskinesis
when they moved position.
A follow-up, redesigned dysarthria assessment study by the same authors helped to address
some of the issues identified above in terms of the original design of the pilot study, certain
limitations of the telerehabilitation system and the assessment protocol used (Hill, et al., 2009b).
Twenty-four participants with dysarthria (type and severity levels not specified) resulting primarily
from CVA and TBI took part in the study. An improvement to the previous study design was the
simultaneous assessment of participants in the face-to-face and online environment. This particular
design helped to reduce potential bias relating to participant test-retest as identified previously.
Greater visualisation of participants online was also made possible through the improved features of
the telerehabilitation system that allowed the online SLPs to control the web cameras at the
participant site. Better background lighting and contrast also helped to enhance visualisation of the
participants online. Finally, the improved assessment protocol consisted of: (1) an informal
oromotor and perceptual speech assessment with clear parameter and rating definitions (five-point
rating scale); (2) sections of the ASSIDS; (3) determining an overall dysarthria diagnosis for each
participant; and (4) a participant satisfaction questionnaire relating to the online-led assessments.
These overall considerations were very appropriate and helped to achieve improved outcomes to the
pilot study which included a good strength of agreement between the online and face-to-face ratings
and high inter- and intra-rater reliability for all parameters (Hill, et al., 2009b). Together with the
high participant satisfaction with the online-led assessments, the study findings further suggested
that valid assessment of dysarthria could be achieved via telerehabilitation.
In their apraxia study, 11 participants with acquired apraxia of speech following CVA
(severity levels not specified) and with concomitant mild to moderate aphasia and/or dysarthria took
part (Hill, et al., 2009a). The participants were assessed simultaneously in the face-to-face and
72
online environments using the Apraxia Battery for Adults-2 (Dabul, 2000). Similar to the previous
studies, the authors demonstrated the feasibility of online assessment with non-significant
differences reported between the online and face-to-face environments, and reasonable intra- and
inter-rater reliability (exact levels could not be determined due to small sample size). Participant
satisfaction with the online-led assessments was also reported as high. However, certain aspects of
the online environment did make it more difficult to assess participants with severe apraxia. It was
noted that with these individuals, the occasional break-up in the audio and difficulties viewing
subtle facial features with videoconferencing made it more challenging for the SLPs to assess them
(Hill, et al., 2009a). The authors suggested that for some patients with severe apraxia of speech,
face-to-face assessments may be better suited as SLPs may be able to gain further information from
nonverbal cues and pragmatics in this environment to assist with assessment rating, which is often
more difficult online due to reduced audio and video quality.
Despite the small sample sizes, the general study findings have been promising and have
suggested that online assessment of speech and voice is feasible and largely valid for patients with
dysarthria and apraxia of speech. The results may further suggest that valid and reliable
assessments of the speech and voice difficulties associated with PD may be achieved via
telerehabilitation. This modality may help to substantially lessen the access issues to speech-
language pathology services that currently exist for this population (section 1.6). Regardless of the
technology used, valid and reliable assessment procedures need to be established for people with
PD to ensure effective telerehabilitation services where appropriate assessments underpin treatment
programs. Therefore, the current study aimed to investigate the validity and reliability of a PC-
based assessment protocol in a laboratory-based setting comprising acoustic and perceptual
measures that was specifically designed to evaluate the speech and voice disturbances associated
with PD, by comparison with clinical face-to-face assessment. It was hypothesised that the online
assessment of the speech and voice disturbances in PD on the acoustic and perceptual measures can
be achieved to a level comparable to standard face-to-face assessment.
73
3.2 Method
3.2.1 Sample Size Calculation
Prior to commencement of the study, ethical clearance was obtained from the Behavioural
and Social Sciences Ethical Review Committee of The University of Queensland. The sample size
required for this study was determined prior to participant recruitment. As the study hypothesis
predicted that the outcomes between the online and face-to-face environments would be
comparable, the sample size calculation was based on that for an equivalence trial (Matthews,
2000). For this calculation, the primary outcome measure for assessment of PD which was also
relevant to treatment (LSVT®), was defined as the mean SPL during a 30 s monologue. This
measure was chosen as it provided the best indication of the participant‟s loudness level in
functional communication. As part of the assessment protocol specifically designed for PD, it was
essential that changes in mean SPL on the monologue task could be adequately represented pre- and
post-treatment. Therefore, the minimum clinically important difference of the primary outcome
measure used in the sample size calculation was 4.5 dB. This value represented the minimum
clinically relevant improvement in monologue loudness with treatment, as identified in the LSVT®
efficacy studies conducted in the face-to-face environment (Ramig, et al., 1996; Ramig,
Countryman, et al., 1995). However, as there were no published reports at the time of the present
study that compared online and face-to-face LSVT® performance, the standard deviation of the pre-
to post-LSVT® difference scores between the two treatment environments that was necessary for
the sample size calculation, could not be adopted from the literature. Consequently, the standard
deviation of 2.48 dB was adopted from preliminary treatment data collected in this thesis (Chapter
4), following analysis of the first eight participants to receive the LSVT® in the online and face-to-
face environments. By using the aforementioned reference levels in the sample size calculation,
together with an alpha level of 0.05, a statistical power of 80% and allowing for a further 10%
attrition rate, a total of 16 participants were required for the study.
3.2.2 Participants
Participant recruitment was able to exceed the minimum numbers required for adequate
power and a total of 61 participants with PD and hypokinetic dysarthria (42 males; 19 females) aged
74
between 52 and 89 years (M = 69.23 years; SD = 8.60) volunteered for the study. Difficulties
recruiting females resulted in an uneven participant sample. Participants were diagnosed as having
PD by a neurologist experienced in movement disorders. Time post-diagnosis ranged from 6
months to 30 years (M = 6.52 years; SD = 6.53). Fifty-seven of the participants had been diagnosed
with IPD, of which nine participants had undergone surgical treatment for PD including DBS
(seven participants) and pallidotomy (2 participants). The remaining four participants in the cohort
had been diagnosed with PPS, including PSP (3 participants) and MSA (1 participant). Stages of
PD as per the Hoehn and Yahr Scale (1967) for the participants ranged from I to IV, with 48
participants rated at Stages I and II and 13 participants rated at Stages III and IV. For all
participants, an overall severity level for hypokinetic dysarthria was determined by the principal
investigator. This classification was based on the assessment data obtained in the present study
which consisted of mean SPL and perceptual ratings of overall speech intelligibility in conversation
(OIC) on a five-point scale (1 = normal, completely intelligible speech; 5 = severely unintelligible
speech with difficulties deciphering many words) that were rated from a monologue. Please see
section 3.2.4 for further details on the assessments. Specific criteria were used to classify the
severity levels. Mild hypokinetic dysarthria was considered to be consistent with SPLs above 65
dB in monologue loudness and a mild reduction in overall speech intelligibility. Moderate
hypokinetic dysarthria corresponded to SPLs ranging from 60 dB to 65dB with an accompanying
mild to moderate reduction in speech intelligibility. Severe hypokinetic dysarthria corresponded to
SPLs below 60 dB and a moderate or severe reduction in speech intelligibility. Using this
classification, the dysarthria levels for the participants in this study ranged from mild to severe, with
40.98% of participants considered as mild, 47.54% of participants as moderate, 3.28% as moderate-
severe and 8.19% as severe. Descriptive characteristics of the participants are summarized in Table
3.1.
75
Table 3.1 Descriptive Characteristics of Participants
Participant Age Sex Time post Hoehn & Dysarthria PD type ±

(years) diagnosis Yahr severity neurosurgery
(years) Stage level
1 80 M 2 1 Mild IPD
2 56 M 6 1 Mild IPD
3 69 M 2 1.5 Mild IPD
4 59 M 4 1 Mild IPD
5 77 M 2 1 Mild IPD
6 66 M 1.5 1 Mild IPD
7 71 F 1 1 Mild IPD
8 54 M 4 1 Mild IPD
9 58 M 0.7 1.5 Mild IPD
10 68 M 3 1 Mild IPD
11 59 M 11 1 Mild IPD
12 67 M 2.5 1 Mild IPD
14 66 M 5.5 1 Mild IPD
15 74 F 1 1.5 Mild IPD
17 76 F 5 1 Mild IPD
19 77 M 1 1.5 Mild IPD
20 67 F 7 1.5 Mild IPD
76
(years) Stage level
22 74 M 6.5 1.5 Mild IPD
24 59 F 0.5 1 Mild IPD
26 81 M 7 2.5 Moderate IPD
28 74 M - 2.5 Moderate IPD
29 75 F 14 3 Moderate IPD
30 69 M 22 2 Moderate IPD
36 62 F 4 2.5 Moderate IPD
77
(years) Stage level
48 76 F 14 4 Moderate IPD, DBS

Unilateral
49 58 M 2 3 Moderate IPD DBS

Bilateral
50 63 M 10 2.5 Moderate IPD DBS

Bilateral

Bilateral
52 66 M 12 3.5 Moderate IPD DBS

Bilateral

Bilateral
54 53 F 1 2 Moderate PSP
55 75 M 16 3.5 Moderate- IPD

Severe Pallidotomy
Unilateral
56 85 F 3 2 Moderate- IPD
78
(years) Stage level
severe
57 65 M 13 3 Severe IPD DBS

Bilateral
58 52 F 13 3 Severe IPD
Pallidotomy
Unilateral
59 60 F 3 3.5 Severe PSP
60 89 M 19 4 Severe PSP
61 62 F 1 3.5 Severe MSA
Group 69.23 6.52 1.90

Mean (SD) (8.60) (6.53) (0.95)
Note. – unspecified; PD = Parkinson‟s disease; DBS = deep brain stimulation; IPD = idiopathic
Parkinson‟s disease; PSP = Progressive supranuclear palsy; MSA = Multiple system atrophy; M =
male; F = female; SD = standard deviation.
Participants were recruited from various support groups of Parkinson‟s Queensland

Incorporated, public hospitals and from private neurologists in Brisbane, Australia. Proficiency in
the use of computers was not a requirement for inclusion in the study as all aspects of the online
assessment delivery were performed by the online assessing SLP. Exclusion criteria included a:
(1) speech and/or language disturbance or a co-existing neurological disorder inconsistent with PD;
(2) severe uncorrected auditory and/or visual disturbance; (3) cognitive disturbance inconsistent
with the capacity to provide informed consent; (4) respiratory dysfunction unrelated to the
neurological disorder; and (5) positive history of alcohol abuse. The participants were not formally
assessed in order to rule out any of the above difficulties. However, participants were excluded
from the study if they and/or their families reported significant difficulties in any of these areas
during the initial case history, and/or where these difficulties where evident to the principal
investigator.
79
The primary mode of assessment (online or face-to-face led) was randomly selected for
each participant using a computerised random-number generator (Dallal, 1997). Randomisation
codes were generated and assigned by the principal investigator to the participants in the order that
they entered the study. In total, 31 assessments were led face-to-face and 30 led online. Figure 3.1
is a schematic flow chart of the participants through the study.
Figure 3.1 Schematic flow chart of participants through the study.
3.2.3 Assessors
Three SLPs experienced in the assessment of motor speech disorders and PD took part in
the study. Assessments were conducted at The University of Queensland within the
Telerehabilitation Research Unit. Prior to the commencement of the study, SLP training was
conducted by the principal investigator in a three-hour session which covered the administration of
80
all assessments in both the online and face-to-face assessment environments. The SLPs were
deemed competent with online administration when they could adequately deliver a mock session
within a one-hour time frame and also agree on the level of severity of five dysarthric speakers who
were not involved in the study. The speakers were judged on the perceptual measures of voice,
overall articulatory precision (OAP), OIC, and non-speech oromotor function. During the study,
two of the three SLPs took part in each assessment session, where one SLP led the session, while
the second SLP acted as a silent rater and did not interact with the participant. One SLP assessed
the participant in the face-to-face environment (within the same room as the participant), while the
second SLP conducted the assessment in the online environment, through a videoconferencing link
via the Internet. The SLPs were also randomized to the assessment environments and were blind to
the participants and their level of hypokinetic dysarthria prior to assessment. In total, SLP 1 took
part in 25 of the online assessments (16 as leader and nine as silent rater) and 20 of the face-to-face
assessments (nine as leader and 11 as silent rater); SLP 2 took part in 21 online assessments (seven
as leader and 14 as silent rater) and 24 face-to-face assessments (11 as leader and 13 as silent rater);
and SLP 3 took part in 14 online assessments (seven as leader and seven as silent rater) and 18 face-
to-face assessments (12 as leader and six as silent rater).
3.2.4 Assessment Battery
Each participant underwent a one-hour assessment on one occasion on a battery of acoustic

and perceptual measures specifically designed for this study. The battery consisted of an
instrumental evaluation of mean SPL, maximum duration of sustained vowel phonation and
maximum F0 range, and perceptual ratings of voice and non-speech oromotor parameters, OAP and
speech intelligibility in reading and conversation. These measures were chosen for the study as
they are commonly used to diagnose and define the level of severity of hypokinetic dysarthria
associated with PD. Furthermore, as this study formed part of a larger validation trial that also
evaluated online treatment (Chapter 4), the measures were chosen as they have been used in the
LSVT® literature as sensitive predictors of treatment change (Ramig, et al., 1996; Ramig,
Countryman, et al., 1995).
81
3.2.4.1 Acoustic Measures
The LSVT® Evaluation Protocol (Ramig, Pawlas, et al., 1995) was used to assess the
participant‟s mean SPLs, maximum duration of sustained vowel phonation and maximum F0 range
during several speaking tasks. This protocol has been widely used in the LSVT® literature as a
routine assessment.
3.2.4.1.1 Sound pressure level and maximum duration of sustained vowel phonation
The mean SPLs of the participant‟s speech were recorded during six maximum sustained
vowel phonations of /a/, readings of the Rainbow Passage (Fairbanks, 1960) and The Grandfather
Passage (Darley, et al., 1975), and during a 30 s monologue about a topic of interest such as family,
hobbies or a recent holiday trip. The duration of each maximum sustained vowel phonation was
also measured in seconds. For all tasks, the participants were instructed to speak in a comfortable
voice and no reference was made to their loudness level. Following the assessment, the SPL and
duration levels were then averaged to provide mean levels for each participant.
3.2.4.1.2 Maximum fundamental frequency range
Each participant performed a series of six vocal glides, reaching their highest and lowest F0
levels respectively. No reference was made to their loudness level. The average highest and lowest
F0 levels (Hz) obtained for each participant were then converted to a maximum range in semitones
(de Pijper, 2007). The levels were converted to semitones so that treatment comparisons in the
subsequent studies in this thesis could be made to face-to-face literature reporting treatment changes
in semitones for this task (see section 2.4.4.1 for previous difficulties in converting results in the
literature that used semitones to hertz).
82
3.2.4.2 Perceptual Measures
In the absence of a standardized assessment protocol for the perceptual measures,

evaluations of these parameters were made on a five-point rating scale using a battery of measures
developed for the study.
3.2.4.2.1 Perceptual voice parameters
The reading of The Grandfather Passage in the acoustic section was also used to rate
perceptual measures of vocal breathiness, roughness (lack of clarity), strain-strangled quality,
tremor, pitch and phonation breaks. Modal pitch and loudness levels and pitch and loudness
variability were rated from the 30 s monologue as per the acoustic section. The vocal parameters
were evaluated using a five-point rating scale (1 = normal; 5 = severely impaired).
3.2.4.2.2 Oromotor function
An informal assessment of non-speech oromotor function was developed to evaluate

specific parameters using a five-point rating scale (1 = normal; 5 = severely impaired). The
parameters included masked facial expression, lip movement (retraction, pucker, seal, alternate
movement), tongue movement (symmetry, protrusion, elevation/depression, lateral and alternate
movement), breath support and diadochokinetic (DDK) rates (alternate motion rate, AMR /pʌpʌ/
and sequential motion rate, SMR /pʌtʌkʌ/).
3.2.4.2.3 Overall articulatory precision
A perceptual rating of each participant‟s articulatory precision was also made from the
speech sample obtained during the reading of The Grandfather Passage. OAP was rated on a five-
point scale (1 = normal, precise production of sounds; 5 = severe distortion or imprecision that
interferes with speech intelligibility).
83
3.2.4.2.4 Measures of speech intelligibility
The ASSIDS was used to measure speech intelligibility at the single word and sentence
level, as well as communication efficiency. For this assessment, participants read or repeated a
series of 50 words and 22 sentences of increasing length. The words and sentences had been
randomly generated prior to the assessment, in accordance with test procedure. The reading
material was displayed on the participant‟s screen or presented as per the test booklet, depending on
the assessment environment. Copyright approval was obtained from the publishers (Pro-Ed, Austin,
TX) to enable conversion of test materials to an online format. Audio recordings of participant
speech samples were made in both environments. Following assessments of all participants, the
speech samples obtained face-to-face and online were numerically coded, randomized and saved to
CD for analysis and scoring. Two independent SLPs who did not participate in the study and who
were blind to the study intent transcribed the speech samples obtained in each environment.
Following the ASSIDS ratings, the values given by the two SLPs from the online and face-to-face
recordings were averaged to express a single mean value for each sample obtained in that
environment. Scores for the word and sentence intelligibility tasks were expressed as percent
correct. The communication efficiency ratio was determined by dividing the participant‟s rate of
intelligible speech (intelligible words per minute) by the mean rate of intelligible words per minute
for normal speakers (190 words per minute) (Yorkston & Beukelman, 1981a). An additional rating
of the participant‟s OIC was made from the 30 s monologue sample using a five-point scale
(1 = normal, completely intelligible speech; 5 = severely unintelligible speech with difficulties
deciphering many words). For this task, the rating was made by the two SLPs who had taken part
in the online and face-to-face assessment for the particular participant.
3.2.4.3 Participant Satisfaction Questionnaire
The 30 participants in the online-led assessments completed a brief questionnaire. On a

five-point scale, the questionnaire evaluated the level of participant satisfaction with: (1) the online
assessment sessions (possible responses ranging from would not participate again to would prefer
these types of sessions to face-to-face sessions); (2) the audio and video quality during the sessions
(responses ranging from poor to excellent); and (3) overall satisfaction with the online assessment
(ranging from not at all satisfied to very satisfied).
84
3.2.4.4 Speech-Language Pathologist Comments
In addition to the participant satisfaction questionnaire, the SLPs were invited to provide
feedback on any aspects relating to the online assessment delivery. This information was obtained
anecdotally.
3.2.5 Assessment Environment
3.2.5.1 Online Environment
The PC-based telerehabilitation system described in Chapter 2 was used for the online
assessment. The application operated on a 128 kbit/s Internet connection and videoconferencing at
320 x 240 pixel resolution was conducted between the SLP and participant site. Additional features
of the system which were used in this study included the ability to: (1) display printed material and
instructional video clips on the participant‟s screen; (2) control the remote camera with the use of a
robotic arm and adjust its alignment for optimal viewing of the participant‟s head and upper torso;
and (3) capture high-quality video (640 x 480 pixel resolution compressed with the windows media
video codec Version 8 at 384 kbit/s) and audio recordings (windows media audio codec Version 8
at 368kbit/s) independent of videoconferencing for the perceptual measures, and then store-and-
forward these audio and video files back to the online SLP for later review (see section 2.2 for
further information of the feature set of the telerehabilitation system).
To standardize acoustic measures across the two assessment environments, the

telerehabilitation system was used as the objective measurement tool in both the online and face-to-
face environments for all acoustic measures. Therefore, where tasks required acoustic measures,
both the online and face-to-face SLPs were able to view and sample real-time calibrated average
recordings of SPL, F0 and duration data. This was achieved via input from the telerehabilitation
system‟s acoustic speech processor that was located at the participant site, and conversion to
calibrated data via the acoustic measurement software tool, both of which were specifically
developed for the studies in this thesis (section 2.2.6).
85
The online-led assessments were conducted between two rooms on separate floors of the
university building. Prior to the session, the face-to-face SLP would physically turn on the PC and
activate the application at the participant site. For standardization purposes, the participant was
seated in front of the telerehabilitation system at a distance of approximately 50 cm from the PC
monitor and wore a headset microphone to enable interaction with the online SLP during
videoconferencing. The microphone distance was set at 5 cm from the corner of the participant‟s
mouth in order to reduce sound distortion, maximize visibility of the participant‟s face, and allow
for accurate recordings of SPL and F0. The distance was measured at the start of each session by
the face-to-face SLP. The output from the telerehabilitation system was also verified against a
Digital SLM held at approximately 30 cm from the participant‟s mouth on a sample of three
sustained /a/ phonations. During the assessment, the online SLP wore a headset microphone
attached to the telerehabilitation system for communication with the participant. The SLP
controlled all displays on the participant‟s screen, without the need for the participant to operate the
system.
During the online-led assessment, the online SLP administered the various tasks and
interacted with the participant over the 128 kbit/s Internet videoconferencing link. At this
bandwidth, live ratings of SPL and F0 were possible, however, judgments of fine movements and
precision on the oromotor assessment were more difficult due to a low picture frame rate and
resolution picture quality. In addition, the real-time detection of subtle features of speech
production for perceptual ratings of voice, OAP and speech intelligibility was also more difficult on
occasion due to the degradation of audio quality. Therefore, to improve the video and audio quality
for rating, the online SLP used the store-and-forward feature of the telerehabilitation system to
record the task and store the video and audio files for later viewing and analyses. The store-and-
forward feature was used routinely by all online SLPs (leading and silent assessors). To standardize
the perceptual measures, SLPs in both environments rated the assessments live (where possible) and
then reviewed the sessions off-line using the equipment available in that environment. A summary
of the online assessment procedure is displayed in Table 3.2.
In addition, the effects of the audio-visual difficulties with videoconferencing on the

participants‟ ability to follow task instructions were minimized with the use of pre-recorded task
demonstrations of the oromotor assessment. Where necessary, these demonstrations were displayed
on the participant‟s screen by the online SLP. Throughout the online-led assessment, the face-to-
86
face SLP acted as the silent rater at the participant site. The face-to-face SLP wore headphones and
was able to follow the assessment instructions given to the participant. The online assessment
environment is represented in Figures 3.2 and 3.3.
87
Table 3.2 Assessment Procedure for Online and Face-to-Face Environment
Assessment measure Online instrument Online scoring procedure FTF instrument FTF scoring procedure
Acoustic Measures
SPL Online acoustic speech Real-time Online acoustic speech Real-time

processor processor
Duration of phonation Online acoustic speech Real-time Online acoustic speech Real-time
processor processor
Maximum F0 range Online acoustic speech Real-time Online acoustic speech Real-time
processor processor
Perceptual Measures
Perceptual voice Online store and forward Real-time where possible Minidisk recorder Real-time where possible
audio and reviewed off-line and reviewed off-line

parameters
Oromotor parameters Online store and forward Real-time where possible Video camera Real-time where possible
video and reviewed off-line and reviewed off-line
OAP Online store and forward Real-time where possible Minidisk recorder Real-time where possible
88
Assessment measure Online instrument Online scoring procedure FTF instrument FTF scoring procedure
OIC Online store and forward Real-time where possible Minidisk recorder Real-time where possible
ASSIDS Online store and forward Off-line Minidisk recorder Offline
audio
Participant satisfaction Paper based End of online-led N/A N/A
questionnaire assessment session
Note. SPL = sound pressure level; F0 = fundamental frequency; OAP = overall articulatory precision; OIC = overall speech intelligibility in
conversation; ASSIDS = Assessment of Intelligibility of Dysarthric Speech; N/A = not applicable; FTF = face-to-face assessment environment.
89
Figure 3.2 Online-led assessment by online SLP and equipment at site. Note. (1) the
videoconferencing system displaying the participant; (2) acoustic measurement software tool
displaying SPL and F0 data; and (3) web camera.
Figure 3.3 Online-led assessment at participant site with face-to-face SLP as the silent rater (left).
Note. (1) the videoconferencing system displaying the online SLP; (2) web cameras; (3) video
camera; and (4) acoustic measurement software tool displaying SPL and F0 data for the face-to-face
SLP.
90
3.2.5.2 Face-to-Face Environment
During the face-to-face led assessment, the participant was seated in front of the
telerehabilitation system with the monitor turned off. Standard face-to-face test administration
procedures were used. The online SLP became the silent rater and viewed, listened, and recorded
tasks while the face-to-face SLP interacted with the participant. In keeping with the online
procedure, the face-to-face SLP obtained real-time measures of mean SPL, F0 and duration via the
telerehabilitation system. A dual screen at the participant site allowed the face-to-face SLP to
sample the acoustic measures during the face-to-face and online-led assessments on one of the
screens. The second screen was utilised during the online-led assessments for videoconferencing
between the participant and online SLP site (Figure 3.3). The face-to-face SLP also used a video
camera and minidisk recorder to collect video and audio data respectively for later analyses, as per
standard clinical practice. The video camera was positioned as close as possible behind the web
cameras and a microphone on a stand was connected to a minidisk recorder and placed 30 cm from
the participant. The face-to-face SLP wore headphones and was able to hear the online SLP if there
was a need for a task repetition or further online recording. A summary of the face-to-face
assessment procedure is also displayed in Table 3.2.
3.2.6 Statistical Analyses
Online and face-to-face ratings and measurements for all participants were compared on
each assessment task to determine the level of agreement between the two environments. Analysis
of the ASSIDS and acoustic parameters (SPL tasks, maximum duration of sustained vowel
phonation and maximum F0 range) were performed using the Bland and Altman LA method for
continuous data (see section 2.4.2.1 for explanation of this method). For those parameters
consisting of ordinal data (perceptual ratings of voice and oromotor parameters, OAP and OIC),
percent close agreement (PCA) and the quadratic weighted Kappa (ĸw) statistic (Landis & Koch,
1977) were calculated.
91
3.2.6.1 Bland and Altman Limits of Agreement
In the absence of a reported minimal, clinically important difference in word and sentence
intelligibility on the ASSIDS assessment, the clinical criterion was established on the test-retest
variability reported in the manual for dysarthric speakers assessed in the face-to-face environment
(Yorkston & Beukelman, 1981a). Values of ± 3.2% and ± 8.6% were set for the respective
measures. In addition, the clinical criterion for the communication efficiency ratio was set at ± 0.27
which was consistent with the criterion determined by Hill et al. (2006) for dysarthric speakers.
The authors had determined the clinical criterion by applying a 95% confidence interval to the
standard deviation of 0.14 which reflected the mean difference between ratings on the ASSIDS
assessments for a group of participants with hypokinetic dysarthria reported previously (Farrell, et
al., 2005). As the criterion was based on a dysarthric population consistent with the current study, it
was considered appropriate for use here. Furthermore, as the assessment battery used in the study
was designed ultimately to determine treatment outcome, the clinical criteria for the SPL, maximum
F0 range, and maximum duration of sustained vowel phonation tasks were set at levels below the
minimal improvement expected following the LSVT®. For all SPL tasks (maximum sustained
vowel phonation, reading and monologue loudness), the clinical criterion was set at ± 4 dB
difference between the two raters, a level below the 4.5 mean level of improvement reported in the
LSVT® efficacy studies for monologue tasks (Ramig, et al., 1996; Ramig, Countryman, et al.,
1995). The clinical criterion for the maximum duration of sustained vowel phonation task was set
at ± 3 s, as the minimal change in phonation time expected with the LSVT® has been reported to be
a mean of 3.72 s (Ramig, Countryman, et al., 1995). For measures of maximum F0 range, the
clinical criterion was set at ± 3 ST, which was below the 4 ST minimum improvement on this task
post-LSVT® (Ramig, et al., 1994). This clinical criterion was also in keeping with the level of
subject variability in healthy adults that can range from 2 to 4 ST (Gelfer, 1986).
3.2.6.2 Percent Close Agreement
PCA was chosen as it is commonly used to quantify agreement in perceptual ratings of

dysarthria. PCA was also selected for the present study to further verify ĸw as it has been reported
in some instances that non-linear distribution of data can negatively impact on ĸw creating a
paradox (Cicchetti & Feinstein, 1990). PCA expressed the percentage of ratings where differences
were within ± 1 scale point on the perceptual rating scales (Kearns & Simmons, 1988; Kreiman,
92
Gerratt, Kempster, Erman, & Berke, 1993; Rey, Plapp, Stewart, Richards, & Bashir, 1987; Sheard,
Adams, & Davis, 1991). In keeping with previous studies that examined dysarthric speech using
perceptual rating scales, the clinical criterion for an acceptable level of agreement in the present
study was considered to be equal to or greater than 80% agreement within ± 1 scale point (Kearns &
Simmons, 1988; Kreiman, et al., 1993).
3.2.6.3 Quadratic Weighted Kappa Statistic
The ĸw is widely used in telerehabilitation studies for ordinal data and provides an
indication of agreement between raters (Landis & Koch, 1977). In the present study, the statistic
provided a measure of agreement beyond chance between the online and face-to-face measures.
The ĸw assigned weights to the observed and chance agreement and presented levels of agreement
where ĸw less than 0.20 is interpreted as poor; 0.21 to 0.40 is fair; 0.41 to 0.60 is moderate; 0.61 to
0.80 is good and 0.81 to 1.00 indicates very good agreement (Landis & Koch, 1977). For this
study, the clinical criterion for an acceptable level of agreement was set at ĸw > 0.6 (good
agreement).
3.2.6.4 Reliability
Reliability between the online and face-to-face environments was conducted for all the
perceptually based assessments and the acoustic F0 measure (Hz). Although the F0 data was
objectively obtained via the telerehabilitation system, the SLPs were required to select a sample
frequency level from a section of the vocal glide thus introducing a subjective element to this task.
The SLPs used the audio, video and F0 data captured during the assessment session in the respective
environments to rate and score the various parameters. Intra- and inter-rater reliability between the
online and face-to-face assessors was calculated using two-way, random effect intra-class
correlations (ICC(2,1)) for 20% (n = 13) of participants in each environment. For inter-rater
reliability, the third SLP who did not take part in a particular assessment session became the
additional rater and was randomly assigned to the online or face-to-face ratings. Intra- and inter-
rater reliability was calculated collectively using the ratings from each of the three SLPs in the
particular environment. ICC values below 0.40 corresponded to poor-to-fair reliability; between
93
0.40 and 0.75 to moderate-to-good reliability; and values above 0.75 represented very good
reliability (Fleiss, 1981).
3.3 Results
3.3.1 Acoustic Measures
The Bland and Altman LA are displayed in Figure 3.4 for sustained vowel phonation
(LA = -1.97 to 1.35 dB), the reading of the Rainbow Passage (LA = -1.05 to 1.04 dB) and The
Grandfather Passage (LA = -1.18 to 1.11 dB), and monologue loudness (LA = -1.07 to 0.81 dB).
For all SPL tasks, the LA were within the predetermined clinical criterion of ± 4 dB. Similarly, the
LA for the maximum duration of sustained vowel phonation task (LA = -2.74 to 2.70 s) were within
the clinical criterion of ± 3 s for online and face-to-face ratings (Figure 3.5). The LA for the
maximum F0 range (LA = -2.03 to 2.19 ST) were also within the clinical criterion of ± 3 ST (Figure
3.6).
94
Figure 3.4 Bland and Altman LA for maximum sustained vowel phonation, readings of The
Rainbow and Grandfather Passages and monologue task. Note. Clinical criterion (CC) for all SPL
tasks = ± 4 dB. LA within clinical criterion.
Figure 3.5 Bland and Atlman LA for maximum duration of sustained vowel phonation.
Note. Clinical criterion (CC) = ± 3 sec. LA (-2.74 to 2.70 s) within the clinical criterion.
95
Figure 3.6 Bland and Altman LA for maximum F0 range. Note. Clinical criterion (CC) = ± 4 ST.
LA (-2.03 to 2.19 ST) within the clinical criterion.
3.3.2 Perceptual Measures
3.3.2.1 Perceptual Voice Parameters
All individual voice parameters met the pre-determined clinical criterion of 80%
agreement for PCA (Table 3.3). However, when using ĸw, seven of the 10 parameters (breathiness,
roughness and strained-strangled vocal quality, pitch breaks, phonation breaks, modal pitch and
loudness variability) were below the clinical criterion of good agreement (ĸw > 0.6).
96
Table 3.3 Perceptual Voice Parameters between Face-to-Face and Online Environments
Voice parameters PCA ĸw
Breathiness 91.80 0.36 (fair)*
Roughness 95.08 0.33 (fair)*
Strained-strangled 95.08 0.41 (moderate)*
Vocal tremor 100 0.69 (good)
Pitch breaks 96.66 0.11 (poor)*
Phonation breaks 95.00 0.37 (fair)*
Modal pitch 96.66 0.38 (fair)*
Pitch variability 96.72 0.63 (good)
Loudness level 100 0.69 (good)
Loudness variability 98.36 0.49 (moderate)*
Note. PCA = percent close agreement; ĸw = quadratic weighted Kappa statistic; * = achieved lower
than the clinical criterion of ĸw > 0.6.
3.3.2.2 Oromotor Function Parameters
Analyses revealed that all individual oromotor parameters reached the clinical criterion for
PCA (Table 3.4). The ĸw indicated that only two parameters (masked facial expression and lip
retraction) fell outside of the clinical criterion of good agreement.
97
Table 3.4 Perceptual Oromotor Parameters between Face-to-Face and Online Environments
Oromotor parameters PCA ĸw
Masked facial expression 86.89 0.31 (fair)*
Lip movement
Retraction 98.36 0.56 (moderate)*
Pucker 98.36 0.77 (good)
Seal 95.08 0.66 (good)
Alternate 100 0.95 (very good)
Tongue movement
Symmetry 100 0.66 (good)
Protrusion 100 0.94 (very good)
Elevation/depression 98.36 0.93 (very good)
Lateral 100 0.89 (very good)
Alternate 100 0.85 (very good)
Breath support 96.72 0.83 (very good)
DDK
AMR /pʌpʌ/ 100 0.75 (good)
SMR /pʌtʌkʌ/ 100 0.87 (very good)
Note. PCA = percent close agreement; ĸw = quadratic weighted Kappa statistic; * = achieved lower
than clinical criterion ĸw > 0.6; DDK = diadochokinetic; AMR = alternate motion rate; SMR =
sequential motion rate.
98
3.3.2.3 Overall Articulatory Precision
For ratings of OAP between the online and face-to-face environments, PCA (100) and ĸw
(0.67 good agreement) were within the clinical criteria.
3.3.2.4 Measures of Intelligibility
For the ASSIDS assessment, the Bland and Altman LA are displayed in Figures 3.7 to 3.9
for word (LA = -10.27 to 8.77%) and sentence intelligibility reading tasks (LA = -5.59 to 6.16%),
and the communication efficiency ratio (LA = -0.12 to 0.10). The LA for sentence intelligibility
and communication efficiency ratio were within the respective clinical criteria (± 8.6% and ± 0.27),
while the word intelligibility LA fell outside of the clinical criterion of ± 3.2%. In addition,
perceptual ratings of OIC were within the clinical criteria for PCA (98.36) and ĸw (0.79 good
agreement).
Figure 3.7 Bland and Altman LA for word intelligibility. Note. Clinical criterion (CC) = ± 3.2%.
LA (-10.27 to 8.77%) outside of the clinical criterion.
99
Figure 3.8 Bland and Altman LA for sentence intelligibility. Note. Clinical criterion (CC) =
± 8.6%. LA (-5.59 to 6.16 %) within the clinical criterion.
Figure 3.9 Bland and Altman LA for communication efficiency ratio. Note. Clinical criterion
(CC) = ± 0.27. LA (-0.12 to 0.10) within the clinical criterion.
100
3.3.3 Reliability
Intra-class correlations ranged from moderate to very good intra-rater reliability in both
assessment environments (ICC = 0.43 to 0.99 face-to-face; ICC = 0.48 to 0.99 online), indicating
comparable intra-rater reliability between environments. Inter-rater reliability was also found to be
comparable between environments, with reliability values between moderate to very good for the
majority of the face-to-face (ICC = 0.43 to 0.99) and online assessments (ICC = 0.48 to 0.99)
(Table 3.5).
101
Table 3.5 Intra-class Correlation Values for Intra-Rater and Inter-Rater Reliability for Online and
Face-to-Face Ratings
Assessment task FTF intra-rater Online intra-rater FTF inter-rater Online inter-rater
reliability reliability reliability reliability
Oromotor parameters 0.85 0.81 0.74 0.76
Perceptual voice 0.60 0.68 0.44 0.56

parameters
OAP 0.43 0.48 0.43 0.48
OIC 0.63 0.69 0.75 0.82
Maximum F0 range 0.99 0.99 0.99 0.99
ASSIDS
WI 0.93 0.99 0.89 0.94
SI 0.97 0.77 0.94 0.82
CER 0.99 0.97 0.97 0.98
Note. FTF = Face-to-face environment; OAP = overall articulatory precision; OIC = overall speech
intelligibility in conversation; F0 = fundamental frequency; WI = percentage word intelligibility;
SI = percentage sentence intelligibility; CER = communication efficiency ratio; ICC values below
0.40 = poor-to-fair, 0.40 to 0.75 = moderate-to-good, above 0.75 = very good reliability; Reliability
was calculated for three assessors.
3.3.4 Online Assessment Delivery
For all 30 online-led assessments conducted in this study, the sessions were delivered
successfully using the PC-based telerehabilitation system operating on a 128 kbit/s Internet
connection. There were no failed sessions or technical issues and all sessions were administered
with adequate audio and video quality over videoconferencing.
102
3.3.5 Participant Satisfaction Questionnaire
Figures 3.10 to 3.13 display the participant ratings on the satisfaction questionnaire. The
majority of participants in the online-led assessments: (1) were comfortable to very happy while
participating in the online sessions (93.32%); (2) found the audio quality during videoconferencing
as adequate or excellent (70%); and (3) found the video quality as adequate or more than adequate
(63.33%). Overall, participants were more than satisfied or very satisfied (80%) with the online
modality.
Figure 3.10 Participant satisfaction with the online assessment sessions.
Figure 3.11 Participant satisfaction with the online audio quality.
103
Figure 3.12 Participant satisfaction with the online video quality.
Figure 3.13 Participant satisfaction overall with online modality.
3.4 Discussion
The results of the present study indicated that a PC-based assessment of the disordered
speech and voice associated with PD was generally valid and reliable. For the majority of the
acoustic and perceptual parameters, the face-to-face and online ratings were within the clinical
criteria as reported in previous face-to-face studies.
104
Objective measures of mean SPL, maximum duration of sustained vowel phonation and
maximum F0 range were obtained in real-time via the telerehabilitation system in both
environments. Although the verification trials demonstrated the validity of the telerehabilitation
system as an acoustic measurement tool (Chapter 2), it was important to assess the performance of
the system during each of the different assessment modes to determine if there was an effect of
transmission across the Internet. Acoustic measures were analysed using the Bland and Altman LA.
For the SPL tasks (maximum sustained vowel phonation, reading and monologue loudness), all
measures were within the clinical criterion of ± 4 dB. This finding is not surprising as the
telerehabilitation system provided objective measures of SPL which were consistent between
environments. The minor differences in values most likely reflected the slightly unsynchronized
start of SPL sampling by the two SLPs in each environment, while transmission across the Internet
appeared to have little effect on the task (Figure 3.4). Measures of maximum duration of sustained
vowel phonation were also within the clinical criterion of acceptable differences (± 3 s), and the
minor differences in duration may also have reflected the subjective element in initiating the
sampling (Figure 3.5). For the maximum F0 range task, the LA were also within the predetermined
clinical criterion of ± 3 ST (Figure 3.6). This further demonstrated that SLPs in both environments
were able to obtain comparable frequency values within a clinically acceptable level, and the
subjective element of selecting a sample from a section of the vocal glide for analysis did not
impact greatly on the results. Moreover, reliability measures revealed very good intra- and inter-
rater reliability for this task between the two environments. Collectively, the comparable values
obtained for all acoustic measures suggested that the telerehabilitation system was a sensitive
acoustic measurement tool that could be used to detect minimal changes in vocal SPL, maximum F0
range and duration in a PD assessment battery.
Analysis of individual voice parameters showed PCA between the online and face-to-face
ratings to be within the clinical criterion. This suggested that the PC-based telerehabilitation system
105
was a valid tool for administering a voice evaluation for PD and the audio store-and-forward
capabilities of the system were sensitive enough for the SLP to determine the presence and level of
severity of specific voice parameters online. However, at the more stringent level of analysis (ĸw),
the online and face-to-face assessments of several of the individual voice parameters failed to reach
acceptable levels of agreement. The ĸw revealed greater variability than PCA for seven of the ten
voice parameters (breathiness, roughness and strained-strangled vocal quality, pitch breaks,
phonation breaks, modal pitch and loudness variability) that fell below the clinical criterion
(achieving poor to moderate agreement). These lower values may reflect rater variability
commonly seen in perceptual rating of voice and/or the nature of the ĸw.
Findings of lower agreement for voice parameters in an online assessment were similarly
reported by Hill and colleagues in their dysarthria studies where consensus agreement was below
80% for ratings of breathiness (68.42%), roughness (63.16%), strained-strangled vocal quality
(73.68%) (Hill, et al., 2006), and overall vocal quality (68.75 %) (Hill, et al., 2009b). Hill and
colleagues (2006) attributed aspects of the lower agreement to possible inter-rater variability that is
inherent in perceptual ratings of voice. It has been acknowledged that naturally occurring
variability is commonly associated with traditional face-to-face evaluations and reflects the
subjective nature of perceptual rating scales (Chan & Yiu, 2002; Kreiman, et al., 1993). The
listener often applies variable internal standards of pathological vocal qualities from their own
experiences to the evaluation (Kreiman & Gerratt, 1998; Kreiman, et al., 1993). Consequently,
achieving high agreement between different raters in relation to vocal qualities is often problematic.
In future, the use of Direct Magnitude Estimation (DME) may assist to reduce rater variability, as
this scaling method uses an external standard of the parameter in question to rate against (see
section 4.2.7.2 for further descriptions of DME and its application to the treatment studies in this
thesis).
Moreover, for those voice parameters with ĸw values below the clinical criterion,
variability between ratings occurred primarily within one-scale point. Online and face-to-face
ratings differed between normal and mild levels for phonation breaks (80% of the time ratings
differed), modal pitch (69.56%), and pitch variability (47.61%), and between mild and moderate
levels for vocal breathiness (62.5% of the time), roughness (56.25%), loudness level (66.66%) and
loudness variability (77.41%). These levels of variability are consistent with previous reports
regarding face-to-face evaluations of voice parameters, where raters were more likely to agree on
106
the presence of normal or severe vocal qualities and have greater variability when rating mild to
moderate severity levels of vocal breathiness and roughness (Gerratt, Kreiman, Antonanzas-
Barroso, & Berke, 1993; Kreiman & Gerratt, 1998; Kreiman, et al., 1993; Rabinov, Kreiman,
Gerratt, & Bielamowicz, 1995).
Along with rater variability, a further explanation for the lower ĸw values relates to the
statistic per se. It has been found that ĸw may be negatively influenced by the data distribution,
where despite high inter-rater agreement, calculations using non-linear data can result in low ĸw
(Cicchetti & Feinstein, 1990). For example, in the present study, strain-strangled vocal quality
(non-linear data), where although receiving comparable PCA to the vocal tremor parameter (linear
data), obtained only moderate agreement according to ĸw, in contrast to good agreement for vocal
tremor (Table 3.3). It is hypothesized, therefore, that the levels of agreement for some of the voice
parameters may have been influenced by this statistical phenomenon. For this reason, it may be
necessary to interpret ĸw alongside PCA for perceptual evaluations of speech and voice. Together,
the results in the present study indicated that an accurate and adequate evaluation of speech and
voice could be achieved online. Further support for the comparability of the online and face-to-face
assessment environments in the evaluation of all vocal parameters was provided by the reliability
measures obtained for these ratings. Intra- and inter-rater reliability were found to be similar in
each assessment environment (Table 3.5), suggesting that the online assessment was as reliable as
the traditional face-to-face assessment on this task.
3.4.2.2 Oromotor Parameters
For the purpose of assessing oromotor function, the two assessment environments were
found to be clinically comparable (PCA between 86.89% and 100%). See Table 3.4. Two of the 14
variables (masked facial expression and lip retraction), however, were below the clinical criterion
using ĸw. As noted with the voice parameters, a certain level of variability inherent in perceptual
ratings may have contributed to the lower ratings for these oromotor parameters. Previous face-to-
face evaluations of facial expression in PD using a range of rating scales and statistical analyses
have shown varying levels of intra- and inter-rater reliability including fair, moderate and
substantial agreement (Geminiani et al., 1991; Goetz et al., 1995; Hely et al., 1993; Martinez-Martin
et al., 1994). The authors of these studies attributed some level of the variability to the subjective
107
interpretation of severity levels, the possible inexperience of a few of the raters, and some
variability in consensus prior to rating.
Similarly, labial judgements which have been investigated predominantly in the cleft palate
literature have been associated with rater variability (Asher-McDade, Roberts, Shaw, & Gallager,
1991; Morrant & Shaw, 1996; Ritter, Trotman, & Phillips, 2002). Face-to-face evaluations of lip
retraction in participants with repaired unilateral cleft lip have shown poor (Morrant & Shaw, 1996)
and moderate levels of inter-rater agreement (Ritter, et al., 2002). The subjective interpretation of
severity levels has also been reported to affect rater agreement in these studies (Morrant & Shaw,
1996; Ritter, et al., 2002). In the present study, variability between face-to-face and online ratings
of masked facial expression and lip retraction was primarily within one-point on the rating scale,
and between the normal and mild severity levels for lip retraction (61.90% of the time ratings
differed), and mild to moderate levels for masked expression (80%). Further, it is possible that with
the progressive decline in smile intensity and facial expression in people with PD with the disease
process (M. C. Smith, Smith, & Ellgring, 1996; Tickle-Degnen & Lyons, 2004), the greater
consensus that occurred between raters in the present study for the moderate to severe facial
features compared to the milder levels of severity may be due to their more pronounced
manifestations. This may have made the features easier to rate compared to the milder levels.
Analyses of the oromotor parameters including masked facial expression and lip retraction using ĸ w
may also require a level of cautious interpretation due to the non-linear distribution of the data, and
results may need to be interpreted alongside PCA. A final consideration for the lower agreement on
the masked facial expression and lip retraction parameters may relate to factors in the online
environment that occasionally made ratings more difficult. For example, it was occasionally
evident that: (1) background lighting created a level of shadowing on the participant‟s face; (2) the
lack of webcam zoom function reduced contrast of the facial features; and (3) a level of image
pixelation occurred with the store-and-forward recordings. On these occasions, the online factors
primarily impacted upon the ratings of lip and tongue symmetry, tongue deviation and general
facial features. On the whole, however, the high intra- and inter-rater reliability obtained for the
oromotor parameters collectively, and the comparable levels between the two environments for the
majority of parameters were very encouraging.
108
The complete agreement obtained between the two assessment environments for OAP was
consistent with the previous studies by Hill et al. (2006; 2009b) where consensus agreement of
89.47 and 100% were observed between online and face-to-face ratings for consonant precision in
the earlier and later study respectively. Additionally, the high level of agreement between the
online and face-to-face ratings in the present study and the comparable intra- and inter-rater
reliability values (moderate agreement) between environments lends further support to the validity
of an online application.
3.4.2.4 Measures of Speech Intelligibility
The Bland and Altman LA were used for online and face-to-face scores of the ASSIDS.
For the sentence tasks, both the LA for sentence intelligibility and communication efficiency ratio
were within the clinical criterion, indicating that comparable measures of speech intelligibility can
be achieved between the online store-and-forward method and traditional face-to-face audio
recordings using this assessment. Hill and colleagues (2006; 2009b) similarly reported sentence
intelligibility values within their pre-determined clinical criterion, while communication efficiency
only reached the criterion in their later investigation. In the present study, the LA for word
intelligibility were outside the clinical criterion of ± 3.2% between the environments. It is possible
that speaker severity may have influenced these results. Differences of three or more words
between raters, which were outside the clinical criterion, occurred predominantly for participants
with moderately and severely reduced intelligibility (66.66% of the time), as identified on the OIC
task. The reduced speaker intelligibility may have contributed to the differences in ratings between
the two environments. Yorkston and Beukelman (1981a) acknowledge that transcription of word
tasks (as used in this study) in the traditional face-to-face environment is difficult with more
severely dysarthric speakers. For all the ASSIDS tasks, intra- and inter-rater reliability was largely
comparable between the two environments and showed very good agreement overall. These values
are in keeping with the very good reliability measures reported for the ASSIDS in previous face-to-
face literature (Yorkston & Beukelman, 1980, 1981b). Comparable levels between assessment
environments and previous literature lend further support to the use of this assessment in an online
application.
109
For the additional ratings of OIC, PCA within the clinical criterion was achieved. This
finding is in keeping with previous reports of high inter-rater agreement within one-scale point for
overall speech intelligibility in traditional face-to-face ratings (Sheard, et al., 1991). It is also in
keeping with the online consensus (95.83%) within the clinical criterion reported by Hill et al.
(2009b) in their re-designed study. In the present study, the ĸw further reflected good agreement
within the clinical criterion. Together with the comparable reliability measures, these findings
suggest that ratings of OIC can be made reliably using an online system.
3.4.3 Benefits and Challenges of Online Assessment
The current trials largely supported the validity and reliability of online assessment, and
have demonstrated that the features of the PC-based telerehabilitation system developed for this
study were appropriate for assessment delivery and ratings of the acoustic and perceptual
parameters. The telerehabilitation system described in this study allowed for objective
measurement of mean SPL, phonation duration and F0 in real-time and also provided adequate
quality audio-visual recordings via the store-and-forward function for the perceptual ratings of
speech and voice and oromotor parameters. Moreover, the majority of participants in the online-led
environment felt more than satisfied or very satisfied (80%) with the online modality overall, which
further supports the uptake of this modality. However, a number of challenges unique to the online
environment were identified by the assessing SLPs.
3.4.3.1 Audio Quality
The 128 kbit/s Internet videoconferencing connection provided adequate audio quality for
the delivery of task instructions and general communication between the SLPs and participants
during the assessments. Occasional audio delays of up to three seconds did occur during the
assessments, which had the potential to affect the participant-SLP interactions. However, the SLPs
and participants were able to quickly compensate for any disturbances in the audio by actively
waiting until the other had finished speaking before replying. The use of this compensatory strategy
and the ability to quickly adjust to the audio delay have also been noted in other studies using low-
bandwidth videoconferencing for psychotherapy and parent training in speech-language pathology
(Ghosh, et al., 1997; McCullough, 2001). In the present study, the audio delay was not reported by
110
those involved in the assessments to greatly impact on participant-SLP rapport during the sessions.
Furthermore, the majority (70%) of the participants in the online-led environment rated the audio
quality as adequate or excellent.
Although the audio quality during videoconferencing was appropriate for assessment
delivery, real-time ratings of the perceptual speech and voice parameters at this level were quite
challenging for the SLPs. In particular, it was difficult to rate the subtle speech features on the OAP
and speech intelligibility tasks in real-time. Consequently, to maximize the audio quality for
ratings, the store-and-forward feature of the telerehabilitation system was employed during the
sessions. With the store-and-forward, the audio quality was comparable to that of the minidisk
recorder used in the face-to-face environment and contributed to the high levels of agreement
between environments. In future, the use of telerehabilitation systems with high-bandwidth
videoconferencing would assist to improve the audio clarity in real-time, and potentially eliminate
the need for store-and-forward use in rating perceptual speech and voice parameters.
3.4.3.2 Video Quality
Unlike the audio quality, the video quality was the most compromised aspect of
videoconferencing. As a result of the frame rate and pixelated image especially with movement, the
facial features of the participants and SLPs were generally more difficult to discriminate in real-
time during the sessions. Difficulties in viewing have also been identified in other studies utilising
low-bandwidth videoconferencing for otolaryngology and occupational therapy consultation and the
assessment of schizophrenia and dysarthria (Chae, Park, Cho, Hong, & Cheon, 2000; Dreyer,
Dreyer, Shaw, & Wittman, 2001; Hill, et al., 2006; Hill, et al., 2009b; Sclafani et al., 1999; Zarate,
et al., 1997). Despite the limitations, the video quality in the present study did not negatively affect
online assessment or general participant-SLP rapport. The adequate or more than adequate ratings
of the video quality by the majority (63.33%) of participants in the online-led assessments further
confirmed this.
For the actual rating of oromotor parameters, however, the store-and-forward feature of the
telerehabilitation system was required. This improved the video quality and allowed for enhanced
111
detection and discrimination of fine motor movements and precision, and generally high levels of
consistency between the online and face-to-face ratings. However, even with store-and-forward,
some ratings were still more difficult, although not significantly compromising the level of
agreement and reliability between the two environments. As outlined previously (section 3.4.2.2),
the background lighting, lack of webcam zoom and level of image pixelation with the store-and-
forward recordings did make it more difficult on occasion to rate lip and tongue function and
general facial features. Difficulties in rating such movements due to webcam zoom and lighting
issues have also been reported by Hill et al. (2006) in their dysarthria pilot study. The authors
proposed that the use of web cameras with greater zoom and focus capabilities and telerehabilitation
systems using higher Internet bandwidth would possibly enhance the accuracy of future online
ratings (Hill, et al., 2006).
For the SLPs in particular, further difficulties in viewing the participants in real-time and
with the store-and-forward arose as a result of participant factors such as head and body dyskinesias
and stooped forward posture. For six of the 61 participants where these factors were pronounced, it
was more difficult for the SLPs to clearly view their entire face and judge aspects of lip retraction,
pucker, and tongue movements. However, these difficulties in oromotor judgment occurred in both
the online and face-to-face environment.
3.4.3.3 System Operation and Speech-Language Pathologist Challenges
Although the telerehabilitation system was very user friendly, on occasion, one aspect of
the online delivery presented a unique challenge for the online SLPs. Specifically, this related to
the differences in establishing eye contact online. Unlike in the face-to-face setting where eye
contact was achieved naturally, the online SLPs needed to actually look up at the web camera on
top of the monitor rather than at the participant on the screen in order to be perceived as making eye
contact. This process, and the need to operate the telerehabilitation system at the same time, made
the interaction with the participants somewhat unnatural for the SLPs. Montani and colleagues
(1997) suggested that difficulties in achieving eye contact could greatly impact on the success of the
online session as professionals may become distracted from the actual consultation as they try to
create the impression of eye contact. In the present study, however, the pre-assessment training in
operating the telerehabilitation system and the SLPs‟ growing experience in the use of the
112
telerehabilitation system during the course of the assessment phase helped to eventually create a
level of automaticity and normality, with minimal impact on assessment delivery. Nevertheless,
future applications with improved positioning of the web camera closer or juxtaposed to the
participant image would help to create a more natural representation of eye contact and reduce some
of the difficulties encountered for the online SLPs (Montani, et al., 1997).
3.4.4 Study Limitations and Future Directions
The present study contained a number of limitations which should be addressed in future
research in order to provide a comprehensive view of the validity and reliability of online speech
and voice assessment for PD. Firstly, only a small number of SLPs were involved in the assessment
administration and ratings. Future research should involve greater numbers of SLPs of varying
degrees of proficiency in online administration, in order to determine the possible effects of
proficiency on assessment outcome. Analyses of rater reliability in both the online and face-to-face
environments using larger groups of raters should also be investigated.
Secondly, an analysis of the effects of PD and hypokinetic dysarthria severity on the online
assessment administration and ratings was beyond the scope of this study. Although it was
identified above that participants with more pronounced physical symptoms such as head and body
dyskinesias and stooped forward posture were more difficult to assess online, it would be necessary
to look in-depth at such factors in order to determine the circumstances where telerehabilitation may
not be an appropriate mode of assessment.
Thirdly, as the online and face-to-face assessments in this study were conducted in a
laboratory setting, it is difficult to generalise the findings to the wider clinical setting. The next
stage of research should involve online assessment delivery in the “real-world” at either the
participant‟s home or within a community health centre (home telecare) to provide a realistic look
at the benefits and limitations of telerehabilitation assessment services for people with PD.
113
Finally, the present study only touched on a small number of issues relating to participant
satisfaction in the online environment which were mainly pertinent to the laboratory setting. In
order to identify the total benefits of telerehabilitation and its possible success as a service delivery
method, a comprehensive analysis of participant and as well as SLP satisfaction with this modality
in the real-world setting is required. An in-depth cost-benefit analysis (which was also beyond the
scope of this study) of the telerehabilitation service from the perspective of patients and healthcare
providers should also follow.
3.5 Conclusion
This study has demonstrated the validity and reliability of online assessment in the
laboratory setting for evaluating the speech and voice disturbances associated with hypokinetic
dysarthria and PD. The comparable ratings achieved for the majority of parameters between the
online and face-to-face environments and high rater reliability have confirmed the study hypothesis.
Additionally, high participant satisfaction with the online modality was achieved overall. Although
some challenges arose with online assessment, the SLPs and participants were able to appropriately
manage these without significant impact on the sessions. The PC-based telerehabilitation system
described in this study provides a basis for the delivery of online assessment for PD. This modality
may prove to be a necessary alternative or addition for people with PD, whereby lessening the
difficult access issues that exist for this population.
114
Chapter 4
Treating Disordered Speech and Voice in Parkinson’s Disease Online:
A Randomised Controlled Non-inferiority Trial
4.1 Introduction
As demonstrated in Chapter 3, online assessment of the speech and voice disorder

associated with hypokinetic dysarthria and PD using the custom made PC-based telerehabilitation
system was valid and reliable. In order to further determine the full potential of telerehabilitation as
a complete speech-language pathology service delivery for PD, the validity of treatment delivery
(LSVT®) via this modality also required investigating. Chapter 1 outlined the paucity of large-scale
PC-based telerehabilitation studies focusing on speech-language pathology treatment for adults with
neurological disorders (section 1.11.2). Further, only two proof-of-concept studies have actually
focused specifically on the online LSVT® delivery for PD, with only Howell et al. (2009)‟s study
being PC-based. In that study, three participants with IPD and mild to moderate hypokinetic
dysarthria were treated online within their own homes (Howell, et al., 2009). The treatment
sessions were conducted over videoconferencing via a broadband connection over Skype and the
system had the additional capability of recording audio samples for later review. Post-treatment
results were promising, with mean improvements noted in SPL for maximum sustained vowel
phonation (approximately 14.7 dB), reading (7.7 dB) and monologue tasks (6.6 dB) (Howell, et al.,
2009).
In an additional study, Tindall and colleagues (2008) used videophones to deliver the
LSVT® remotely to the homes of 24 participants with IPD and hypokinetic dysarthria (severity
level not specified). The authors reported statistically significant increases in mean SPL with
treatment for measures of maximum sustained vowel phonation (13.3 dB), reading (5.8 dB),
monologue (3.6 dB) and picture description (5.9 dB). High participant satisfaction with the use of
videophones was reported, as was a direct time and cost saving to participants.
115
Although both studies have touched on the potential of online LSVT® delivery, a number
of study limitations make it difficult to generalize the findings and to determine whether the
benefits obtained via online LSVT® delivery are comparable to those obtained face-to-face.
Limitations common to both studies relate to their design and to the telerehabilitation systems used.
Firstly, in relation to the study design, both studies did not employ a comparison face-to-face
treatment group. Additionally, only a small number of participants were included in Howell et al.‟s
(2009) study, and there was a potential performance bias for participants as they were assessed post-
treatment by their treating SLP.
Secondly, the relatively basic features of the telerehabilitation systems impacted on

treatment delivery. The systems in both studies did not allow the SLPs to incorporate all essential
elements of the LSVT®, in line with face-to-face delivery. For the LSVT® to ultimately be
successful, the SLP must be able to monitor the participant‟s vocal loudness (SPL) during all
treatment tasks, in order to aid calibration (Ramig, Pawlas, et al., 1995). In Tindall et al.‟s (2008)
study, an attempt at monitoring SPL was made by having the participants position the Digital SLM
during the session at their end, so that it could be viewed by the SLP over the videophone.
However, this practice may have introduced a level of variability to treatment, as the positioning of
the Digital SLM and its distance relative to the participant‟s mouth may have varied between
sessions, potentially affecting the accuracy of the SPL readings. In Howell et al.‟s (2009) study,
monitoring of SPL was not possible with videoconferencing and therefore, one treatment session
each week was conducted face-to-face in order to monitor SPL, build participant-SLP rapport and
provide treatment materials. This mixed treatment modality was therefore not truly representative
of face-to-face delivery. Both studies were also unable to transmit treatment materials to the
participants via the online systems, highlighting the potential difficulties of applying these methods
to the home setting.
There is a recognized need for online delivery of the LSVT® due to the intensive nature of
the treatment and the difficulties that arise for people with PD in accessing this treatment. The
option of receiving treatment remotely would provide numerous benefits for people with PD and
would help to drastically reduce the current barriers to service access resulting from the physical
incapacity of individuals, transport, travel difficulties and cost, and the large distances to specialised
healthcare facilities (section 1.6). To help overcome these access barriers and ensure best practice,
there is a need for further studies utilising online technology to closely replicate face-to-face
116
treatment delivery, and for research designs to incorporate comparison groups. The present study
aimed to investigate the validity of online LSVT® delivery using the PC-based telerehabilitation
system in a laboratory setting compared to face-to-face LSVT® for participants with IPD and mild
to moderate hypokinetic dysarthria. It was hypothesised that LSVT® can be effectively delivered
online and that the online treatment outcomes on the acoustic and perceptual measures will not be
inferior to those obtained face-to-face. Specifically, the treatment outcomes will not be inferior
compared to the: (1) participant group receiving face-to-face LSVT® in the laboratory setting; and
(2) outcomes described in the literature for traditional face-to-face LSVT® delivery.
4.2 Method
4.2.1 Study Design
This study was a single-blinded, prospective, randomized controlled non-inferiority trial.

The study was conducted according to the extension of the CONSORT guidelines for non-
inferiority trials (Piaggio, Elbourne, Atlman, Pocock, & Evans, 2006). Whereas the previous study
(Chapter 3) was concerned with the equivalence of the assessment, the present study was concerned
with treatment outcome and therefore, a non-inferiority design was appropriate.
4.2.2 Sample Size Calculation
Ethical clearance was obtained prior to commencement of the study from the Behavioural
and Social Sciences Ethical Review Committee of The University of Queensland. Prior to
participant recruitment, the sample size for a non-inferiority trial (Jones, Jarvis, & Ebbutt, 1996)
was calculated using the study‟s primary outcome measure of mean change in SPL following the
LSVT® on a 30 s monologue task. As stated previously (section 2.4.2.1), an improvement in
monologue loudness following the LSVT® provides a good indication of the level of carryover of
the louder voice to daily life, and a minimum improvement of 4.5 dB on this parameter following
the LSVT® is considered clinically relevant (Ramig, et al., 1996; Ramig, Countryman, et al., 1995).
The non-inferiority margin was set at 2.25 dB which is half of the minimum improvement of
4.5 dB, as recommended by Jones et al. (1996). Due to the lack of previously published studies
117
comparing performance with the LSVT® between the online and face-to-face environment at the
time of the study, the sample size was calculated using preliminary data collected in this study.
Analysis of the first eight participants to receive the online and face-to-face LSVT® in the current
study revealed a standard deviation of pre- to post-treatment difference scores of 2.48 dB. This
value was included in the power calculation. Using an alpha level of 0.05 and a statistical power of
80%, a minimum of 15 participants were required for each LSVT® environment (online and face-to-
face).
4.2.3 Participants
Allowing a 10% loss to follow-up, a total of 34 participants with IPD and hypokinetic
dysarthria were recruited to the study. The participants (27 males; 7 females) were aged between
54 and 85 years (M = 70.12 years; SD = 8.56). Descriptive characteristics of the participants are
summarized in Table 4.1. Difficulties recruiting females resulted in an uneven participant sample.
All participants had been diagnosed with IPD by a neurologist experienced in movement disorders,
with time post-diagnosis ranging from eight months to 30 years (M = 6.16 years; SD = 6.48).
Stages of PD as per the Hoehn and Yahr Scale ranged from I to IV with 27 participants rated at
Stages I-II and seven participants rated at Stages III-IV.
118
Table 4.1 Descriptive Characteristics of Participants who Received the LSVT® in the Online and
Face-to-Face Environments
Participant Age (years) Sex Time post Hoehn & Dysarthria

diagnosis (years) Yahr Stage severity level
Participants who received the LSVT® online (n =17)
1 80 M 2 1 Mild
2 56 M 6 1 Mild
3 69 M 2 1.5 Mild
4 59 M 4 1 Mild
5 77 M 2 1 Mild
6 66 M 1.5 1 Mild
7 71 F 1 1 Mild
8 54 M 4 1 Mild
9 58 M 0.7 1.5 Mild
10 81 M 7 2.5 Moderate
12 74 M - 2.5 Moderate
13 75 F 14 3 Moderate
14 69 M 22 2 Moderate
17 82 F 8 1 Moderate
Group Mean 70.65 5.39 1.59

(SD) (10.26) (5.76) (0.81)
119
Participant Age (years) Sex Time post Hoehn & Dysarthria
diagnosis (years) Yahr Stage severity level
Participants who received the LSVT® face-to-face (n =17)
18 68 M 3 1 Mild
19 59 M 11 1 Mild
20 67 M 2.5 1 Mild
21 68 M 1 1 Mild
22 66 M 5.5 1 Mild
23 74 F 1 1.5 Mild
24 75 M 7 1 Mild
25 76 F 5 1 Mild
26 69 M 9 2 Mild
29 62 F 4 2.5 Moderate
30 59 M 10 2.5 Moderate
32 84 M 16 1.5 Moderate
33 68 F 3 1.5 Moderate
Group Mean 69.59 6.88 1.53

(SD) (6.71) (7.19) (0.82)
Note. – unspecified; M = male; F = female; SD = standard deviation.
120
The participants were recruited from the assessment study (Chapter 3) if they met the
inclusion criteria for treatment which were: (1) the presence of hypokinetic dysarthria (ranging from
mild to severe) associated with diagnosed IPD that was impacting on communication; (2) a
videolaryngoscopic evaluation of the vocal fold structure and movement consistent with IPD; (3)
stimulability to increased loudness; and (4) a consistent drug regimen for IPD. Proficiency in the
use of computers was not a requirement for study inclusion as all aspects of the online treatment
were delivered remotely by the online treating SLPs. All exclusion criteria were consistent with the
assessment study (Chapter 3) and included a: (1) speech and/or language disturbance or a co-
existing neurological disorder inconsistent with PD; (b) severe uncorrected auditory and/or visual
disturbance; (3) cognitive disturbance inconsistent with the capacity to provide informed consent;
(4) respiratory dysfunction unrelated to the neurological disorder; and (5) a positive history of
alcohol abuse. Participants who had received the LSVT® within 12 months of the present study
were also excluded. The participants did not display any of the aspects that warranted exclusion.
Prior to the commencement of treatment, the participants underwent a videolaryngoscopic

examination by an Ear Nose and Throat specialist to examine the vocal fold structure and
movement. The examinations excluded any laryngeal pathology inconsistent with the effects of
IPD. Evaluations identified bowed vocal folds in 14 participants, vocal tremor in one participant
and a slight supraglottal constriction in another. For the remaining 20 individuals, the examinations
were unremarkable. All participants were cleared for inclusion in the treatment study.
Prior to treatment, all participants were classified by the principal investigator on the basis
of an overall severity level for hypokinetic dysarthria. This classification was made using the
assessment data in Chapter 3 which included mean SPL and perceptual ratings of OIC on a five-
point scale (1 = normal, completely intelligible speech; 5 = severely unintelligible speech with
difficulties deciphering many words) that were both rated from the monologue. As outlined in
Chapter 3 (section 3.2.2), SPLs above 65 dB in monologue loudness, together with a mild reduction
in overall speech intelligibility were classified as mild hypokinetic dysarthria. SPLs ranging from
60 dB to 65dB with an accompanying mild to moderate reduction in speech intelligibility
corresponded to moderate hypokinetic dysarthria. A severe hypokinetic dysarthria level was
considered to be consistent with SPL values below 60 dB and a moderate or severe reduction in
speech intelligibility. Using this classification, 18 participants in the study demonstrated overall
mild hypokinetic dysarthria, and the remaining 16 participants demonstrated moderate dysarthria.
121
Based on their dysarthria severity classification, the participants were then stratified, and randomly
assigned by the principal investigator to receive the LSVT® in either the traditional face-to-face or
online treatment environment. Randomisation occurred using a computerized random-number
generator (Dallal, 1997). As the participants entered in the study, randomization codes were
generated and assigned by the principal investigator to the participants in their order of recruitment
to the study. This process resulted in 17 participants in each treatment environment. Each group
comprised of nine participants with mild dysarthria and eight participants with moderate dysarthria
(Table 4.1). Figure 4.1 is a schematic flow chart of the participants through the study.
Figure 4.1 Schematic flow chart of participants through the study.
122
4.2.4 Speech-Language Pathologists
Four experienced and LSVT® certified SLPs from The University of Queensland
conducted the treatment sessions. The SLPs were randomized to both the treatment environments
and the participants within each environment using a computerized random-number generator
(Dallal, 1997). Due to other work commitments, the SLPs were randomly allocated to the study
within bounds of their availability. The same SLP delivered the entire treatment programme for a
participant. In total, SLP 1 delivered nine treatments in the online environment and seven face-to-
face; SLP 2 delivered five treatments online and seven face-to-face; SLP 3 delivered two treatments
online and face-to-face; and SLP 4 delivered a single treatment online and face-to-face.
4.2.5 The Lee Silverman Voice Treatment® Programme
All online and face-to-face laboratory treatment sessions were conducted within the
Telerehabilitation Research Unit at The University of Queensland. The treatment was delivered
intensively, one hour a day, four days a week over a four-week period, in accordance with the
LSVT® programme (Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ramig & Fox, 2004;
Ramig, Pawlas, et al., 1995). Daily sessions consisted of repetitive speech loudness drills and
hierarchical tasks that were performed at high intensity and maximum effort. The LSVT ® protocol
aims to promote increased respiratory drive, vocal fold adduction and carryover of the louder voice
into functional communication (see Figure 1.3). During the treatment tasks, the SLP monitored the
participant‟s vocal loudness (dB) and quality, and aided calibration (i.e. the participant‟s ability to
self-monitor and consistently use their louder voice in everyday communication). The LSVT® tasks
include:
(1) Daily Task 1: Maximum Duration of Sustained Vowel Phonation
This drill aimed to improve vocal loudness and duration of sustained phonation through
efficient respiratory drive and vocal fold adduction. In each session, participants sustained loud and
long phonations of /a/ during 15 repetitions, while maintaining good vocal quality.
123
(2) Daily Task 2: Maximum Fundamental Frequency Range
This exercise targeted improved F0 range through increased cricothyroid muscle

movement. For this task, participants practiced vocal glides, sustaining loud phonations of /a/ at the
highest and lowest F0 possible, with good vocal quality. Fifteen repetitions at each frequency level
were produced every session.
(3) Daily Task 3: Maximum Functional Speech Loudness Drill
Participants practiced a set of 10 “functional phrases”, five times each session, in a loud
voice. The phrases were common everyday expressions that were typically used by each
participant, such as “Good morning” or “Would you like a cup of coffee?” The habitual nature of
the phrases and focus on loud phonation facilitated carryover of the louder voice into daily
communication, by reminding participants to “think loud” and to maintain the high effort level
required to produce the phrases.
(4) Hierarchical Speech Loudness Drills
For the remainder of every session, participants practiced speech loudness tasks. Over the
course of the treatment, these tasks increased in difficulty from single word, phrase, sentence, and
paragraph reading, to conversational speech. The increased loudness and phonatory effort targeted
at each stage promoted calibration of the louder voice to everyday speech.
(5) Homework Tasks
Participants also practiced the daily tasks and hierarchical speech loudness drills at home
on a daily basis. This encouraged further carryover of the loud voice into the participant‟s everyday
environment.
124
4.2.6 Treatment Environment
4.2.6.1 Online Lee Silverman Voice Treatment® Environment
The PC-based telerehabilitation system described in the previous chapters was used for the
online treatments. For this study, an Internet connection was established at 128 kbit/s between the
two ends of the telerehabilitation system. The specific features of the telerehabilitation system
designed to deliver the LSVT® included the ability to: (1) conduct real-time videoconferencing at
320 x 240 pixel resolution; (2) display functional phrases and reading material for the hierarchical
speech tasks on the participant‟s screen; (3) adjust the remote web cameras for optimal viewing of
the participant; (4) capture real-time calibrated average recordings of SPL, F0 and duration via the
telerehabilitation system‟s acoustic speech processor and acoustic measurement software tool
during the daily tasks and speech loudness drills; and (5) capture high quality audio (windows
media audio codec Version 8 at 368 kbit/s) and video recordings (640 x 480 pixel resolution
compressed with the windows media video codec Version 8 at 384 kbit/s) as a separate function to
videoconferencing, and to review the files when needed via the store-and-forward function of the
telerehabilitation system.
Consistent with the assessment study (section 3.2.5.1), the two ends of the
telerehabilitation system were located in separate rooms and on different floors of the university
building. The online SLP was located at the SLP site and delivered the sessions over the
videoconferencing link to the participant. At the other site, the participant was seated comfortably
at approximately 50 cm from the PC monitor and wore a headset microphone at a set distance of
5 cm from the corner of the mouth. Prior to the sessions, the online SLP turned on the PC and
activated the application at the participant site. Further, to maintain the accuracy of the online
acoustic data (SPL, F0 and duration), the microphone distance was measured at the beginning of
each session by the online SLP, and the online mean SPL values obtained on a sample of at least
three sustained /a/ phonations were verified against a conventional Digital SLM held at
approximately 30 cm from the participant‟s mouth. All aspects of treatment were delivered
remotely by the SLP. Clinician training on the use of the telerehabilitation system for LSVT®
delivery was conducted in a three-hour session prior to the commencement of the study by the
principal investigator. The SLPs were familiarised with the specific features of the
telerehabilitation system that were necessary for the delivery of each LSVT® daily task. The SLPs
125
were deemed competent with online LSVT® when they could adequately deliver a mock treatment
session to face-to-face standard, within a one-hour time frame. The online treatment environment is
represented in Figures 4.2 and 4.3.
Figure 4.2 Online LSVT® environment at participant site. Note. (1) the videoconferencing system
displaying the SLP; (2) web cameras; and (3) telerehabilitation system‟s acoustic speech processor.
Figure 4.3 Online LSVT® environment at SLP site. Note. (1) the videoconferencing system
displaying the participant; (2) acoustic measurement software tool displaying SPL and F0 data; and
(3) the web camera.
126
4.2.6.2 Face-to-Face Lee Silverman Voice Treatment® Environment
The face-to-face LSVT® sessions were conducted in accordance with standard practice
1995). In this environment, the participant and SLP were seated at a table, facing each other. The
Digital SLM was kept at a set distance of 30 cm from the participant‟s mouth and was used by the
SLP to monitor the mean SPLs throughout the session. For the maximum duration of sustained
vowel phonation and maximum F0 range tasks, a generic stopwatch and chromatic tuner (KORG
Model DT-7) were used respectively. For the hierarchical speech loudness drills, the same reading
material used online was provided in hard copy. The face-to-face treatment environment is
represented in Figure 4.4.
Figure 4.4 Face-to-face LSVT® environment. Note. (1) the use of the Digital SLM; (2) reading
material; and (3) homework exercises.
4.2.7 Assessment of Outcomes
Within 48 hours of the final LSVT® sessions, the participants were re-assessed on the
primary outcome measure of mean change in SPL during the 30 s monologue, and secondary
measures on a section of the perceptual and acoustic parameters outlined in the assessment chapter
127
that have been used in the LSVT® literature as sensitive predictors of treatment change (Ramig, et
al., 1996; Ramig, Countryman, et al., 1995). For the post-LSVT® assessments, the participants
were assessed in the face-to-face environment. In addition, the three SLPs who had conducted the
pre-treatment assessments in Chapter 3 were randomized to the post-treatment assessments. Only
one SLP was required for each assessment. If, however, one of the SLPs had been the treating
clinician for a particular participant, they were not allocated to the post-treatment assessment.
During the post-treatment assessments, the participants were asked not to divulge their treatment
environment to the assessors. The set-up and procedure of the face-to-face assessments were
consistent with those conducted pre-treatment (section 3.2.5.2), with the only exception being that
the SLP was not required to wear headphones for communication with the online SLP as no online
assessments were performed post-treatment.
4.2.7.1 Acoustic Measures
The LSVT® Evaluation Protocol was used both pre-treatment (Chapter 3) and post-
treatment to obtain measures of mean SPL, maximum F0 range and maximum duration of sustained
vowel phonation. The protocol has been used routinely in the LSVT® literature for pre- and post-
treatment assessment. The assessment procedures for this section, including the use of the
telerehabilitation system to capture data, the averaging of values and the conversion of frequency
(Hz) data to semitones (ST), were consistent with the assessment study (see section 3.2.4.1). As
outlined earlier, each participant wore the headset microphone used in the assessment and online
treatment sessions, with the microphone distance measured at 5 cm from their mouth. The acoustic
measures were displayed on the PC in the face-to-face assessment room and were sampled by the
assessing SLP. During the assessments the participants were instructed to perform all tasks in a
comfortable voice and were not prompted by the SLPs to monitor their loudness level, phonation
time or frequency range. Mean SPLs were recorded on three assessment tasks. These included six
maximum phonations of /a/, reading of The Grandfather Passage and a 30 s monologue on a topic
of interest. The maximum duration of sustained vowel phonation was also recorded in seconds.
The maximum F0 range was calculated for each participant from the difference between the highest
and lowest F0 mean levels (Hz) obtained during the performance of six vocal glides to the highest
and lowest levels respectively. The range in Hertz (Hz) was then converted to a maximum
semitone range (ST) (de Pijper, 2007). For each participant, the data obtained pre-treatment in the
face-to-face environment (Chapter 3) and post-treatment (current study) were used in the analyses.
128
4.2.7.2 Perceptual Speech and Voice Parameters
Pre- and post-LSVT® readings of the Rainbow Passage were rated for each participant on
the parameters of vocal breathiness, roughness and OAP. In addition, the 30 s monologue
performed by each participant for the acoustic loudness measure was also rated according to
loudness level, pitch and loudness variability, and OIC. The recordings in the face-to-face
environment had been obtained using the minidisk recorder in both the assessment study and post-
treatment in the current study. Further, perceptual ratings of the speech and voice parameters were
made using DME. This method of scaling allows the listener to rate a speech or voice parameter
numerically, with no limits to the endpoints of the scale (Gescheider, 1976; Schiavetti, Metz, &
Sitler, 1981; Stevens, 1974). It has been suggested that this freedom in rating allows for more
accurate representation of parameters than equal appearing interval scaling (EAI), as raters may not
identify intervals as equal on a scale, and therefore should have greater freedom in their ratings
(Stevens, 1974). Together with the use of a “standard”, or an external reference of the speech or
voice parameter in question to rate against, DME scaling allows for improved validity and
reliability for rating speech and voice parameters compared to EAI scaling (Eadie & Doyle, 2002;
Schiavetti, et al., 1981; Schiavetti, Sacco, Metz, & Sitler, 1983; Toner & Emanuel, 1989; Whitehill,
et al., 2002; Zraick & Liss, 2000). DME scaling has been used to rate a number of perceptual
parameters relevant to the current study, including vocal breathiness and roughness, articulatory
precision, dysarthria severity and speech intelligibility (Hillenbrand, Cleveland, & Erickson, 1994;
Mackenzie & Lowit, 2007; Samlan & Weismer, 1995; Schiavetti, et al., 1981; Schliesser, 1985;
Toner & Emanuel, 1989; Toner, Emanuel, & Parker, 1990; Weismer, et al., 2001; Weismer &
Laures, 2002; Whitehill, et al., 2002; Zraick & Liss, 2000). For the current treatment study, DME
scaling was chosen as the most appropriate rating method for detecting post-LSVT® changes in the
perceptual speech and voice parameters. It was anticipated that DME scaling would allow for
higher rater reliability on the perceptual parameters than achieved with EAI scaling in the
assessment study of this thesis (section 3.3.2), and consequently, provide a better representation of
treatment change.
To obtain the speech samples for DME rating, pre- and post-LSVT® recordings of the
Rainbow Passage and 30 s monologues were made. The middle three sentences of the Rainbow
Passage were used for the ratings:
129
These take the shape of a long round arch, with its path high above, and its two ends
apparently beyond the horizon. There is, according to legend, a boiling pot of gold at one
end. People look, but no one ever finds it.
The pre- and post-LSVT® recordings were numerically coded and randomized to CD using
a computerized random-number generator (Dallal, 1997). The standards for each parameter were
also added to CD and these were readings and monologues of additional speakers with PD who had
taken part in the assessment study only. A different standard was used for each speech feature and
represented a moderate degree of impairment for each parameter (Stevens, 1975; Weismer &
Laures, 2002).
The DME ratings were made by the two independent SLPs that had rated the ASSIDS
parameters in the assessment study (section 3.2.4.2.4). These SLPs had not taken part in the
assessments or delivery of the LSVT®, and were blind to the intent of the study. Rater training was
conducted for 30 min prior to the ratings using additional stimuli that were not included in the data
analyses. A definition of each parameter was provided to the raters to assist them in identifying the
specific speech feature. Following standard DME procedures, one parameter was rated at a time
and the raters first listened to the standard, which was then repeated after every four speech samples
(Whitehill, et al., 2002). If a sample needed to be heard a second time, the rater would again listen
to the standard and then the four speech samples in that set. The volume was also kept at a
consistent level for all speech samples. For rating, a set value of 100 was assigned to each standard
(Schiavetti, et al., 1981; Whitehill, et al., 2002), and raters assigned values greater than 100 to
denote improvement for measures of loudness level, pitch and loudness variability, OAP and OIC.
For these parameters, a value of 200 indicated that the sample was rated as twice as improved in
quality as the standard, while a value of 50 denoted that the sample was only half as clear in quality
compared to the standard (Eadie & Doyle, 2002; Weismer & Laures, 2002). In contrast, for the
vocal breathiness and roughness parameters, values greater than 100 denoted increased impairment
in vocal quality. Following the DME ratings, the values given by the two independent SLPs were
averaged to express a single mean value for each sample. As the DME ratings are considered to
have a log normal distribution (Engen, 1971; Stevens, 1975), the means were then converted to
logarithmic values for inclusion in the statistical analyses, and represented as geometric means and
standard deviations.
130
Consistent with reliability analyses in the assessment chapter (section 3.2.6.4), reliability
of the DME method was also determined for the perceptual voice parameters as a group (vocal
breathiness and roughness, loudness level, loudness variability and pitch variability), OAP and OIC.
Intra- and inter-rater reliability of the logarithmic values for these parameters was calculated using
random effect intra-class correlation coefficients (ICC(2,1)) for 47% (n = 16) of participants. Half of
the data for analyses were taken from the participants‟ pre- and post-LSVT® assessments,
respectively. ICC values below 0.40 corresponded to poor-to-fair reliability; between 0.40 and 0.75
to moderate-to-good reliability; and values above 0.75 represented very good reliability (Fleiss,
1981).
In addition to the DME ratings for OIC, the ASSIDS was used to measure change in word
and sentence intelligibility (reading tasks) and communication efficiency with treatment. The pre-
and post-treatment face-to-face audio recordings for each participant were randomized and rated by
the two SLPs that were blind to the intent of the studies. Section 3.2.4.2.4 provides further details
on the ASSIDS procedure.
4.2.7.3 Participant Satisfaction Questionnaire
In addition to the perceptual and acoustic measures obtained, the participants who had
received online-LSVT® completed a satisfaction questionnaire post-treatment. A five-point rating
scale was used to assess satisfaction with: (1) the online treatment sessions (possible responses
ranging from would not participate again to would prefer these types of sessions to face-to-face
sessions); (2) the audio and video quality during the sessions (responses ranging from poor to
excellent); and (3) overall satisfaction with online treatment (ranging from not at all satisfied to
very satisfied).
4.2.7.4 Speech-Language Pathologist Comments
At the conclusion of the study, the SLPs were given the opportunity to comment
anecdotally on any aspects relating to the delivery of the LSVT® in the online environment.
131
4.2.8 Statistical Analyses
The primary analysis (non-inferiority of the online LSVT® modality) was evaluated by
firstly computing the within-group differences from baseline on the primary outcome measure of
mean SPL on the monologue task, and the 95% confidence interval (CI) of these differences. It was
then determined whether the CI, centred on the observed difference between the online and face-to-
face LSVT® groups, did lie entirely between the non-inferiority margin (± 2.25 dB) and zero.
Statistical analyses of the pre- and post-LSVT® data were then performed individually for each
parameter using repeated-measures general linear model (GLM). Time (pre- to post-LSVT®) was
the within-subjects variable, while the mode of the LSVT® delivery (online or face-to-face
environment) and participant severity level (mild or moderate dysarthria) were the between-subjects
variables. The interactions between the mode of the LSVT® delivery, participant severity level and
time were also calculated using GLM. For the analyses, the level of significance was set at p < .05.
4.3 Results
For the primary outcome measure of mean change in SPL with the LSVT® on the
monologue task, the 95% CI of the within-group differences from baseline was 1.41. This value
was found to lie entirely between the non-inferiority margin of ± 2.25 dB and zero. Therefore, non-
inferiority of the online LSVT® modality was confirmed.
Table 4.2 displays the pre- and post-LSVT® values, standard deviations, changes from
baseline and F and p values for the acoustic measures. For all acoustic measures (SPL parameters,
maximum duration of sustained vowel phonation and maximum F0 range), statistic analyses using
repeated-measures GLM showed non-significant main effects for the LSVT® environment,
dysarthria severity and interaction effects (p > .05). Statistically significant increases in mean SPL
on all parameters and maximum F0 range (p < .05) were evident with time (pre- to post-LSVT®).
However, for the maximum duration of sustained vowel phonation parameter, there was no
significant increase evident with time for either treatment environment.
132
Table 4.2 Acoustic Parameters: Means, Standard Deviations (in parentheses), F and p values
LSVT Env x
Online LSVT Face-to-Face LSVT LSVT Env Severity Severity x Time
Time
Parameter Pre Post Change Pre Post Change F p F p F p F p
Maximum sustained 73.49 83.78 10.29 73.45 83.75 10.30 0.01 .940 0.02 .894 1.29 .266 197.17 <.001*
vowel phonation (dB) (4.55) (3.88) (4.01) (5.49) (4.38) (4.42)
Reading (dB) 68.81 73.84 5.04 67.84 72.81 4.97 <.01 .956 1.51 .228 0.03 .867 94.15 <.001*
(2.86) (2.52) (2.85) (3.23) (3.65) (3.07)
Monologue (dB) 67.39 71.25 3.87 66.74 70.62 3.88 <.01 .984 2.52 .123 0.28 .599 65.17 <.001*
(3.01) (2.75) (2.27) (3.08) (3.68) (3.24)
Maximum duration 10.65 11.26 0.61 11.08 11.32 0.24 0.60 .810 <.01 .998 0.06 .814 0.29 .598
of sustained vowel (4.57) (4.71) (4.09) (5.22) (5.15) (4.84)
phonation (s)
Maximum F0 11.04 13.71 2.68 9.69 11.98 2.29 0.02 .880 0.03 .871 0.03 .877 4.31 .046*
range (ST) (5.91) (4.82) (8.11) (5.01) (4.92) (4.94)
Note. pre = pre-LSVT®; post = post-LSVT®; = higher or lower post-treatment value; * = significant effect; severity = hypokinetic dysarthria
severity; time = pre- to post-LSVT®; Env = environment; change = treatment change; F0 = fundamental frequency.
133
The pre- to post-LSVT® values, standard deviations, F and p values on the perceptual
measures are displayed in Table 4.3. Analyses using repeated-measures GLM revealed non-
significant main effects (p > .05) for the LSVT® environment, participant severity level and
interaction effects on the LSVT® outcomes for all perceptual measures (perceptual voice
parameters, OAP and measures of intelligibility). Statistical analyses did reveal a significant effect
of time (pre- to post-LSVT®) for the majority of perceptual measures (Table 4.3). This effect
included improvements pre- to post-treatment on all the perceptual voice parameters (breathiness,
roughness, loudness level, loudness variability and pitch variability), the word and sentence
intelligibility tasks of the ASSIDS, and OIC. OAP and the communication efficiency ratio
determined from the ASSIDS were the only perceptual parameters that did not improve
significantly with time (p > .05).
134
Table 4.3 Perceptual Parameters: Means, Standard Deviations (in parentheses), F and p values
Face-to-Face LSVT Env x

Online LSVT LSVT LSVT Env Severity Severity x Time Time
Pre Post Pre Post F p F p F p F p

Perceptual Parameter
Breathinessa 77.20 62.14 84.12 63.53 0.42 .523 0.12 .737 0.53 .473 26.21 <.001*
(12.52) (9.59) (17.14) (11.24)
Roughnessa 88.39 85.33 97.10 84.84 2.02 .166 0.35 .557 2.88 .101 6.14 .019*
(10.50) (11.74) (12.48) (10.90)
Loudness levela 158.74 195.16 157.07 201.23 0.42 .522 0.02 .894 0.21 .653 11.18 .002*
(10.81) (21.12) (10.70) (21.78)
Loudness variabilitya 127.20 153.43 125.40 146.25 1.10 .303 2.79 .106 0.65 .429 73.70 <.001*
(10.63) (7.37) (14.89) (14.13)
Pitch variabilitya 100.00 115.77 101.20 110.48 0.23 .636 0.80 .378 2.32 .139 92.87 <.001*
(11.88) (11.19) (16.42) (14.20)
OAPa 132.74 142.00 127.91 128.03 1.60 .216 0.15 .703 4.14 .051 1.42 .243
(12.83) (11.86) (12.36) (18.72)
135
Face-to-Face LSVT Env x
Online LSVT LSVT LSVT Env Severity Severity x Time Time
Pre Post Pre Post F p F p F p F p

Perceptual Parameter
OICa 117.60 123.85 108.49 123.82 2.58 .119 2.42 .131 .024 .877 13.05 .001*
(12.73) (11.97) (19.20) (21.01)
WI (%) 94.38 96.31 88.47 92.53 1.81 .189 1.42 .243 0.04 .849 13.92 .001*
(4.65) (2.47) (12.45) (9.68)
SI (%) 98.02 99.23 95.73 98.31 1.87 .182 2.11 .157 0.10 .750 13.69 .001*
(2.38) (1.22) (5.87) (3.38)
CER 1.01 0.96 0.89 0.84 <.01 .972 0.58 .451 0.49 .492 3.63 .067
(0.22) (0.21) (0.22) (0.26)
Note. pre = pre-LSVT®; post = post-LSVT®; * = significant effect; severity = hypokinetic dysarthria severity; time = pre- to post-LSVT®; Env =
environment; aParameters measured using Direct Magnitude Estimation; OAP = overall articulatory precision; OIC = overall speech intelligibility in
conversation; WI = word intelligibility; SI = sentence intelligibility; CER = communication efficiency ratio.
136
4.3.3 Reliability of Direct Magnitude Estimation Scaling
For the perceptual voice parameters using DME scaling, intra-class correlations revealed
very good intra-rater (ICC = 0.96 Rater 1; ICC = 0.86 Rater 2) and inter-rater reliability (ICC =
0.86). Moderate intra-rater (ICC = 0.55 Rater 1; ICC = 0.50 Rater 2) and very good inter-rater
reliability (ICC = 0.82) was achieved for OAP. Additionally, for OIC, very good intra-rater (ICC =
0.89 Rater 1; ICC = 0.96 Rater 2) and moderate inter-rater reliability (ICC = 0.44) was obtained.
4.3.4 Online Lee Silverman Voice Treatment® Delivery
For all 17 participants who received the LSVT® online in this study, the treatment was
successfully delivered using the PC-based telerehabilitation system operating on a 128 kbit/s
Internet connection. Out of the 272 online sessions delivered during the study, none failed for any
of the participants. Additionally, for seven of the 17 participants, all of their sessions ran very
smoothly, without technical difficulties and with adequate audio and video quality for treatment
delivery. The remaining 10 participants encountered some difficulties during one of their sessions.
These difficulties included: (1) some networking issues that considerably compromised the audio
and video quality (five sessions); (2) the headset microphone becoming faulty and making it
difficult for the SLP to hear the participant (one session); and (3) the telerehabilitation system‟s
acoustic speech processor malfunctioning and the SPL and F0 data not being able to be obtained
(four sessions). However, all difficulties were appropriately addressed and the sessions were able to
continue.
4.3.5 Participant Satisfaction Questionnaire
Figures 4.5 to 4.8 display the participant ratings on the satisfaction questionnaire for the
participants who received online LSVT®. The majority of participants: (1) were comfortable or
very happy while participating in the online sessions (94.12%); (2) rated the audio as adequate or
more than adequate (76.47%); and (3) overall, were more than satisfied or very satisfied with online
treatment (82.35%). For the video quality, only 52.94% of participants found the quality to be
adequate or more than adequate for treatment, with the remainder rating it as inadequate or poor.
137
Figure 4.5 Participant satisfaction with the individual online treatment sessions.
Figure 4.6 Participant satisfaction with the online audio quality.
138
Figure 4.7 Participant satisfaction with the online video quality.
Figure 4.8 Participant satisfaction overall with online treatment.
4.4 Discussion
For the primary outcome measure of mean change in SPL on the monologue task with the
LSVT®, the 95% CI of the within-group differences from baseline were within the non-inferiority
margin, thus confirming non-inferiority of the online LSVT® modality. Furthermore, the results
obtained in the present study on the secondary outcome measures supported the study hypothesis
139
that online LSVT® is valid. For the participants in both the online and face-to-face LSVT®
environments, non-significant effects of treatment environment, participant severity level and the
interaction of these factors were found for all the acoustic and perceptual measures post-LSVT®.
The treatment gains for the majority of the parameters were statistically significant with time and
comparable between the two LSVT® environments. While findings indicated that comparable
treatment outcomes could be achieved in the online and face-to-face environments, the study also
revealed that the online LSVT® was clinically effective in improving the speech and voice of people
with PD. Indeed, significant improvements were identified across the majority of acoustic and
perceptual parameters.
4.4.1.1 Sound Pressure Level Tasks
For the maximum sustained vowel phonation, reading and monologue tasks, statistical
analyses disclosed significant improvements in mean SPL with the LSVT® for both the online and
face-to-face participants. These findings are consistent with the statistically significant
improvements reported on such SPL tasks in previous LSVT® studies conducted face-to-face with
IPD participants of similar dysarthria severity levels (El Sharkawi, et al., 2002; Liotti, et al., 2003;
Ramig, et al., 1996; Ramig, Countryman, et al., 1995; Ramig, Sapir, Countryman, et al., 2001;
Ramig, Sapir, Fox, et al., 2001; M. E. Smith, et al., 1995; Ward, et al., 2000). Post-treatment
changes in SPL are seen as direct effects of improved respiratory drive and vocal fold adduction
following the LSVT®, and the further carryover of the louder voice into functional communication
(Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ramig, Pawlas, et al., 1995). It has also
been suggested that these changes in SPL contribute to the perceived improvements in voice
parameters and speech intelligibility (Baumgartner, et al., 2001; Ramig, et al., 1994; Ramig,
Countryman, et al., 1995; Sapir, et al., 2002), which have been similarly observed in the present
study.
The treatment gains in mean SPL in the present study were comparable between the online
and face-to-face LSVT® groups. However, these changes in maximum sustained vowel phonation
(M = 10.29 dB, SD = 4.23 online; M = 10.3 dB, SD = 5.26 face-to-face), reading (M = 5.04 dB, SD
140
= 2.85 online; M = 4.97, SD = 3.14 face-to-face), and monologue loudness (M = 3.87 dB, SD = 2.22
online; M = 3.88 dB, SD = 3.23 face-to-face) were lower than the minimal improvements in mean
SPL reported with face-to-face LSVT® for IPD participants with mild to moderate hypokinetic
dysarthria. Efficacy studies by Ramig and colleagues (1995, 1996) have reported mean
improvements in SPL following the LSVT® of 13 dB and 14.03 dB for maximum sustained vowel
phonation, 8.03 dB and 8.75 dB for reading loudness, and 4.5 dB and 4.68 dB for monologue
loudness respectively (SD not reported in these studies). However, the clinical relevance of the
treatment gains in the present study should be interpreted alongside the pre- and post-treatment
mean SPL values. On the majority of tasks, participants in both the online and face-to-face LSVT®
groups achieved higher pre- and post-treatment mean SPL values than the previous studies.
Additionally, the post-treatment mean SPL values in the current study were comparable to the
values reported for two groups of healthy older adults speaking at comfortable loudness levels in
the face-to-face setting (Fox & Ramig, 1997; Ramig, Sapir, Fox, et al., 2001). Figure 4.9 displays
the post-treatment mean SPLs of the comparative groups mentioned above. Please note that for the
healthy older individuals, the figure only displays the highest mean SPL values that were obtained
from their two phonation trials on each task.
Figure 4.9 Comparison of post-treatment mean SPLs for IPD participants in the online and face-to-
face laboratory LSVT® study with those reported in the efficacy studies of face-to-face LSVT® and
healthy-older adults. Note. IPD Online and IPD FTF are the participants who received online and
face-to-face LSVT® in this study respectively; Ramig, Countryman, et al. (1995) and Ramig et al.
(1996) are studies of face-to-face LSVT® for IPD; Ramig et al. (2001) and Fox and Ramig (1997)
are studies of healthy older adults. Sustained „ah‟ is the maximum sustained vowel phonation task.
141
It is acknowledged that comparisons were made to control data involving small samples
sizes (n = 14 healthy participants in each study). However, these studies were similar in their
assessment procedures to the current study. It was therefore felt that comparisons with the control
data were appropriate and provided a good indication of treatment success and clinical significance.
Overall, the statistically significant improvements achieved on all SPL tasks in the present study,
together with the comparable post-treatment values to the control data suggest that participants with
IPD in both online and face-to-face treatment environments were able to achieve near-normal
loudness levels post-treatment.
4.4.1.2 Maximum Duration of Sustained Vowel Phonation
For this acoustic measure, statistically significant main effects of time (pre- to post-
LSVT®) were not observed for either the online or face-to-face IPD participants. Overall, however,
the treatment changes for both environments (M = 0.61 s, SD = 4.09 online; M = 0.24 s, SD = 4.84
face-to-face) were markedly lower than the 3.72 s and 4.9 s mean improvements reported by Ramig
and colleagues in the efficacy studies (1994; 1995). The post-treatment mean duration values
(M = 11.26 s, SD = 4.71 online; M = 11.32 s, SD = 5.15 face-to-face) in the present study were also
lower than the 17.94 s (SD = 5.01) mean recording for healthy older adults, which was the longest
duration of their three trials (Fox & Ramig, 1997). It is possible that the absence of change in
duration on this task is related to the participants‟ focusing mainly on vocal quality during the task
rather than duration, and as a result, not demonstrating a significant change in duration in
conjunction with an increase in loudness. The clear quality of the post-treatment phonations for
participants in both groups, compared to the pitch breaks, vocal roughness and breathiness
perceived pre-treatment, further suggests that participants learned to self-monitor the quality of their
phonations as part of calibration, which was an important treatment outcome.
4.4.1.3 Maximum Fundamental Frequency Range
For the maximum F0 range task, statistically significant improvements with LSVT® were
achieved for the participants in both treatment environments. This finding is in keeping with the
significant gains reported face-to-face for maximum F0 range, F0 and variability on reading,
monologue tasks and vocal glides (Ramig, et al., 1994; Ramig, et al., 1996; Ramig, Countryman, et
142
al., 1995; Ramig, Sapir, Countryman, et al., 2001). The findings are also consistent with the
perceived changes in pitch variability in reading, as noted in the present study, and previously
(Ramig, et al., 1994; Ramig, Countryman, et al., 1995). Improvements in maximum F0 range have
been associated with improved cricothyroid muscle activity (Ramig, et al., 1994). Although
significant gains were achieved on this parameter, the post-treatment mean increases in maximum
F0 range (M = 2.67 ST, SD = 8.11 online; M = 2.29 ST, SD = 4.94 face-to-face) in the present study
were lower than the 4 ST minimum improvement (SD not included) reported by Ramig and
colleagues (1994).
Participants in the online and face-to-face treatment environments showed significant

improvements on all of the perceptual voice parameters following treatment (Table 4.3). These
changes included perceived reductions in the levels of vocal breathiness and roughness in reading,
and increases in pitch variability, loudness level and loudness variability in conversational speech.
The results are in keeping with the literature that has reported significant post-LSVT®
improvements in vocal breathiness and hoarseness, loudness level and pitch variability for IPD
participants with mild to moderate dysarthria who were treated in the conventional manner. In
these studies, improvements were perceived by the participants themselves, their families, and
assessing SLPs (Baumgartner, et al., 2001; Ramig, et al., 1994; Ramig, Countryman, et al., 1995;
Sapir, et al., 2002). Such improvements in vocal quality have been associated with increased
respiratory drive and vocal fold adduction resulting from increased SPL, as well as more efficient
cricothyroid muscle movement with the LSVT® training (Baumgartner, et al., 2001; Fox, et al.,
2002; Ramig, et al., 1994; Ramig, Countryman, et al., 1995; Ramig, Pawlas, et al., 1995). As the
previous studies have used visual analogue scales rather than DME to determine treatment changes,
and primarily participant and family ratings, direct comparisons of results with the current study are
difficult. Nevertheless, the improvements made by participants in the present study which were
largely comparable between the online and face-to-face groups suggested that online delivery of the
LSVT® is valid.
143
Significant gains in OAP with treatment were not achieved for the participants in either
environment. This finding is somewhat surprising, as previous studies of IPD participants with
predominantly mild to moderate dysarthria such as in this study have highlighted changes in
articulatory movements with the LSVT®. These post-LSVT® changes were measured as
improvements in: (1) extent and rate of movement; (2) vowel and whole word duration; (3)
transition duration; (4) rise time; (5) vowel formants; (6) frication duration; (7) enhanced tongue
strength and endurance; and (8) increased tongue-base swallowing function (Dromey, et al., 1995;
El Sharkawi, et al., 2002; Sapir, et al., 2007; Ward, et al., 2000). Changes at the articulatory level
are suggestive of the global effects of loud phonation extending beyond the respiratory and
phonatory subsystems (Dromey, et al., 1995; Fox, et al., 2006; Sapir, et al., 2007; Ward, et al.,
2000). In the present study, only minor increases in OAP were perceived for the participants
(M = 9.26, SD = 0.13 online; M = 0.12, SD = 0.10 face-to-face). This finding may suggest that the
perceptual rating process for this parameter was unable to detect specific changes in articulatory
precision with treatment. This may be due to the fact that the pre-LSVT® DME values
(M = 132.74, SD = 12.83 online; M = 127.91, SD = 12.36 face-to-face) were already above the
standard. It is possible that the participants in general, did not present with obvious impairments in
articulatory precision per se prior to treatment, and as such, substantial post-treatment changes were
not identified. In future studies, physiological investigations of tongue function with
instrumentation such as electromagnetic articulography may provide more conclusive evidence of
treatment specific changes in articulatory precision.
4.4.2.3 Speech Intelligibility
On the evaluation of OIC using DME scaling, statistically significant improvements with
treatment were found for the participants in both treatment environments. These results are in
keeping with the improvements in conversational speech intelligibility in previous face-to-face
LSVT® studies, which were perceived by assessing SLPs, participants and their families using
visual analogue scales (El Sharkawi, et al., 2002; Ramig, et al., 1994; Ramig, Countryman, et al.,
1995). This improvement in speech intelligibility is also consistent with the significant
improvements noted on the word and sentence intelligibility tasks of the ASSIDS below.
144
On the ASSIDS, participants in both treatment environments made statistically significant
gains in word and sentence intelligibility with treatment. Moreover, the post-LSVT® sentence
scores, which have been suggested by Yorkston and Beukelman (1981a) to be the most
representative of functional speech out of all ASSIDS tasks, demonstrated that participants in both
treatment environments had improved in speech intelligibility from relatively mild impairment to
near normal levels. Such findings of improved speech intelligibility with the LSVT® are also
consistent with a face-to-face study of a group of 18 IPD participants with mild to severe
hypokinetic dysarthria who made significant gains on the ASSIDS sentence intelligibility task post-
treatment, approaching the normal range (Ward, et al., 2000). Improved intelligibility with the
LSVT® has been associated with the global effects of loud phonation, where enhanced motor drive
and coordination of movement throughout the speech subsystems are perceived at this overarching
level (Countryman, et al., 1994; Dromey & Ramig, 1998; Fox, et al., 2006; Scott & Caird, 1983;
Ward, et al., 2000). Although the treatment gains for word and sentence intelligibility were
significant in the present study, the levels of improvement on both parameters fell within the levels
of test-retest variability reported in the ASSIDS manual for dysarthric speakers assessed in the face-
to-face environment (Yorkston & Beukelman, 1981a). This is likely due to the ceiling effect of the
ASSIDS where it has been suggested that ongoing recovery for speakers with mild dysarthria and
those with sentence intelligibility scores above 90% (as seen pre-treatment in both the online and
face-to-face groups), may not be adequately represented on this assessment (Yorkston &
Beukelman, 1981a). Furthermore, if participants in the present study were to demonstrate treatment
gains above the speaker variability on the sentence intelligibility task, the post-LSVT® values
online and face-to-face would have exceeded the 100% upper limit of the assessment range. In the
absence of more representative standardized assessments, the current treatment gains on the various
measures of intelligibility were very promising.
On the final measure of the ASSIDS, significant post-treatment improvements in

communication efficiency were not achieved for the IPD participants in either treatment
environment. The post-LSVT® communication efficiency ratios were in fact lower than pre-
treatment (M = 4.95% lower online; M = 5.62% lower face-to-face), which was unexpected. This
finding can be explained in relation to the calculation of the communication efficiency ratio. For
the majority of the participants in both treatment environments (64.71% of online participants;
70.59% face-to-face), the reading time was longer post-LSVT® (M = 0.20 min longer, SD = 0.16
online; M = 0.32 min longer, SD = 0.26 face-to-face). This increase in time contributed to lower
efficiency values in the calculations, despite the significant gains obtained in sentence intelligibility.
145
A longer reading time is an anticipated positive effect of the LSVT® that contributes to improved
speech intelligibility. The louder phonation with treatment has been found to decrease speech rate
by increasing pause length and reducing the duration of utterances during reading and
conversational talks (Ramig, Countryman, et al., 1995). Goberman and Elmer (2005) also
identified slower articulation rates in reading and conversational tasks for participants with PD in
their study following prompting to use clear speech. With these factors in mind, the communication
efficiency parameter alone may not clearly represent the post-LSVT® changes and results should be
interpreted alongside those of word, sentence and conversational speech intelligibility.
4.4.3 Benefits and Challenges of Online Lee Silverman Voice Treatment® Delivery
The results of the present study indicate the validity of online LSVT®. The features of the
telerehabilitation system were conducive to treatment delivery. The 128 kbit/s videoconferencing
rate provided sufficient audio and video quality for the treatment sessions. Additionally, the
features of the telerehabilitation system including the real-time display and sampling of average
recordings of SPL, F0 and duration data and the store-and-forward facility, incorporated all the
essential components for successful treatment delivery, to face-to-face standard. Furthermore, the
application was user-friendly and easy to operate and allowed the SLPs to focus their attention
appropriately on the LSVT® goals of training loud phonation and carryover. Moreover, there were
no failed treatment sessions and the great majority of online participants (82.35%) were either more
than satisfied or very satisfied overall with their participation in the online sessions, which was very
encouraging (Figure 4.8). It is acknowledged, however, that a number of challenges unique to the
online environment arose for the participants and SLPs. These challenges were identified
anecdotally by the SLPs and are discussed in the following sections.
Participant-SLP rapport is considered an important feature for general face-to-face

consultations (Mekhjian, Turner, Gailiun, & McCain, 1999). For the delivery of the LSVT®,
establishing a good working relationship between the participant and SLP is particularly important,
as it aids the optimal treatment delivery over the one month period, and promotes carryover of
treatment effects into everyday communication (Ramig, Pawlas, et al., 1995). For online LSVT®,
146
the audio quality was considered integral to treatment delivery and participant-SLP rapport. In
general, the audio quality with videoconferencing at 128 kbit/s was sufficient for all treatment
related purposes. This included: (1) the delivery of treatment instructions and feedback; (2) shaping
of phonations when needed; (3) monitoring loudness level and vocal quality throughout the
sessions; as well as (4) maintaining appropriate participant-SLP rapport. The appropriateness of the
audio quality for treatment was reflected in the participant satisfaction surveys where the majority
(76.47%) of participants rated the audio quality as adequate or more than adequate (Figure 4.6).
Similar reports of adequate audio quality with videoconferencing at 33 kbit/s and 128 kbit/s have
been noted in online psychiatry studies where the audio quality was sufficient for participant self-
expression and rapport during assessment and treatment (Chae, et al., 2000; Ghosh, et al., 1997).
Although the audio quality with videoconferencing was adequate for treatment delivery in
the present study, occasional audio delays of up to three seconds were encountered during some of
the online sessions. As noted in the assessment study (section 3.4.3.1), these delays could
potentially affect the communicative interactions. However, the participants and SLPs were able to
quickly accommodate for any disturbances caused by the audio delay even within the first few
treatment sessions. A useful strategy that aided this process was waiting until the speaker had
clearly finished before responding. This strategy was consistently applied throughout the treatment
block, and sessions were able to continue with minor disturbances. Similar adaptations to the audio
delay leading to minimal disruptions to sessions have also been reported in other online studies
utilising 128 kbit/s Internet connections for psychotherapy and parent training in speech-language
pathology (Ghosh, et al., 1997; McCullough, 2001).
Where necessary, the SLPs in the present study also adopted further strategies to help
minimise the effects of the audio delay on treatment delivery. The SLPs used shorter and more
precise instructions during tasks to facilitate the flow of the conversation and to minimize the
likelihood of the SLP and participant talking on top of each other. On occasion, the audio delay did
make it more difficult for the SLPs to provide prompt feedback or additional encouragement or
coaching during the actual task performance (e.g. during sustained vowel phonations). With
feedback, participants would often stop speaking to attend to what the SLPs were saying, or due to
the delay, the feedback would not be heard at the appropriate time. To overcome such difficulties,
the SLPs would often wait until the task was completed before providing feedback or use obvious
and easy to detect hand-cues for quick input. The cues included “stop” to terminate the task, hand
147
raising to indicate the need to increase loudness or “thumbs-up” for appropriate performance and to
continue with the task. The SLPs were able to quickly adopt the use of such strategies during
treatment, and the majority of participants were also able to easily follow this routine. On occasion,
the participants with moderate dysarthria and a level of cognitive difficulty would require greater
prompting from the SLPs for shaping the desired vocal responses, especially during the first few
treatment sessions. For these participants, the SLPs successfully utilized as many verbal and non-
verbal strategies as necessary. The predictable nature of the LSVT® sessions in general, greatly
aided the uptake of the online modality for all participants and minimized any difficulties that
potentially related to the audio quality. In future, the use of applications with higher Internet
bandwidth may assist in more closely resembling the face-to-face modality, by further reducing any
audio delays and the need for compensatory strategies.
Finally, on a few occasions (six of the 272 sessions), some unexpected disturbances in the
audio quality arose during the treatment sessions. In five instances, suspected congestion on the
local area network considerably compromised the audio quality. On another occasion, the faulty
participant headset microphone created a great level of static which made it difficult for the SLP to
hear the participant. These difficulties were quickly resolved by the SLPs during the sessions by
disconnecting and re-establishing the videoconferencing connection between the two ends of the
telerehabilitation system and by replacing the headset microphone, respectively. As all sessions
were able to continue, these additional and intermittent difficulties in the audio quality were not
seen to adversely affect the study.
In comparison to the face-to-face modality where viewing of the participant was optimal,
the video quality with videoconferencing at 128 kbit/s was relatively poor. The same difficulties
encountered in the assessment study such as the frame rate and pixelated image especially during
movement (section 3.4.3.2), made it more difficult for the participants and SLPs to clearly view
each other during the sessions. Consequently, the participants rated the video quality during
treatment less favourably than the audio quality, with only 52.94% of participants indicating that the
audio quality was adequate or more than adequate for treatment (Figure 4.7). Lower participant
satisfaction ratings for video quality have similarly been reported in online physiotherapy and
148
psychiatry studies using 18 kbit/s and 384 kbit/s Internet bandwidth respectively, where factors such
as reduced picture quality and pixelation compromised satisfaction (Russell, et al., 2003; Zarate, et
al., 1997). As suggested in the assessment study of this thesis, the use of higher Internet bandwidth
for online systems would assist to improve the clarity of the visual image during sessions and in
turn, participant satisfaction (Hill, et al., 2006; M. Waite, et al., 2006; Zarate, et al., 1997).
For the SLPs, the reduced video quality presented additional challenges in delivering the
LSVT®. The web cameras allowed for only limited viewing of the participant during the session
compared to the face-to-face environment. Even with the adjustment of the web cameras to the
participant‟s face, the SLP was limited to viewing only the head to the upper torso. This restriction,
together with the pixelated image during videoconferencing, often made it more difficult for the
SLP to detect the presence of certain factors such as: (1) whether the participant was phonating
effectively at the top of the breath with an open-mouth position (during sustained phonation and
maximum F0 range drills); (2) the presence of muscle tension in the head and neck regions
(contraindicated for treatment); and (3) the optimal participant positioning and posture during the
sessions. It has been recommended with the LSVT® that such factors need to be closely monitored
during the sessions as they may impact on the ability to achieve the target loudness level, vocal
quality and carryover (Ramig, Pawlas, et al., 1995).
However, as noted with the audio quality above, the participants and SLPs were able to
compensate for the reduced video quality. Indirectly, the structured nature of the LSVT® and
reliance primarily on verbal communication helped to promote the most optimal viewing of the
participants and SLPs that was possible with videoconferencing at this bandwidth. As the LSVT®
is not a physical type of therapy, both the participants and SLPs were able to sit relatively still in
front of the PC during the sessions, which allowed for the least image pixelation. Additionally,
where clearer viewing of the participants was needed, such as to monitor the factors mentioned
above, the SLPs were able to record the desired task using the store-and-forward function and play
it back quickly before providing feedback. One drawback of the store-and-forward, however, was
that continual reviewing of tasks during the treatment session would often disrupt the flow of the
session, and feedback to aid calibration was not provided in a timely manner. As a result, the store-
and-forward was only used when necessary. For the majority of participants, this was mainly
required during the first few treatment sessions to correctly shape the desired behaviour. On the
additional rare occasions (five sessions) where the suspected networking difficulties that had
149
disrupted the audio quality also resulted in even poorer viewing during the sessions, the SLPs were
able to resolve these problems by disconnecting and re-establishing the videoconferencing
connection between the two ends of the telerehabilitation system. In general, the aforementioned
challenges relating to the video quality online did not significantly impact on treatment outcomes,
as comparable treatment gains were achieved with online as face-to-face LSVT®. Overall, the
video quality was not seen as a major limitation to online treatment delivery and the adequate audio
quality, structured nature of the LSVT® and emphasis on verbal communication greatly
compensated for any challenges relating to this less-than-optimal quality.
4.4.3.3 System Operation and Speech-Language Pathologist Challenges
The features of the telerehabilitation system including the ability to sample real-time
calibrated average recordings of SPL, F0 and duration data, and display therapy materials online,
incorporated the essential information necessary for the delivery of the LSVT®. This objective
information allowed the SLPs to appropriately monitor task performances and provide timely
feedback to participants. However, there were more steps involved in obtaining this information
online compared to face-to-face delivery. For instance, during the sustained phonation task, the
face-to-face SLPs were able to: (1) record the average SPLs as they were automatically displayed
on the Digital SLM; (2) attend to the participants where necessary to observe their task
performances (e.g. vocal quality, use of open mouth position, general posture and presence of
muscle tension); and (3) maintain appropriate eye contact with the participants. For this same task,
the SLPs in the online environment needed to: (1) perform the additional step of sampling the mean
SPLs before recording the values; (2) attempt to monitor task performances either in real-time with
videoconferencing or via the store-and-forward feature for later review; and (3) give the impression
of maintaining eye contact with the participants by looking directly at the web camera on top of the
monitor rather than at the participants on the PC screen. These additional steps in online delivery
had the potential to disrupt the flow of treatment as the SLPs manipulated the telerehabilitation
system. However, the online application in the present study was easy to operate and the SLP
training in using the telerehabilitation system, together with the prescriptive nature of the LSVT®
enabled quick uptake of skills and proficiency. As a result, the components of the treatment were
delivered online and within the necessary one-hour time frame and comparable outcomes to face-to-
face delivery were achieved. This further demonstrated that the SLPs were able to focus their
attention appropriately on the LSVT® goals of training loud phonation and calibration. In future,
150
the use of systems with higher bandwidth would further assist SLPs during the online delivery of
the LSVT®, as clearer viewing of participants with videoconferencing would reduce the need for the
additional store-and-forward review. Furthermore, the repositioning of the web camera closer to
the participant image (Montani, et al., 1997), would create a more natural environment where
similar to the face-to-face setting, the online SLPs could maintain eye contact at the same time as
monitoring participants‟ performance via the telerehabilitation system. Moreover, it would also be
useful if future telerehabilitation systems could enable audio playback of task performances to
participants during the online sessions, which would further aid calibration. This additional
capability of the telerehabilitation system would ensure that online LSVT® further resembles face-
to-face treatment.
Finally, it is acknowledged that during the course of the study, the telerehabilitation system
malfunctioned on a particular day, and this event presented further challenges for the treating SLPs
in operating the system. On this occasion, the telerehabilitation system‟s acoustic speech processor
malfunctioned and the SPL, F0 and duration data could not be obtained. This hardware difficulty
occurred on the day that four participants had their scheduled treatment sessions. However, as all
other features of the telerehabilitation system including videoconferencing were unaffected, the
online sessions were able to continue as per normal, with minor inconvenience. This was largely
due to the fact that each SLP was still able to monitor the participants‟ vocal loudness and quality
auditorily, and that at this stage of treatment (second or third week of the programme), calibration
of the loudness level was already underway for each participant. Overall, as the aforementioned
technical difficulties occurred very rarely and were subsequently resolved with no further impact on
sessions, they were not considered significant factors limiting online treatment delivery.
4.5 Study Limitations and Future Directions
The present study had a number of limitations. Firstly, due to the laboratory nature of the
investigation, the findings are difficult to generalise to the wider clinical setting. As outlined
previously (section 3.4.4), future large-scale studies should be conducted remotely at either the
patient‟s home or within a community health centre setting (home telecare), in order to fully
determine the benefits and challenges of online treatment and the likelihood of its uptake as a mode
of service delivery for PD.
151
Secondly, the principal investigator was one of the assessing and treating SLPs and
therefore not blind to the study. This was an unavoidable factor due to the shortage of SLPs who
were available to take part in the study and who were certified in the LSVT®. However, to
minimise bias, no SLP that had treated a participant also assessed them post-treatment. Further,
possible bias arising from the involvement of the SLPs was minimal, as the LSVT® was
prescriptive, the assessments were administered using standard instructions, and the assessment data
were either recorded using objective measures or rated by independent and blind assessors.
A third limitation was the combination of the SLP shortage and the laboratory design that
allowed for some face-to-face interaction between the SLPs and participants receiving online
LSVT® at the beginning and/or end of the sessions. For those participants with mobility
difficulties, the treating SLPs needed to assist them from the taxi or waiting room to the treatment
room. Further, with the current proof-of-concept design, the online SLPs were required to: (1)
physically turn on the PC and activate the online application at the participant site; (2) measure the
microphone distance for all participants at the beginning of each online sessions; and (3) verify the
mean SPLs against the Digital SLM. As a result, face-to-face contact and a level of interaction
were inevitable. Similar difficulties were also reported by Mashima and colleagues (2003) in their
online laboratory proof-of-concept voice treatment study, where some level of interaction was
inevitable between the treating SLPs and participants who were in adjacent rooms. However, in the
present study, all possible attempts were made to minimize the level of face-to-face contact and to
maintain the online treatment environment. Most importantly, the participants and SLPs were
aware of keeping the conversation to a minimum during the face-to-face contact, and discussions
relating to treatment or prompting of vocal loudness were discouraged. These practices were
similarly observed for the participants treated face-to-face. Such issues will be negated in the next
stage of research in which participants will be treated in their own homes (Chapter 6).
A fourth limitation related to the participant satisfaction questionnaire. As identified in the

assessment study (Chapter 3), the relevance of the questionnaire was mainly to the laboratory
setting and included only a small number of issues. The future investigation of participant as well
as SLP satisfaction with online treatment should be comprehensive and also relevant to the real-
world remote setting. An in-depth cost benefit analysis (which was beyond the scope of this study)
relating to remote online LSVT® delivery for the participant and healthcare provider would also be
pertinent.
152
Finally, the present study concentrated only on participants with IPD and mild to moderate
hypokinetic dysarthria. Further studies are needed to investigate the feasibility of online LSVT ® for
participants with more complex PD, in order to identify any challenges to treatment that may exist
for this population. Together, this research would outline the full benefits of online LSVT®
delivery as an alternative or additional mode of service delivery for people with PD.
4.6 Conclusion
This study confirmed the validity of online laboratory-based LSVT® for participants with
IPD and mild to moderate hypokinetic dysarthria. The results supported the hypothesis that
participants treated in the online environment would achieve non-inferior treatment outcomes on the
acoustic and perceptual measures to those participants treated in the face-to-face setting, and the
efficacy trials in the literature. On the majority of the parameters, the treatment outcomes for the
participants in both LSVT® environments were also statistically and clinically significant.
Although the online modality presented some unique challenges for treatment delivery, the
participants and SLPs were able to quickly adapt to this modality, and a high level of participant
satisfaction was achieved overall. The PC-based telerehabilitation system described in this study
provides a basis for the delivery of online LSVT® in the clinical setting, whereby lessening the
access barriers that currently exist for people with PD.
153
154
Chapter 5
Online LSVT® for Complex Cases of PD: Case Studies
5.1 Introduction
Following on from the successful online LSVT® delivery for participants with IPD and
mild to moderate hypokinetic dysarthria as demonstrated in the previous chapter, it was considered
important that the next stage of research in this thesis investigate the feasibility of online LSVT® for
more complex cases of PD. Individuals with IPD at more advanced stages of the disease and those
with PSP face greater physical, communication and cognitive difficulties than the classic IPD
patients with mild to moderate hypokinetic dysarthria reported previously. Consequently, these
multiple and complex participant factors further complicate the ability for individuals to access
speech-language pathology services (detailed in section 1.6).
For these complex cases, the motor symptoms associated with PD are typically more
pronounced and individuals may experience increased muscle rigidity, hypokinesia, bradykinesia,
akinesia, tremor, postural and balance disturbances (please refer to section 1.3.1 for a detailed
review of these symptoms). The former three features are also generally prominent across the
speech subsystems, resulting in greater severity of hypokinetic dysarthria that is often more resistant
to treatment change (Colton, et al., 2006; Duffy, 2000; Ramig, et al., 1994). Together with the
additional cognitive difficulties that are often apparent for individuals with IPD at advanced stages
of the disease, and those with PSP, it is often more difficult for these individuals to achieve marked
success with the LSVT® and to maintain long-term treatment gains (Adams, 1997; Countryman &
Ramig, 1993; Countryman, et al., 1994; Lewis, et al., 2003; Owen, 2004; Theodoros, et al., 1999;
Ward, et al., 2000). It has been suggested that for these complex cases of PD, realistic expectations
following the LSVT® may include improved functional communication to the level where the
patient is intelligible most of the time or at the single word level and with the improved ability to
communicate daily needs. These goals differ to those expected for IPD patients with milder
hypokinetic dysarthria where near-normal speech may be a realistic probability (Ramig, Pawlas, et
al., 1995).
155
An example of the more variable performance with the LSVT® for PSP has been
demonstrated in a single case study of face-to-face treatment (Countryman, et al., 1994). In the
study, the participant presented with moderate to severe hypokinetic dysarthria at stage IV on the
Hoehn and Yahr Scale, and with significant physical weakness requiring her to be wheelchair-
bound. Following the LSVT®, the participant and family rated perceptual improvements in
loudness (34% and 37% respectively), monotonicity (76% and 89%), slurring (62% and 93%) and
overall speech intelligibility (55% and 92%). Substantial treatment gains were also achieved
acoustically for mean SPL (7.55 dB increase in maximum sustained vowel phonation; 7.38 dB
reading; 5.53 dB monologue). However, the post-treatment levels were markedly lower than those
reported in the LSVT® efficacy studies for IPD participants with mild to moderate dysarthria
(Ramig, et al., 1994; Ramig, Countryman, et al., 1995). It was suggested in the study that the
dysarthria severity and disease progression largely impacted on treatment performance
(Countryman, et al., 1994). Overall, it was encouraging to note that the participant benefitted from
improved functional communication skills, which provided support for the use of the LSVT® as an
appropriate treatment option in this case.
In addition to the aforementioned factors of greater physical, communication and cognitive

difficulties, for individuals with IPD at more advanced stages of the disease that have undergone
neurosurgical intervention for the management of their motor symptoms (surgical IPD), the effects
of neurosurgery can further impact on the LSVT® performance. Such findings have been primarily
reported in three studies in the literature. In one of the studies, Ward and colleagues (2000) found
that their surgical IPD group (n = 12; Hoehn and Yahr Stages I-II = 2 participants; Stages III-IV
= 10 participants) with pallidotomy and/or thalamotomy and mild to severe hypokinetic dysarthria
made significant progress with the LSVT® in sentence intelligibility (9% improvement as measured
by a SLP) and mean SPL (12.7 dB increase in maximum sustained vowel phonation; 13.3 dB in
reading). However, these gains were overall lower than those of the non-surgical IPD participants
in the same study with similar levels of dysarthria severity (n = 18; Hoehn and Yahr Stages I-II = 10
participants; Stages III-IV = 8 participants). Lower performances of the surgical IPD participants
were particularly evident on the above measures of mean SPL during reading (3.7 dB lower) and
sentence intelligibility (10.7% lower). The surgical IPD group also failed to show significant gains
in tongue strength and endurance in line with the non-surgical group.
156
In a case study by Countryman and Ramig (1993), the LSVT® performance was reported
for a participant who had undergone bilateral thalamotomy and presented at Stage III on the Hoehn
and Yahr Scale, with mixed hypokinetic-spastic dysarthria of moderate impairment. Post-treatment
gains were demonstrated both perceptually (increase of 54.66% in loudness level; increase of
45.69% in articulatory precision; decrease of 28.81% in monotonicity as rated by the participant),
and acoustically by the participant (mean increases of 5.91 dB in maximum sustained vowel
phonation; 3.49 dB in reading; 2.55 s in maximum duration of sustained vowel phonation)
(Countryman & Ramig, 1993). However, as per Countryman et al. (1994) and Ward et al. (2000),
the treatment gains were generally lower than those reported in the LSVT® efficacy studies for IPD
participants with mild to moderate hypokinetic dysarthria (Ramig, et al., 1994; Ramig, Countryman,
et al., 1995).
A similar trend to the above studies was also observed in another case study where a
participant with severe hypokinetic dysarthria at Stage IV on the Hoehn and Yahr Scale who had
undergone thalamotomy and pallidotomy made some gains post-LSVT® (Theodoros, et al., 1999).
However, these overall gains were lower than in the efficacy studies. Specifically for this
participant, improvements were only perceived in speech intelligibility (increases of 22.5 % in word
intelligibility; 101.8% in sentence intelligibility as rated by a SLP), and on a number of acoustic
parameters (mean increases of 22.3 dB in maximum sustained vowel phonation; 13 dB in reading;
15.9 s in maximum duration of sustained vowel phonation).
It has been suggested that the effects of neurosurgery may have a direct negative impact on
the degree of improvement that is possible with the LSVT® for surgical IPD patients (Countryman
& Ramig, 1993; Theodoros, et al., 1999; Ward, et al., 2000). Due to the vast neural connectivity
between the cerebral cortex and basal ganglia (Figure 1.2) it is likely that the direct interference to
the basal ganglia control circuit with neurosurgery may have a wide and negative influence on
aspects of motor speech function, language, cognition and behaviour (Heo et al., 2008; Murdoch,
2001; Okun et al., 2009; Zangaglia et al., 2009). Clinically, for surgical IPD patients, these
influences are likely to appear as difficulties: (1) achieving the desired loudness levels; (2) self-
monitoring speech production; (3) maintaining the louder phonations long-term; as well as
(4) generalising the global effects of loud phonation throughout the speech subsystems, in line with
non-surgical IPD cases (Countryman & Ramig, 1993; Theodoros, et al., 1999).
157
Overall, despite the limited number of investigations into the feasibility of face-to-face
LSVT® for complex cases of PD and the variable treatment gains, clinically, the LSVT® continues
to be recommended as an appropriate intervention for these individuals (Ramig, Pawlas, et al.,
1995). Additionally, as some of the barriers to treatment access (e.g. increased physical difficulties
impacting on capacity to travel) may be even more significant for such individuals, online LSVT®
delivery may prove to be a logical and necessary service delivery option. This modality may aid
access to treatment and in turn, allow for improved and/or maintained functional communication
and quality of life. However, due to the paucity of telerehabilitation investigations in this area, it is
unclear whether a telerehabilitation approach to treatment is feasible for complex cases of PD. In
particular, it is unknown whether the additional participant factors such as greater communication,
physical and cognitive difficulties and the effects of neurosurgery encountered by these individuals
would present unique challenges to online treatment. The intent of the current study was to
investigate these issues for a small number of complex cases of PD. Specifically, the study aimed
to investigate the feasibility of the LSVT® delivered online in a laboratory setting using the PC-
based telerehabilitation system for a series of case studies of individuals with complex PD. It was
hypothesised that despite the complex issues, online LSVT® would be feasible for these cases and
result in functional gains.
5.2 Method
5.2.1 Participants
and Social Sciences Ethical Review Committee of The University of Queensland. Four participants
with complex PD took part in the laboratory treatment pilot study at The University of Queensland.
The participants were recruited from the assessment study (Chapter 3) having met the inclusion
criteria for treatment as outlined in Chapter 4 (section 4.2.3), in addition to having undergone
neurosurgical intervention for IPD or having been diagnosed with PSP. Furthermore, the
participants were considered complex cases of PD and suitable for inclusion in the study as they
demonstrated marked impairment of a number of factors including physical, communication and
cognitive function that were impacting on daily activities. The participants did not display any
aspects described previously (section 4.2.3) that warranted exclusion. The participants recruited to
the study had not taken part in the randomized controlled treatment trial in Chapter 4. An additional
158
six complex cases of PD has been identified from the assessment study (Chapter 3) as eligible for
inclusion in this study. However, these individuals had declined involvement in the study as certain
access barriers including the travel distance to the university, their physical incapacity and the
intensive nature and commitment of the LSVT® programme prohibited their participation.
Before commencement of the study, an overall hypokinetic dysarthria level (mild to

severe) was determined for each participant by the principal investigator (Table 5.1). This was
based on their pre-treatment assessment results (Chapter 3) of mean SPL and perceptual ratings of
OIC from the monologue task (see section 4.2.3 for further description of the dysarthria
categorisation). The severity of the prominent disordered speech and voice characteristics and their
impact on overall communication were also taken into account when determining the dysarthria
level. The descriptive characteristics of the four participants included in the study are displayed in
Table 5.1. For consistency between the studies in this thesis, the original participant number
allocation as per the assessment study was maintained for each participant. The first three
participants in this study had IPD at advanced stages and had undergone neurosurgical intervention.
Specifically, Participant 55 presented at Stage III-IV on the Hoehn and Yahr Scale with moderate-
severe hypokinetic dysarthria, a moderate degree of cognitive, postural and balance difficulties and
had undergone a pallidotomy. Participant 49 presented at Stage III on the Hoehn and Yahr Scale
with moderate hypokinetic dysarthria, some mild postural difficulties and had undergone DBS.
Participant 48 had also undergone DBS and presented at Stage IV on the Hoehn and Yahr Scale
with moderate hypokinetic dysarthria, moderate postural and balance difficulties, and mild
cognitive difficulties. The fourth participant in this study, Participant 54, had been diagnosed with
PSP and was at stage II on the Hoehn and Yahr Scale. Participant 54 presented with moderate
hypokinetic dysarthria, mild balance difficulties, and mild fluctuating neck pain and tightness.
Videolaryngoscopic examinations conducted by an Ear Nose and Throat specialist prior to
treatment were unremarkable for all participants.
159
Table 5.1 Descriptive Characteristics of Participants
Participant Age Sex PD type Time post Neurosurgery Time post Hoehn & Dysarthria Prominent characteristics
(years) diagnosis neurosurgery Yahr severity
(years) Stage level
55 75 M IPD 16 Pallidotomy 7 3.5 Moderate- +++Reduced speech

Severe intelligibility; ++Imprecise
Unilateral
articulation; ++Roughness;
++Monopitch; ++Cognitive
difficulties; ++Postural
difficulties and balance;
+Reduced loudness;
+Breathiness; +Fast rate of
speech
49 58 M IPD 2 DBS 1 3 Moderate +++Fatigue due to wearing-off

effects of medication;
Bilateral
+++Breathiness; ++Reduced
loudness; ++Roughness;
++Imprecise articulation;
++Monopitch; ++Fast rate of
speech; +Reduced speech
160
(years) Stage level
intelligibility; +Postural
difficulties; +Word repetitions
48 76 F IPD 14 DBS 3 4 Moderate ++Roughness; ++Breathiness;

++Reduced loudness;
Unilateral
++Monopitch; ++Postural
difficulties and balance;
++Fatigue due to wearing-off
effects of medication; +Labile;
+Cognitive difficulties
54 53 F PSP 1 __ __ 2 Moderate ++Roughness;

++Monoloudness;
++Monopitch; ++Imprecise
articulation; +Reduced loudness;
+Breathiness; +Fatigue due to
wearing-off effects of
161
(years) Stage level
medication; +Labile; +Balance

difficulties; +Fluctuating neck
pain and tightness; +Some
reduced lip and tongue
movement
Note. M = male; F = female; IPD = idiopathic Parkinson‟s disease; PSP = progressive supranuclear palsy; DBS = deep brain stimulation; Dashes
correspond to non-applicable information; + corresponds to mild impairment; ++ corresponds to moderate impairment; +++ corresponds to severe
impairment.
162
5.2.2 Procedure
One of the four LSVT® certified SLPs who participated in the previous laboratory
treatment study (Chapter 4) delivered the online treatment to all participants. The sessions were
conducted within the Telerehabilitation Research Unit at The University of Queensland and the
LSVT® was delivered according to standard clinical practice (Ramig, et al., 1994; Ramig,
Countryman, et al., 1995; Ramig & Fox, 2004; Ramig, Pawlas, et al., 1995) as detailed in section
4.2.5. The set-up and delivery of the online sessions using the telerehabilitation system were also
consistent with the online procedure of the previous study (please refer to section 4.2.6.1).
In keeping with the protocol in Chapter 4 (see section 4.2.7), the participants were re-
assessed within 48 hours of their final LSVT® session at The University of Queensland. The
participants were assessed face-to-face and the SLPs who had administered the pre-treatment
assessments (Chapter 3) took part. Only one SLP was required for each assessment and the treating
SLP was not assigned to the assessments. All speech and voice perceptual ratings were made by the
two additional SLPs that had not taken part in any of the assessment or treatment studies and who
were independent and blind to the intent of all studies in this thesis (see section 4.2.7.2).
Consistent with the previous treatment study (section 4.2.7), the assessments included
acoustic measures of: (1) mean SPLs of the participant‟s speech during six maximum sustained
vowel phonations of /a/, reading of The Grandfather Passage and a 30 s monologue; (2) maximum
duration of six sustained vowel phonations (s); and (3) maximum F0 range (ST) from a series of six
vocal glides where the average of the highest and lowest F0 levels (Hz) were then converted to a
maximum range in semitones (ST) (de Pijper, 2007). As outlined previously (section 4.2.7.1), the
telerehabilitation system was used to obtain all acoustic measures. The perceptual measures
included: (1) DME ratings of voice parameters (vocal breathiness and roughness, loudness level,
loudness variability and pitch variability), OAP and OIC (where a lower DME value for the vocal
breathiness and roughness parameters only denoted improved performance); and (2) word and
sentence intelligibility and communication efficiency on the ASSIDS (section 4.2.7.2). The
participants also completed a questionnaire relating to their level of satisfaction with the online
163
treatment sessions, audio and video quality during the sessions and overall satisfaction with online
treatment (section 4.2.7.3). Further to the assessment outcomes in Chapter 4, the participants were
invited to comment on any aspects relating to their online management and their speech and voice
features. The SLP in this study was also invited to comment on any aspects relating to the online
treatment of complex cases of PD.
5.3 Results and Discussion
The following sections discuss the study findings for the four complex cases. As statistical
analyses could not be performed due to the case study design, descriptive comparisons were made
for each participant on the pre- and post-LSVT® values and degree of treatment change on the
acoustic and perceptual measures. Furthermore, in the absence of large-scale studies of online and
face-to-face LSVT® performance for more complex cases of PD, the success of treatment in this
study was difficult to quantify. Consequently, a number of criteria from previous chapters were
used in this study to provide some indication of treatment success. These measures included: (1)
minimum improvement levels on the acoustic parameters following face-to-face LSVT® as reported
in the efficacy studies for IPD participants with mild to moderate hypokinetic dysarthria (13 dB
change for maximum sustained vowel phonation; 8.03 dB for reading loudness; 4.5 dB for
monologue loudness; 4 ST for maximum F0 range; 3.72 s for maximum duration of sustained vowel
phonation) (Ramig, et al., 1994; 1996; Ramig, Countryman, et al., 1995); (2) percent improvement
on the word and sentence intelligibility tasks of the ASSIDS above the test-retest variability for
dysarthric speakers (above 3.2% for word intelligibility; 8.6% for sentence intelligibility) (Yorkston
& Beukelman, 1981a); (3) improvement in communication efficiency on the ASSIDS above 0.27
(Hill, et al., 2006); and (4) improvement on the DME parameters above the level of inter-rater
variability observed overall on the parameters on the pre-treatment ratings in Chapter 4 (MAD =
21.34, SD = 18.16). The latter criterion was calculated specifically for this study.
5.3.1 Participant 55
Participant 55 was a retired gentleman who resided in a care facility as a result of his
physical difficulties relating to PD. He presented with moderate-severe hypokinetic dysarthria. His
primary speech and voice concerns were difficulties being understood and “getting words out”,
164
which had significantly limited his communication interactions with family members, residents and
staff. He would often keep to himself throughout the day as it was becoming too difficult to make
himself understood. Participant 55‟s prominent speech and voice difficulties included severely
reduced speech intelligibility, moderately impaired articulatory precision, pitch variability and a
moderately rough vocal quality. The participant also demonstrated a mild reduction in vocal
loudness, a mildly breathy vocal quality and fast rate of speech (Table 5.1).
5.3.1.1 Treatment Outcome
The participant made negligible improvements on the majority of the acoustic parameters
including mean SPL (maximum sustained vowel phonation, reading, and monologue) and
maximum duration of sustained vowel phonation (Table 5.2). Only on the maximum F0 range task
(increase of 5.29 ST) was there a substantial improvement achieved in line with the LSVT® efficacy
studies for nonsurgical IPD participants with mild to moderate hypokinetic dysarthria (Ramig, et al.,
1994). On the perceptual parameters, negligible improvements were evident which largely
supported the minimal changes noted acoustically (Table 5.3). At post-treatment, a number of the
perceptual parameters (vocal breathiness, loudness level, pitch variability, OAP, OIC and word
intelligibility) were somewhat lower than pre-treatment which was unexpected.
165
Participant 55
Task Pre-LSVT Post-LSVT Change from baseline
Maximum sustained 77.93 79.91 1.98

vowel phonation (dB)
Reading (dB) 68.88 71.23 2.35
Monologue (dB) 67.82 70.89 3.07
Maximum duration 5.83 6.00 0.17

of sustained vowel
phonation (s)
Maximum F0 range (ST) 1.11 6.39 5.29
Note. = higher or lower post-treatment value.
166
Participant 55
Breathinessa 64.57 77.62 13.05
Roughnessa 89.13 75.86 13.27
Loudness levela 186.21 181.97 4.24
Loudness variabilitya 138.04 144.54 6.50
Pitch variabilitya 95.50 85.11 10.39
OAPa 95.50 77.62 17.88
OICa 75.86 54.95 20.91
WI (%) 81.00 78.00 3.00
SI (%) 86.68 90.23 3.54
CER 0.42 0.63 0.22
Note. = higher or lower post-treatment value; OAP = overall articulatory precision; OIC =
overall speech intelligibility in conversation. aParameters measured using Direct Magnitude
Estimation; WI = word intelligibility; SI = sentence intelligibility; CER = communication efficiency
ratio.
In general, the lower performance on both the acoustic and perceptual parameters and lack
of generalisation of the global effects of loud phonation throughout the speech subsystems may
reflect the participant factors such as greater motor impairment, dysarthria severity and cognitive
difficulties associated with advanced PD for this individual, and the effects of neurosurgical
intervention. Together, such factors may have made it more difficult for Participant 55 to achieve
success with treatment comparable to the nonsurgical IPD groups reported previously, and in
particular, calibration to the louder level. The postural difficulties experienced by Participant 55
impacted on treatment as the participant found it difficult to maintain a consistent upright sitting
position throughout each session and to correct his posture when prompted by the SLP. This factor
167
affected the participant‟s ability to achieve consistent and adequate breath support and control for
the treatment tasks. Participant 55 also required greater assistance by the SLP during the first few
weeks of the LSVT® delivery to achieve the target loudness level and vocal quality. More frequent
prompting by the SLP was also required each session to then maintain the target levels and assist
with carryover, as Participant 55 found it challenging to monitor and maintain his loudness level
independently. This amount of prompting was considerably greater than that typically required for
the nonsurgical IPD participants treated in the previous study (Chapter 4) in either environments
(online or face-to-face), who were able to achieve calibration more easily. Difficulties in achieving
calibration have also been documented with face-to-face LSVT® for complex cases of PD
(Countryman & Ramig, 1993). This finding would suggest that problems with calibration appear to
occur independently of the treatment environment.
In addition to the participant factors mentioned above, rating difficulties due to the severity
of the hypokinetic dysarthria may have impacted on treatment outcome. It is possible that the
prominent features of articulatory imprecision, vocal roughness and breathiness, reduced pitch
variability and fast rate of speech that persisted with treatment, may have made it harder for the
independent raters to discriminate and attend to the specific parameters in question using DME
and/or to perceive demonstrable changes with treatment. This may possibly explain some of the
discrepancies observed perceptually, where despite the negligible gains, a number of parameters
were unexpectedly perceived as lower post-treatment.
5.3.1.2 Participant Perspectives Regarding Online Lee Silverman Voice Treatment®
All online treatment sessions were conducted successfully without any technical
difficulties. Online LSVT® was well received by Participant 55. On the participant satisfaction
questionnaire, he rated the audio and video quality during videoconferencing as adequate, felt
comfortable while participating in the sessions and found it easy to engage with the SLP. Overall,
Participant 55 was more than satisfied with online treatment and was now even planning to set up
his own computer so that he could talk to his grandchildren over the Internet. Despite the relatively
small quantifiable improvements on the acoustic and perceptual parameters, from a functional
perspective, the treatment made a notable difference to Participant 55‟s communication and
confidence. Participant 55 reported that his speech and voice had improved post-treatment and that
168
he now sounded louder, clearer, less slurred and his speech and voice did not deteriorate quite as
much throughout the day. Friends and family had also noticed an overall improvement in his
speech. Participant 55 stated:
Some people still think some words I can’t pronounce but other people say we can get the
gist of what you’re talking about now. I’m very pleased with it. My wife rang up at 9
o’clock last night and she thought it was very clear.
Participant 55 also reported improved confidence in his communication abilities which

translated to greater initiation of conversations and interactions with the other residents and staff at
the care facility, and speaking more frequently on the telephone with family and friends. His sense
of humour was also more evident with the increased confidence.
Participant 49 was a practicing priest who was finding it increasingly difficult to make
himself heard and understood by the congregation during Mass, and in his other duties. He had also
started to avoid speaking on the telephone as it was difficult to make himself heard. Participant 49
presented with moderate hypokinetic dysarthria. His speech and voice features included a severely
breathy vocal quality and moderately impaired loudness level, pitch variability, rough vocal quality,
articulatory precision and fast rate of speech. Participant 49 also presented with a mild reduction in
speech intelligibility and a mild increase in word repetitions (Table 5.1). Additionally, he reported
fatigue associated with fluctuations in the effects of his PD medication which resulted in even more
pronounced speech and voice difficulties.
Participant 49 achieved substantial improvements following treatment on all acoustic

parameters (Table 5.4). The treatment gains were in line with the minimal improvements reported
in the efficacy studies for participants with IPD and mild to moderate hypokinetic dysarthria
(Ramig, et al., 1994, 1996; Ramig, Countryman, et al., 1995). Furthermore, Participant 49‟s post-
169
treatment mean SPL values were generally within the phonation ranges of the nonsurgical IPD
participants with mild to moderate hypokinetic dysarthria described in Chapter 4, the face-to-face
LSVT® efficacy trials (Ramig, et al., 1996; Ramig, Countryman, et al., 1995), and two groups of
healthy older adults (Fox & Ramig, 1997; Ramig, Sapir, Fox, et al., 2001). These findings
suggested that for this individual, carryover of treatment effects with online LSVT® was achievable,
as was the return to near-normal loudness levels.
Table 5.4 Pre- and Post-LSVT® Values and Treatment Changes for the Acoustic Measures for
Participant 49

Reading (dB) 63.14 72.01 8.87
Monologue (dB) 62.96 69.26 6.30

of sustained vowel
phonation (s)
However, despite these acoustic improvements, Participant 49‟s performance on the

perceptual parameters was variable, with only the parameters of vocal breathiness, OAP, and word
and sentence intelligibility achieving measurable gains post-treatment (Table 5.5). Some of the
parameters including vocal roughness, loudness level and loudness variability were somewhat lower
post-treatment which was an unexpected outcome. When compared to the significant
improvements on almost all of the perceptual parameters for the nonsurgical IPD participants in
Chapter 4, the inconsistency in Participant 49‟s performance may reflect the greater severity of
hypokinetic dysarthria due to the more persistent speech and voice features. It is likely that a level
of carryover of loud phonation throughout the speech subsystems was more difficult to perceive.
170
Table 5.5 Pre- and Post-LSVT® Values and Treatment Changes for the Perceptual Measures for
Participant 49
Breathinessa 158.49 67.61 90.88
Roughnessa 70.79 83.18 12.39
OAPa 102.33 134.90 32.57
OICa 123.03 128.82 5.79
WI (%) 96.00 100 4.00
SI (%) 66.59 96.82 30.23
CER 0.84 0.98 0.14
ratio.
For Participant 49, all online LSVT® sessions were able to be completed without technical
difficulties. The participant was generally positive about the treatment modality. On the participant
satisfaction questionnaire, Participant 49 rated that he was very satisfied with the online sessions
and treatment overall, and found the audio quality to be excellent and the video quality adequate.
Participant 49 reported substantial changes with treatment in speech, loudness, overall intelligibility
and confidence. Specific improvements with treatment included the greater frequency in speaking
171
on the telephone, and an overall increase in the length of time spent talking each day. He also
stated:
I think my voice is louder. I can stay louder for longer and sometimes I just speak
naturally. People can hear me now. I used to whisper a lot, and now I don’t whisper
anymore. I get my message across. I have had quite a lot of feedback from parishioners.
They are very happy going to the Mass now when I’m by myself. They all feel that my
voice is stronger and clearer.
Participant 48 was a retired lady who was becoming increasingly dependent on her
husband to perform activities of daily living due to her physical difficulties. Her primary speech
and voice concerns were not being understood on the telephone and in social settings in background
noise. These difficulties had reduced her communication interactions. She stated, “I can no longer
chat so easily and I often have to think whether it‟s important to say something or skip it”.
Participant 48 presented with moderate hypokinetic dysarthria characterized by a moderately
breathy and rough vocal quality and moderately reduced loudness and pitch variability (Table 5.1).
Her speech and voice difficulties would also become more pronounced with fluctuations in the
effects of her PD medication and fatigue.
Participant 48 showed a substantial improvement on the acoustic parameter of maximum

sustained vowel phonation only, while negligible treatment gains on the remaining acoustic and
perceptual parameters (Tables 5.6 and 5.7). As similarly noted for the previous two participants,
the post-treatment values on a number of these parameters were somewhat lower than pre-treatment
for Participant 48, which was unusual. It is likely that the participant factors including the more
advanced stage of PD, cognitive difficulties, effects of medication and fatigue and effects of
neurosurgery, impacted on the participant‟s ability to achieve calibration on the acoustic measures.
Calibration difficulties were evident during treatment for this participant and included problems
monitoring and maintaining the target SPL and duration levels independently between treatment
172
tasks and sessions. Despite frequent prompting and demonstrations by the SLP throughout
treatment, the participant displayed substantial difficulties with increasing her range of vocal glides
for the maximum F0 range task. It is also likely that the greater dysarthria severity and persistent
features may have added to rater difficulties in perceiving substantial post-treatment improvements
across the perceptual measures. Overall, however, it was encouraging noting clinically that the
quality of Participant 48‟s post-treatment phonations on the maximum sustained vowel phonation
tasks and vocal glides had improved, suggesting a level of carryover to some extent on these tasks
with treatment.
Participant 48

Reading (dB) 66.44 68.08 1.64
Monologue (dB) 65.20 67.60 2.40

of sustained vowel
phonation (s)
173
Participant 48
Roughnessa 79.43 77.62 1.81
OAPa 128.82 125.89 2.93
OICa 120.23 125.89 5.66
WI (%) 94.00 95.00 1.00
SI (%) 98.18 94.32 3.86
CER 0.91 0.80 0.11
ratio.
Participant 48 was also able to complete the 16 online sessions and no technical difficulties
occurred during her treatment. As rated on the participant satisfaction questionnaire, Participant 48
was very satisfied with the online sessions, rated the video and audio quality as excellent and was
overall, more than satisfied with the online treatment. The participant further commented that she
enjoyed all sessions and looked forward to them each day. Despite the negligible gains on the
majority of the acoustic and perceptual parameters, Participant 48 did report post-treatment
improvements in overall communication and in her speaking confidence. The participant was
174
finding it easier to chat to family and friends and would more often engage in conversation post-
treatment. She stated:
The treatment has made a difference. I can manage to make myself heard and understood.
My voice is stronger, louder, and I can talk with greater ease. I’m talking to more people
with greater confidence that they can understand me, and I do talk on the telephone and
that’s quite a lot better.
Participant 54 was the manager of her own company. She had begun to notice some
difficulties in being understood on the telephone, during work seminars, as well as in a one-to one-
setting. Participant 54 has been diagnosed with PSP and presented with moderate hypokinetic
dysarthria. Her prominent speech and voice difficulties included a moderately rough vocal quality,
and moderately reduced pitch and loudness variability and articulatory precision. Participant 54
also demonstrated a mild reduction in her loudness level and a mildly breathy vocal quality.
Fluctuating levels of pain and tightness in the neck region (associated with vocal roughness and
breathiness) were also reported, as were some difficulties with lip and tongue movements during
speech. The wearing-off effects of the PD medication and fatigue were also noted clinically to
reduce Participant 54‟s speech intelligibility to some extent.
On the acoustic parameters, Participant 54 showed substantial improvements with

treatment in mean SPL for the maximum sustained vowel phonation, reading and monologue tasks
(Table 5.8). The post-treatment values for maximum sustained vowel phonation and monologue
were also within the ranges reported for the nonsurgical IPD participants in Chapter 4, the face-to-
face feasibility studies (Ramig, et al., 1996; Ramig, Countryman, et al., 1995) and the healthy older
adult groups (Fox & Ramig, 1997; Ramig, Sapir, Fox, et al., 2001). These finding suggested a
near-normal return in loudness level with treatment on these parameters. However, for the
maximum duration of sustained vowel phonation and maximum F0 range parameters, lower post-
treatment performances and negligible changes were noted respectively. It is likely that the
175
fluctuating muscle tightness and pain experienced in the neck region that occurred each session (and
in the post-treatment assessment), directly impacted on the participant‟s treatment success in these
areas. The symptoms would consistently result in perceived intermittent vocal breathiness and
roughness on all tasks during treatment. Therefore, to help improve the quality of the phonations as
part of calibration, the participant was encouraged to self-monitor the quality of the sustained
phonations and vocal glides during the sessions and cease the trials when the quality diminished.
Improved vocal quality on these tasks post-treatment was noted clinically, despite the lack of
measurable treatment change.
Participant 54

Reading (dB) 67.18 72.66 5.48
Monologue (dB) 65.48 72.10 6.62

of sustained vowel
phonation (s)
On the perceptual parameters, substantial gains with treatment were noted on the
parameters of loudness level, loudness variability, pitch variability and OAP (Table 5.9). It is
possible that some carryover of the effects of loud phonation were perceived for this individual.
However, for the remaining parameters, negligible treatment gains were observed. Similarly to the
previous participants, it is likely that the combination of participant factors including greater
dysarthria severity, physical symptoms and uniquely, the effects of the PSP condition may have
contributed to the more variable treatment outcome for Participant 54.
176
Participant 54
Roughnessa 102.33 83.18 19.15
OAPa 117.49 141.25 23.76
OICa 131.83 128.82 3.01
WI 96.00 98.00 2.00
SI 99.28 99.55 0.27
CER 0.91 0.85 0.06
ratio.
5.3.4.2 Participant Perspectives Regarding Online LSVT®
All 16 online treatment sessions were also successfully conducted for this participant,
without any technical difficulties encountered during treatment. On the participant satisfaction
questionnaire, Participant 54 was very positive about online LSVT®. The participant rated the
audio and video quality as more than adequate and stated that she would prefer online sessions to
face-to-face for the future management of her condition. Overall, the participant was very satisfied
with online treatment. Despite the variable treatment gains, Participant 54 had noticed
improvements in her speaking voice and confidence with treatment. She reported:
177
My voice is absolutely much louder and clearer. I found that if I speak more loudly,
particularly my husband has noticed that it is much clearer. He can understand what I’m
saying. He thinks it’s marvellous. When I’m speaking one-on-one with people I
concentrate more on how I’m saying things and that I am louder. I feel a lot more
comfortable now with my speaking and when I’m talking one-on-one. I know that when I
think more loudly, it seems to go more clearly. Before, people used to say to me “Oh, the
phone was breaking up,” or “Can you repeat that, I can’t quite hear what you said. We
must have a bad line”. I don’t seem to get that as much anymore which is really good.
5.3.5 Speech-Language Pathologist Challenges with Online Lee Silverman Voice Treatment®
Delivery
The results obtained in this study on the acoustic and perceptual parameters largely
suggested that online LSVT® provided variable benefits for these four individuals with complex
PD. The variability in treatment outcomes most likely reflected the participant factors including the
severity of hypokinetic dysarthria, physical and cognitive difficulties and/or the effects of
neurosurgery. From the perspective of the treatment delivery, all treatments were conducted
successfully for the participants, with no failed sessions or technical difficulties and with adequate
audio and video quality. The overall high levels of participant satisfaction with the online modality
were also very encouraging. However, a number of unique challenges did arise with online
delivery, particularly from the SLP‟s perspective. These are discussed below.
5.3.5.1 Participant factors
It became apparent in this study that certain participant factors did make it more difficult at
times to deliver the treatment sessions in the online environment. The most notable of these were
the cognitive difficulties associated with more advanced PD and effects of neurosurgical
intervention. Although not formally assessed, it was evident during treatment that such features
contributed to the difficulties in achieving the target loudness levels during the sessions and overall
calibration for the participants. These were particularly evident for Participants 55 and 48. As a
result of these difficulties, the SLP was required to take more of the treatment onus and provide
frequent prompting each session in an attempt to achieve and maintain the louder levels. As noted
178
previously, this amount of prompting was far greater than that required for the nonsurgical IPD
participants in Chapter 4 who received online and face-to-face LSVT®. It was also necessary to
continue with the prompting for a longer period of the treatment duration than usual (up to the third
week of treatment compared to the first week which is typical), in order to overcome some of the
calibration difficulties. Although the participants generally responded appropriately to the SLP
prompting in the short-term within the sessions, the post-treatment results particularly for
Participants 55 and 48 showed limited overall gains, suggesting that the participant factors may
have directly impacted on the level of treatment success for these individuals. It has been reported
in the literature that cognitive difficulties with more advanced PD generally manifest as deficits in
executive functioning, particularly planning and working memory (Lewis, et al., 2003; Owen,
2004). A similar trend in cognitive difficulties has also been observed for patients following
neurosurgical intervention alongside possible language (e.g. verbal fluency), motor speech and
behavioural difficulties with variable long-term effects (Heo, et al., 2008; Murdoch, 2001; Okun, et
al., 2009; Zangaglia, et al., 2009). When considering the possible impact of such factors on
treatment, it is highly likely that they directly contributed to the notable difficulties in achieving
task performance and calibration observed in the present study. However, from clinical experience,
similar difficulties have also been noted for complex cases of PD treated face-to-face, suggesting
that such difficulties occur independently of the treatment environment. Therefore, the same
considerations as face-to-face would need to be made for online treatment when patients present
with such difficulties at treatment screening. Among these considerations would be whether the
LSVT® is suitable for the needs of the patients and if so, that treatment expectations are realistic.
The physical difficulties experienced by the participants did make treatment delivery more
challenging at times. For Participant 55 in particular, it was generally difficult for him to maintain
an upright sitting position during the sessions, despite frequent prompting by the SLP. As discussed
previously, the postural difficulties impacted on the underlying level of breath support and control
possible for the treatment tasks. Again, such difficulties have been noted to occur independently of
the treatment environment. However, a useful future consideration for the online treatment of
patients with obvious physical difficulties would be to have a family or volunteer present at the
participant site to physically assist them where necessary, in place of the online SLP.
Lastly, the level of communication difficulties experienced by the participants did not
appear to impact on the online treatment delivery any more so than would be anticipated face-to-
179
face. This factor was therefore not considered a challenge to online delivery for the SLP with these
four cases. Overall, the participant factors mentioned here are generally present concomitantly for
complex cases of PD, with varying levels of severity and impact on treatment. It may therefore be
more important to consider the actual appropriateness of the LSVT® itself and realistic treatment
expectations for these individuals prior to the treatment environment (online or face-to-face). Such
considerations would best be made on a case by case basis.
As mentioned in Chapter 4 (section 4.4.3.1), the audio quality with videoconferencing was
adequate for treatment delivery in relation to providing timely directions, feedback and guidance
and for maintaining appropriate participant-SLP rapport. The occasional audio delays that were
encountered did not have a substantial impact on treatment delivery overall, as strategies to
minimise their effects were successfully implemented. However, in comparison with the previous
study in Chapter 4, a greater number of strategies needed to be implemented by the SLP during
online treatment for these complex cases of PD in the present study. These strategies were
necessary in order to: (1) ensure the same standard of delivery as face-to-face; (2) overcome the
effects of the participant factors mentioned above as well as fatigue that were impacting on
calibration; and (3) facilitate the flow of the sessions via appropriate communicative interactions
and turn-taking. Examples of the strategies employed by the SLP included the more frequent use
of: (1) shorter and more precise instructions; (2) demonstrations of the correct vocal behaviour; (3)
hand cues to assist with shaping, maintenance and quick feedback; as well as (4) waiting until the
participant had completed the task before responding. Although these strategies were easily and
successfully implemented by the SLP during the sessions, it was generally evident that the
participants in the present study with more severe hypokinetic dysarthria and greater cognitive
difficulties took longer to adapt to the use of the strategies. This was particularly noted for
Participants 55 and 48 who found the turn-taking behaviour more difficult in this environment.
These participants generally required at least an entire week of treatment training to adapt to the
online routine, as opposed to only the initial one to two sessions that were typically necessary for
the nonsurgical IPD participants in Chapter 4. However, from SLP experience, similar cases of
complex PD treated face-to-face also generally take longer to adapt to the LSVT® routine than
participants with milder difficulties. This may therefore indicate that such difficulties in treatment
delivery can occur regardless of the treatment environment. As a consideration, the future use of
180
applications with higher Internet bandwidth may further reduce any audio delays and the need for
SLPs to utilise a high number of compensatory strategies when working with this population.
As noted previously (section 4.4.3.2), the video quality with videoconferencing was
relatively poor during the sessions, which made viewing of the participant and SLP difficult. For
the SLP, the video quality in terms of the frame rate and pixelated image especially during
movement and the restricted viewing of the participants via the web cameras, presented additional
challenges to achieving calibration. For example, the reduced video quality made it more difficult
for the SLP to demonstrate the desired tasks to the participants and assist with the shaping of
phonations. It was also more difficult for the SLP to detect: (1) whether the phonations were
performed at the top of the breath with an open-mouth position; (2) head and neck muscle tension
(contraindicated for treatment); and (3) subtle changes to participant positioning and posture during
the sessions. The ability to detect such features was considered even more crucial for these more
complex cases in the present study compared to Chapter 4, as greater difficulties in these areas were
more likely to have a direct impact on treatment success. Furthermore, the SLP found it more
challenging to judge the participants‟ mood and levels of interest during the sessions. These
difficulties resulted from the combination of the pixelated visual image which made it difficult to
clearly view facial features, and the participants‟ masked facial expressions. On the occasions
where the participants became emotionally labile, the SLP found it challenging to promptly detect
these emotions due to the reduced video quality.
Despite the challenges arising from the reduced video quality, a number of strategies were
utilised effectively by the SLP, which helped to minimise the impact of the poor video quality on
treatment outcome. As outlined in Chapter 4, the SLP: (1) relied more heavily on verbal
communication to instruct the participants and to judge their mood and levels of interest; (2)
utilized the store-and-forward function of the online application where further verification of visual
information was necessary; (3) instructed the participants to remain relatively still in front of the
web cameras where possible; and (4) kept to the treatment routine to assist with task
comprehension. In future, the use of higher Internet bandwidth would also help to improve the
video quality during the sessions and lessen the need for compensatory strategies. Furthermore,
181
pre-recorded task demonstrations would also assist SLPs when trying to shape the correct vocal
behaviour during treatment.
5.3.5.4 System Operation
The successful delivery of all treatment sessions was greatly facilitated by the user-
friendliness of the telerehabilitation system, particularly at the participant site. As the online SLP
controlled all aspects of the telerehabilitation system remotely, the participants were not required to
operate the system in any way or have any level of proficiency in the use of computers. Such
factors were particularly relevant in this study as any potential barriers to online treatment delivery
resulting from the need for participants with more complex PD and motor and cognitive difficulties
to manipulate the telerehabilitation system in any way was eliminated. However, from the SLP‟s
perspective, online treatment delivery for these four cases did have its challenges. Consistent with
the previous treatment study in Chapter 4, the online SLP was required to perform additional steps
in order to deliver the LSVT® to face-to-face standards. These steps included: (1) sampling the
mean SPL, F0 and duration measures before recording the values; (2) monitoring task performances
and participant features such as muscle tension, open mouth position and posture via
videoconferencing or the store-and-forward function; and (3) looking directly at the web camera
rather than at the participant on the screen to give the impression of eye contact. In comparison to
the previous study where these steps were performed adequately and treatment was delivered with
relative ease, it was evident in the present study that online delivery for complex cases of PD was
more challenging for the SLP. This was as a direct result of the combined need to perform these
additional steps, as well as to employ a greater number of strategies in order to minimize the impact
of the audio-visual difficulties and participant factors on treatment. It is likely that the relatively
positive treatment outcomes and high participant satisfaction with online treatment in the present
study reflected the level of proficiency of the SLP in operating the telerehabilitation system as well
as her experience in treating complex cases of PD. It may therefore be useful for SLPs to firstly
build up proficiency in online delivery by treating patients with milder cases of PD, and by treating
more complex cases face-to-face initially, before undertaking online LSVT® with more complex
cases. As mentioned previously, future applications with higher bandwidth providing clearer audio
and video qualities, and optimal positioning of the web cameras closer to the participant image for
direct eye contact, would assist to decrease some of the challenges of online delivery for SLPs
working with this population.
182
5.3.5.5 Nature of the Online Treatment Modality
In the present study, the distance factor associated with the online modality posed further
unique challenges to treatment delivery for the SLP. These were particularly evident in the
situations where participants were experiencing muscle tightness, postural difficulties, fatigue, or
were emotionally labile. In contrast to standard face-to-face delivery, the online SLP was unable to
assist in reducing muscle tightness through direct massage, correcting participant posture or
providing further physical prompts to combat the fatigue and keep participants on task. In these
situations, the online SLP relied more heavily on verbal instructions for the participants. However,
this strategy was often ineffective in correcting the behaviour due to the severity of the muscle
tightness and postural difficulties, as well as the level of participant fatigue and/or cognitive
difficulties. As mentioned previously, in specific cases such as these, it may be useful to have a
family member or volunteer present at the participant site to physically assist the participant in
place of the online SLP. The online modality was also thought to be somewhat unnatural and a
distant environment by the SLP in the situations where participants became emotionally labile and
the SLP was unable to quickly detect these subtle changes or to physically comfort them as she
would have done face-to-face. This distinct difference in the participant-professional interaction
between the online and face-to-face modality has also been reported in other online psychology
studies where the poor video quality and distance made it challenging for the professionals to detect
the emotional participants, as well as important non-verbal information useful for diagnosis (Ghosh,
et al., 1997; Zarate, et al., 1997). To some extent, the improved video quality with future
applications may assist in creating a more natural environment, resembling face-to-face.
Due to the case study laboratory design, the current findings cannot be generalised to the
wider clinical setting. Further large-scale randomised controlled trials are needed to compare the
benefits of online versus face-to-face LSVT® for this population. These studies should include
greater numbers of complex cases of PD who receive the LSVT® in both environments, as well as
SLPs with a range of experience levels in online delivery. The next stages of research in which
participants are treated in their own homes (home telecare) should also follow. Additionally, the
feasibility of an extended online LSVT® programme of one to four weeks should be considered for
potential participants, in order to maximise the likelihood of long-term maintenance of treatment
183
gains. This is in keeping with the recommendations for face-to-face LSVT® made in previous
studies for complex cases of PD (Countryman & Ramig, 1993; Countryman, et al., 1994;
Theodoros, et al., 1999). Finally, it would be important to determine the circumstances in which
online delivery may not be feasible or where a hybrid treatment environment of both online and
face-to-face delivery may be better suited to the individual. For example, in the situations where
the SLP judges certain patient factors as potentially impacting on treatment delivery and calibration,
the patient may be better served by receiving the first week of treatment face-to-face. This protocol
would allow the SLP to provide: (1) physical support to the patient to reduce muscle tension and
assist with optimal posture where necessary; (2) prompt feedback to assist with calibration; and (3)
to establish the routine of online treatment delivery. Once these behaviours are established, the
remaining treatment sessions could be delivered online. A hybrid approach should also be
considered for maintenance sessions following treatment. Considerations such as these will need
to be investigated further before online LSVT® can be regarded as an appropriate service delivery
option for this more complex population.
5.5 Conclusion
The findings of the present study largely supported the hypothesis that online LSVT® was
feasible for the participants in the study with complex PD. Despite the variable post-treatment
gains, all participants reported improvements in functional communication, which is the ultimate
treatment outcome for these cases. Although the participant factors including the physical,
communication and cognitive difficulties and effects of neurosurgery did present some unique
challenges to online treatment, they did not substantially impact on treatment delivery in this
environment, or on participant satisfaction. If online LSVT® is adopted into mainstream clinical
practice, it would potentially make a substantial difference to the functional communication and
quality of life of individuals with complex PD.
184
Chapter 6
Home-based Speech Treatment for Parkinson’s Disease Delivered
Remotely: A Case Report
6.1 Introduction
As a future direction in Chapter 4, it was recommended that studies investigating the

effectiveness of remote online services for PD need to be conducted outside of laboratory settings in
order to fully identify the suitability of telerehabilitation for this population. Home telecare or the
use of telecommunication technologies to deliver healthcare services to patients in their own homes
is increasingly being utilised for individuals with chronic disability and the elderly. Home telecare
assists individuals to remain independent within their homes through improved and direct access to
quality healthcare (J. Barlow, Singh, Bayer, & Curry, 2007; Onor et al., 2008). Furthermore, this
modality can assist patients to become advocates for their management, resulting in improved long-
term outcomes and overall quality of life (Averwater & Burchfield, 2005). The benefits of home
telecare have been reported in numerous studies, with high participant satisfaction with this
modality found for individuals with various chronic conditions and geriatric related difficulties
(Botsis & Hartvigsen, 2008; Onor, et al., 2008; Shores et al., 2004). The cost savings resulting
from reduced hospitalisation and the potential for increased collaboration and treatment delivery
have also been associated with high levels of professional satisfaction for this mode of service
delivery (Botsis & Hartvigsen, 2008; Brignell, Wootton, & Gray, 2007; Vasquez, 2008).
Home telecare may be ideal for people with PD. This modality could potentially eliminate
many of the current service access barriers resulting from the individuals‟ physical incapacity,
difficulties with travel and transport, cost of travel, and the large distances which may need to be
traversed in order to access healthcare facilities offering speech-language pathology services
including the LSVT®. Consequently many more people with PD could receive the LSVT®, who
would otherwise remain untreated. Moreover, home-based delivery of the LSVT® would assist
with carryover of treatment effects as patients are presented with increased opportunities to practice
185
their louder voices within their natural home environments during the sessions, which is a
challenging task for the SLP to facilitate in the treatment room.
In the two proof-of-concept studies investigating the feasibility of home-based LSVT®

delivery (Howell, et al., 2009; Tindall, et al., 2008), the authors identified certain advantages of this
modality (see sections 1.11.2 and 4.1 for further details on these studies). Howell and colleagues
(2009) reported that it was convenient for: (1) the participants who were more relaxed in their own
home environments; (2) building rapport with the participants‟ families as they could easily be
present during the sessions and ask questions where necessary; and (3) the SLP as she gained a true
indication of how the participants were using their louder voices in routine speaking tasks within
the home. Furthermore, when compared to face-to-face management, Tindall et al. (2008) reported
an average direct saving in time of 35 hours and money of US $1,220 associated with home-based
LSVT® for their participants.
The growing number of studies reporting the benefits of home telecare and the positive
outcomes achieved using online LSVT® in the laboratory trials in Chapter 4, prompted the current
investigation of home-based treatment in a patient‟s most natural and preferred environment.
Specifically, the present study aimed to investigate the feasibility of home-based LSVT® delivery
for a single case of IPD using the PC-based telerehabilitation system. It was hypothesised that
home-based LSVT® can be effectively delivered, with treatment outcomes consistent with those
described in the literature with traditional face-to-face LSVT® delivery, and those obtained in the
laboratory treatment study in this thesis (Chapter 4).
6.2 Method
6.2.1 Participant
and Social Sciences Ethical Review Committee of The University of Queensland. As per the
original number allocation in the assessment study (Chapter 3), Participant 23, a 65 year old retired
gentleman living in a regional city, 90 km north of Brisbane, was recruited to the study. Participant
186
23 had been diagnosed with IPD six years prior to the study and was classified at Stage I as per the
Hoehn and Yahr Scale. In the previous few years, the participant and his wife had noted a reduction
in his vocal loudness and speech intelligibility as well as a breathy vocal quality. As a result,
Participant 23 was finding it more difficult to be understood in group situations, where there was
background noise and occasionally, while speaking on the telephone. His speech and voice
difficulties were beginning to impact on his duties as a coordinator of a community group in his
local area. In daily life, Participant 23 had begun to avoid speaking by telephone and would often
spend the greater part of his day at home, where he did not have to talk to anyone. Participant 23
had not previously received any speech-language pathology services. There were also no public or
private speech-language pathology services offering the LSVT® in his local community at the time
of the study. The intensive nature and commitment of the LSVT® programme, and participant
fatigue when driving, made it difficult for Participant 23 to seek treatment for his speech and voice
difficulties outside of his local area.
The participant had taken part in the assessment study of this thesis (Chapter 3) where his
speech was classified by the principal investigator as demonstrating mild hypokinetic dysarthria.
This was based on his assessment results on the monologue task where he achieved a mean SLP of
68.13 dB and displayed mildly reduced OIC. Please refer to section 4.2.3 for further details on the
dysarthria severity classification. A videolaryngoscopic examination conducted prior to treatment
by an Ear Nose and Throat specialist revealed some vocal fold bowing, a feature consistent with
IPD. Participant 23 wore hearing aids for his mild-moderate bilateral high frequency hearing loss.
Participant 23 was cleared for inclusion in the study, having met the criteria outlined in section
4.2.3 (i.e. the presence of hypokinetic dysarthria that was impacting on communication and vocal
fold structure and movement consistent with diagnosed PD; stimulability to increased loudness; and
a consistent drug regimen for IPD). Furthermore, the participant did not display any of the
following aspects that warranted exclusion including: (1) the presence of a speech and/or language
disturbance or a co-existing neurological disorder inconsistent with IPD; (2) a severe uncorrected
auditory and/or visual disturbance; (3) a cognitive disturbance inconsistent with the capacity to
provide informed consent; (4) respiratory dysfunction unrelated to the neurological disorder; (5) a
positive history of alcohol abuse; and (6) previous LSVT® within the last 12 months of the study.
187
6.2.2 Procedure
One of the four LSVT® certified SLPs who had taken part in the previous treatment studies
was assigned to the study and delivered the LSVT® sessions according to standard clinical practice
1995) as outlined in section 4.2.5. The delivery of the online sessions was also consistent with the
online procedure of the previous studies (section 4.2.6.1). The SLP was located within the
Telerehabilitation Research Unit at The University of Queensland and delivered the treatment to the
participant‟s home via the custom made PC-based telerehabilitation system described elsewhere in
this thesis (see section 2.2). One end of the telerehabilitation system was located at the SLP site,
while the other was a portable version of the system on a laptop computer. The latter was located at
the participant site and was connected to the telerehabilitation system‟s acoustic speech processor
via a universal serial bus (USB) port. As mentioned in section 4.2.6.1, the features of the
telerehabilitation system designed to deliver the LSVT® included the ability to: (1) provide real-
time videoconferencing at 320 x 240 pixel resolution; (2) present reading material for the
participant; (3) manipulate the web cameras at the participant site via a robotic arm to maintain a
clear view of the participant throughout the session; (4) obtain average measures of SPL, F0 and
duration via the telerehabilitation system‟s acoustic speech processor and acoustic measurement
software tool; and (5) capture high quality video (640 x 480 pixel resolution) and audio
(compressed at 384 kbit/s) for later examination. Figures 6.1 and 6.2 display the SLP and
participant treatment sites respectively. A 128 kbit/s Internet connection over the public network
was established between the two sites of the telerehabilitation system, consistent with the
assessment and treatment studies in this thesis. This was achieved using a 128 kbit/s Internet
connection at the SLP end that limited the bandwidth of the Asynchronous Digital Subscriber Line
(ADSL) broadband Internet connection used by the participant.
188
Figure 6.1 SLP treatment site. Note. (1) the videoconferencing system displaying the participant,
(2) acoustic measurement software tool displaying SPL and F0 data ; and (3) web camera.
Figure 6.2 Participant treatment site. Note. (1) laptop computer containing the videoconferencing
system; (2) robotic arm with web cameras attached; (3) mirror used for checking the microphone
distance; (4) telerehabilitation system‟s acoustic speech processor; and (5) homework reading
material.
189
Prior to treatment, the participant‟s end of the telerehabilitation system was set-up in a
quiet room of his home by the principal investigator. This took approximately 10 min to complete.
The participant was shown how to perform the basic functions required for treatment, including
turning on/off the PC, activating/closing the application, and positioning the headset microphone,
which in the laboratory studies had been performed by the SLP. Once the set-up was completed,
the microphone distance was adjusted to 5 cm from the corner of the participant‟s mouth by the
principal investigator. This distance helped to reduce sound distortion, maximize visibility of the
participant‟s face and allowed for accurate recordings of SPL and F0. To verify this optimal
positioning, the SPL data generated by the telerehabilitation system was then confirmed against a
conventional Digital SLM held at approximately 30 cm from the participant‟s mouth during three
sustained /a/ phonations. As the microphone arm was somewhat flexible and manipulation could
result in some changes in the distance, the participant was instructed to avoid adjusting the
microphone arm over the course of treatment and to take care where he placed the headset
microphone when not in use. Participant 23 checked the microphone distance in a mirror at the start
of each session. During treatment, the SLP also wore a headset microphone and communicated
with the participant over the videoconferencing link. The SLP controlled all displays on the
participant‟s screen during the sessions, without the need for the participant to operate the
telerehabilitation system.
In keeping with the protocol in Chapter 4, the post-LSVT® assessment was conducted in
the face-to-face environment at The University of Queensland within 48 hours of treatment
completion. As outlined in section 4.2.7, a battery of acoustic and perceptual measures was used in
these assessments. The acoustic measures included: (1) mean SPLs (dB-C) of the participant‟s
speech during six maximum sustained vowel phonations of /a/, reading of The Grandfather Passage
and a 30 s monologue; (2) maximum duration of six sustained vowel phonations (s); and (3)
maximum F0 range obtained from the average of the highest and lowest F0 levels (Hz) from a series
of six vocal glides that were then converted to a maximum range in semitones (ST) (de Pijper,
2007). The acoustic measures were obtained using the telerehabilitation system.
190
Perceptual measures of voice and speech included ratings of voice parameters (breathiness,
roughness, loudness level, loudness variability and pitch variability), OAP and OIC. The ratings on
all of the parameters were made using DME scaling by the two additional and independent SLPs
who had not taken part in the assessments or treatments and who were blind to the intent of the
study. Consistent with the previous studies in this thesis, a lower DME value for the vocal
breathiness and roughness parameters only denoted improved performance. The DME method and
assessment procedure have been described in detail elsewhere (section 4.2.7.2).
The ASSIDS was not used in the present study as the additional measure of speech
intelligibility. This is because Participant 23‟s pre-treatment results on the ASSIDS were already
within the normal range (98% for word intelligibility; 100% for sentence intelligibility;
communication efficiency ratio of 1.08). Therefore, due to the ceiling effects of the ASSIDS itself,
any expected treatment changes in speech intelligibility would not have been adequately
represented with this assessment.
The participant completed a satisfaction questionnaire (detailed in section 4.2.7.3)

evaluating the level of satisfaction with the online treatment sessions, audio and video quality
during treatment and overall satisfaction with this modality. Further to the outcome measures in
Chapter 4, the participant was invited to comment on any aspects relating to home-based LSVT®
and his speech and voice features. The treating SLP was similarly invited to comment anecdotally
on any aspects relating to the home-based treatment delivery.
6.3 Results
Descriptive comparisons and degree of change were determined between the pre- and post-
LSVT® acoustic and perceptual measures. Treatment performance was compared against the
criteria determined in this thesis to provide some indication of treatment success (section 5.3). For
the acoustic parameters, substantial treatment gains were made on all SPL tasks (mean
improvements of 6.13 dB for maximum sustained vowel phonation; 12.28 dB for reading and 11.32
dB for monologue loudness), and on the maximum duration of sustained vowel phonation
parameter (4 s mean improvement). Maximum F0 range failed to show an improvement with
191
treatment and rather, a lower post-treatment value than baseline was recorded. These results are
summarised in Table 6.1.
Table 6.1 Pre- and Post-LSVT Values and Treatment Changes for the Acoustic Parameters

Reading (dB) 71.42 83.70 12.28
Monologue (dB) 68.13 79.45 11.32

of sustained vowel
phonation (s)
For the perceptual parameters, measurable treatment gains were observed only for vocal
breathiness, while minimal changes were recorded for vocal roughness and OIC. No treatment
changes were evident for OAP, with this parameter remaining at the high pre-treatment DME value
of 162.18. Lower performance with treatment was evident on the remaining parameters of loudness
level, loudness variability and pitch variability. The pre- and post-LSVT® values and treatment
changes are displayed in Table 6.2 for the perceptual measures.
192
Table 6.2 Pre- and Post-LSVT® DME Values and Treatment Changes for the Perceptual
Parameters
Breathiness 75.00 44.67 30.33
Roughness 79.43 64.57 14.86
Loudness level 186.21 151.36 34.85
Loudness variability 120.23 114.82 5.41
Pitch variability 104.71 95.50 9.21
OAP 162.18 162.18 0
OIC 128.82 141.25 12.43
overall speech intelligibility in conversation.
On the participant satisfaction questionnaire, Participant 23 rated the audio and video
quality as excellent and stated that he would prefer online sessions to face-to-face for the future
management of his condition. Overall, the participant was very satisfied with online treatment.
Home-based LSVT® was successfully delivered on a 128 kbit/s PC-based Internet connection via
the public network for all sessions. Overall, there were no failed treatment sessions and the
majority (n = 13) of the sessions ran very smoothly, without technical difficulties and with adequate
audio and video quality for treatment delivery. During the remaining three sessions, some
networking difficulties considerably compromised the audio and video quality. These issues were
able to be resolved by disconnecting and re-establishing the videoconferencing connection between
the two ends of the telerehabilitation system, thus allowing these sessions to continue.
193
6.4 Discussion
The present case report demonstrates the feasibility of home-based delivery of the LSVT®
via a PC-based telerehabilitation system operating on a 128 kbit/s Internet connection via the public
network and supports the study hypothesis. On the whole, Participant 23 showed substantial
improvements with home-based LSVT® for most of the acoustic parameters and one of the
perceptual parameters. Participant 23 also indicated a high level of satisfaction with home-based
LSVT®. Furthermore, there were no failed treatment sessions and the majority of the sessions ran
very smoothly with adequate audio and video quality for treatment delivery.
On all acoustic measures of SPL (maximum sustained vowel phonation, reading and
monologue loudness), Participant 23 showed clinically relevant improvements with home-based
LSVT®. The mean levels of improvement (6.13 dB for maximum sustained vowel phonation; 12.28
dB for reading; and 11.32 dB for monologue loudness), were largely consistent with the treatment
outcomes reported for the IPD participants treated in the online and face-to-face environments in
Chapter 4 (section 4.4.1.1), and with the efficacy studies reporting face-to-face LSVT® (Ramig, et
al., 1996; Ramig, Countryman, et al., 1995). Figure 6.3 displays the mean treatment changes on the
SPL measures for Participant 23 and the participant groups mentioned above. Please note that as
the standard deviations for the SPL measures were not consistently reported in the literature, they
are therefore not displayed for any of the comparison groups or for Participant 23 in this figure.
194
Figure 6.3 Comparison of mean SPL treatment gains for Participant 23 with those reported in the
efficacy studies of face-to-face LSVT® and in the laboratory trials in Chapter 4. Note. Ramig,
Countryman, et al. (1995) and Ramig et al. (1996) are studies of face-to-face LSVT® for IPD; IPD
Online and IPD FTF are the participants who received online and face-to-face LSVT® in Chapter 4
respectively; Sustained „ah‟ is the maximum sustained vowel phonation task.
For all measures of mean SPL, Participant 23‟s baseline values were generally higher than
the comparison groups mentioned above, and reflected his mild level of hypokinetic dysarthria at
the time of the study. As a direct result, less extensive treatment change was observed on the
maximum sustained vowel phonation task. A treatment change in line with the 10-14 dB
improvement reported in the face-to-face literature would have resulted in an SPL value beyond the
acceptable maximum level of 90 dB with the LSVT® (Ramig, Pawlas, et al., 1995). Despite the
higher pre-treatment values, Participant 23 achieved clinically relevant post-treatment mean gains
on the SPL parameters of reading and monologue loudness. Furthermore, the post-treatment mean
values for all SPL parameters were slightly higher than the mean values for two groups of healthy
older adults speaking at a comfortable loudness level (Fox & Ramig, 1997; Ramig, Sapir, Fox, et
al., 2001). Although it is acknowledged that comparisons were made to control data using small
samples sizes (n = 14 participants in each study), the data provide a good comparison for treatment
success and clinical relevance of outcomes. Figure 6.4 displays Participant 23‟s post-treatment
mean values for all SPL measures, in relation to the mean values of the comparison groups
mentioned above. Please note that for the healthy older adults, the figure only displays the highest
SPL values obtained from their two phonation trials on each task.
195
Figure 6.4 Comparison of post-treatment mean SPL values for Participant 23 with those reported in
the efficacy studies of face-to-face LSVT®, healthy older adults and laboratory trials in Chapter 4.
Note. Ramig, Countryman, et al. (1995) and Ramig et al. (1996) are studies of face-to-face LSVT®
for IPD; IPD Online and IPD FTF are the participants who received online and face-to-face LSVT®
in Chapter 4 respectively; Ramig et al. (2001) and Fox and Ramig (1997) are studies of healthy
older adults; Sustained „ah‟ is the maximum sustained vowel phonation task.
For maximum duration of sustained vowel phonation, a clinically relevant treatment gain
of 4 s was obtained. This was comparable with the minimal improvement (3.72 and 4.9 s; SD not
reported) reported in the efficacy studies for IPD participants with mild to moderate hypokinetic
dysarthria following face-to-face LSVT® (Ramig, et al., 1994; Ramig, Countryman, et al., 1995).
The treatment gain was also higher than the post-treatment improvements achieved by the online
(M = 0.61 s; SD = 4.09) and face-to-face IPD participants in this thesis (M = 0.24 s; SD = 4.84).
Please see section 4.4.1.2. Additionally, the post-LSVT® value of 13.67 s in the present study was
also comparable to the duration values (M = 17.94 s, SD = 5.01) reported for healthy older adults
(Fox & Ramig, 1997).
On the maximum F0 range task, the treatment outcome of reduced range following the
LSVT® (3.86 ST less) presented an unexpected result. Previous literature has suggested a minimum
of 4 ST improvement with face-to-face LSVT®, reflecting greater cricothyroid and thyroarytenoid
muscle activity with treatment (Ramig, et al., 1994). However, it is possible that a change on this
parameter following treatment may not be evident consistently during testing at this milder level of
dysarthria severity. Consequently, the post-treatment range remained relatively consistent with pre-
treatment. Participant 23‟s post-treatment phonations on this task, however, were clear in quality
196
compared to pre-treatment, where pitch breaks and a breathy vocal quality were evident. This
finding suggests that the LSVT® training on the maximum F0 range task assisted the participant in
self-monitoring the quality of his phonations as part of calibration, and this practice generalised to
the post-treatment assessment. On the whole, the positive changes with treatment on the majority
of the acoustic parameters have highlighted the feasibility of home-based LSVT® for improving
vocal adduction, respiratory drive, and promoting calibration of loud phonation to daily living,
consistent with expectations of face-to-face delivery.
The post-treatment improvements in mean SPL were only reflected perceptually as a

measurable reduction in vocal breathiness. This change with treatment has also been reported in the
LSVT® literature (Baumgartner, et al., 2001; El Sharkawi, et al., 2002; Ramig, et al., 1994; Ramig,
Countryman, et al., 1995; Sapir, et al., 2002), and observed in Chapter 4 for both the online and
face-to-face IPD participants (sections 4.3.2). It is possible that the lack of perceived improvements
on the remaining speech and voice parameters were related to the participant‟s mild degree of
hypokinetic dysarthria which made it difficult for the raters to detect demonstrable changes on these
parameters following treatment. It is recognised that the perception of speech and voice changes is
more difficult in milder degrees of speech impairment (Kreiman & Gerratt, 1998), and may not be
as accurate as objective measurement. Despite the lack of substantial improvements in specific
speech and voice parameters, the participant reported improved performance in everyday speech
activities following treatment.
197
6.4.3 Perspectives Regarding Home-based Lee Silverman Voice Treatment® Delivery
6.4.3.1 Participant Perspectives
Home-based delivery of the LSVT® was well received by the participant. On the
satisfaction questionnaire, Participant 23 rated that he was overall very satisfied with the online
treatment and found the audio and video quality of the telerehabilitation system to be excellent. His
level of hearing loss, which was corrected by hearing aids, did not interfere with his ability to hear
instructions over the videoconferencing link. Interestingly, on the satisfaction survey, Participant
23 also reported that he would prefer online sessions to face-to-face for the future management of
his condition. He felt that the online method provided: (1) ease of access to treatment without the
hassle of travelling and the need to leave his own home; (2) time-savings from not having to travel;
and (3) a friendly technical interface for treatment using reliable technology. Participant 23 further
commented:
I found it very easy doing it from home. I think it’s a great way to do it, even where there
are not problems of distance involved. It’s just so much simpler and you’re more relaxed
and it’s easier to fit into your daily routine.
The convenience of home-based LSVT® proved to be a highly motivating factor for

Participant 23 to perform well. Similar findings have also been reported in other home telecare
studies where participants were motivated with treatment and accepting of the technology used
when they could be treated in their natural or least restrictive environment (Hornsby & Hudson,
1997; Mashima, et al., 2003). On the whole, Participant 23 benefited from home-based LSVT® and
found the treatment useful for increasing his loudness to pre-morbid levels, and for integrating and
maintaining the treatment gains of improved loudness level and vocal quality in daily life. As a
result of these positive treatment changes, Participant 23 felt that his speech sounded natural and
close to how he remembered it prior to PD. Participant 23 stated:
I’ve been quite pleasantly surprised at the process and particularly the results. I think my
voice has improved a lot. It’s stronger, more confident. As a result of treatment, I have
regained confidence in engaging in conversation with family and friends, speaking on the
telephone and in public at the community group meetings. I even made a brief prepared
speech at my daughter’s wedding, something I would not have contemplated doing before
198
treatment. Friends and family members have also noted improvements in my speech in
loudness, intelligibility and confidence.
6.4.3.2 Speech-Language Pathologist Perspectives
In the present study, the same telerehabilitation system was used as described in the
previous assessment and treatment chapters. The real-time videoconferencing feature of the
application and the ability to capture and display mean SPL, F0 and duration data, as well as therapy
materials online were integral to the successful delivery of the treatment. These features allowed
the SLP to: (1) provide timely instructions to the participant and assist with shaping correct voice
productions and overall calibration; (2) appropriately monitor mean SPL, F0 and vocal quality; and
(3) maintain good rapport with the participant. From the SLP‟s perspective, the telerehabilitation
system was user-friendly and allowed for effective delivery of home-based LSVT® via the public
network. The treatment delivery was in keeping with the previous online laboratory trials in this
thesis, and incorporated the essential components of traditional face-to-face delivery. Overall, there
were no failed treatment sessions and the majority of sessions ran very smoothly, with sufficient
audio and video quality for treatment delivery. This was also greatly facilitated by the user-
friendliness of the telerehabilitation system. On only three occasions was the SLP required to
disconnect and re-establish the videoconferencing connection due to networking difficulties. The
process of re-establishing the connection helped to improve the audio and video quality and allowed
all sessions to continue. As such, these networking difficulties were not seen to adversely affect the
study.
Although occasional audio delays were encountered during treatment, they were
effectively managed by the SLP and participant, who waited until the other had clearly finished
speaking before replying, and further, by the SLP who used shorter and more precise instructions.
Strategies, however, were needed to maximise the video quality, as the frame rate and pixelated
video image especially during movement made it more difficult for the SLP and participant to
clearly view each other during the session. These issues were similarly noted in the laboratory
assessment and treatment studies and certain strategies were also successfully adopted during the
home-based sessions to overcome many challenges resulting from these features. As mentioned
previously (section 4.4.3.2), useful techniques included the SLP: (1) and participant sitting
199
relatively still in front of the PC monitor during the sessions; (2) using easy to detect hand-cues for
quick input; (3) using the store-and-forward function to record the desired task when necessary; and
(4) relying more heavily on specific verbal directions and participant feedback rather than on visual
information. These strategies were easily adopted in the sessions and aided the smooth delivery of
home-based treatment in line with the previous online laboratory studies in this thesis. As
mentioned previously, the use of telerehabilitation systems with higher Internet bandwidth in the
future will help to improve the interface by reducing any audio and video difficulties and lessening
the need for compensatory strategies. This would allow for a closer resemblance of the face-to-face
modality.
Other procedures necessary to ensure the smooth delivery of treatment in the home
environment were identified during the study. The sessions were conducted in a quiet room of the
participant‟s home, which reduced any household distractions and noise, and telephone calls were
not taken during the sessions. Furthermore, the microphone distance from the participant‟s mouth
was kept constant, to ensure that the SPL and F0 acoustic data were consistent between sessions.
The participant also took great care when handling the microphone during the treatment programme
to ensure that this distance was maintained. The future use of headset microphones with rigid
microphone arms and/or a desktop microphone at a set distance that is externally calibrated would
also help to reduce the potential impact of external variables on the accuracy of the acoustic
information obtained.
As the present study was a single case design, the findings cannot be generalised to the
wider PD population. Future large-scale studies are needed to investigate the effectiveness of
home-based LSVT® delivery with a greater number of participants with PD of varying types, stages
of the disease, hypokinetic dysarthria severity levels, and where applicable, participants who have
undergone neurosurgical intervention. Greater numbers of treating SLPs with a range of experience
levels in online delivery should also be included. These studies would provide additional validation
for home-based service delivery and the instances where home telecare is most appropriate.
Furthermore, the investigation of remote assessment of the speech and voice disorder associated
200
with PD should occur, together with cost-analyses, in order to establish the full benefits of this
mode of service delivery for people with PD.
6.6 Conclusion
Overall, home-based LSVT® delivery was feasible for Participant 23, supporting the
hypothesis for the present study. The treatment gains and high participant satisfaction and
motivation with home-based LSVT® illustrated the potential of this mode of service delivery for
people with PD. Home telecare may assist in reducing the effects of physical disability, distance,
transport, travel difficulties and cost for people with PD, which at present represent substantial
barriers to service access. Home telecare may also facilitate earlier access to intervention than
currently exists for people with PD, thus allowing individuals to remain independent and active
longer within their own homes and communities.
201
202
Chapter 7
Conclusions
Due to the progressive and debilitating nature of PD, individuals require ongoing speech-
language pathology rehabilitation services to help improve or maintain their functional
communication. It has been recommended that optimal management of PD should include timely
access to assessment, and to efficacious treatment such as the LSVT® which requires intensive
delivery, and one-to-one participant-SLP interaction. To date, however, patient access to speech-
language pathology services for appropriate management is limited in Australia and worldwide
(Grimm, et al., 2004; Hartelius & Svensson, 1994; Mutch, et al., 1986; Oxtoby, 1982; Peto, et al.,
1997). Certain barriers that impact on service access have been identified such as the physical
incapacity of individuals, difficulties with transport and travel, associated travel costs, and the large
distances that need to be covered to attend healthcare services. The use of telerehabilitation as an
alternate or additional mode of service delivery for this population may provide a possible solution
to these access issues. To date, however, there is a paucity of large-scale telerehabilitation studies
demonstrating the feasibility of this mode of service delivery for PD. The aim of the work
described in this thesis was to investigate the validity of telerehabilitation for the assessment and
treatment of the speech and voice disorder associated with PD, and to provide a framework for the
online delivery of services for this population.
As no available PC-telerehabilitation system at the time of this thesis was able to meet all
the assessment and treatment demands of PD in line with face-to-face management, a custom made
system was developed for the thesis studies. Chapter 2 provided an in-depth description of the
feature set of the telerehabilitation system which included: (1) videoconferencing via a 128 kbit/s
Internet connection; (2) control of the participant web cameras remotely; (3) store-and-forward
capabilities; (4) the capacity to display printed materials on the participant‟s remote computer;
(5) ease of operation; and (6) the ability to view and objectively sample calibrated average measures
of vocal SPL, F0 and duration. The first study in this thesis investigated the validity of the
telerehabilitation system as an acoustic measurement tool through a series of calibration and
verification phases. These phases were conducted successfully in Chapter 2 as evidenced by
comparable acoustic measures of SPL and F0 obtained between the telerehabilitation system and the
203
reference measurement tools. The findings supported the use of the telerehabilitation system as the
acoustic measurement tool for the subsequent studies in this thesis.
The investigations that followed evaluated the validity of the telerehabilitation system for
the assessment and treatment of PD by comparison with the traditional face-to-face modality. Of
these investigations, Chapter 3 was an equivalence study that determined the validity and reliability
of online assessment of the speech and voice disorder associated with PD in a laboratory setting. A
total of 61 participants with PD and hypokinetic dysarthria ranging in severity from mild to severe
were assessed simultaneously in the online and face-to-face environment on a number of acoustic
and perceptual measures specifically designed for this study. The findings indicated that valid and
reliable assessment of the speech and voice disturbances of PD could be achieved via the
telerehabilitation system. This was established from the comparable ratings obtained between the
online and face-to-face environments across the majority of acoustic and perceptual parameters.
Furthermore, comparable intra- and inter-rater reliability was determined on all measures between
the two environments. The participants who received the online-led assessments also expressed a
high level of satisfaction with the modality. Further, there were no technical issues or failed
assessment sessions in the online environment and administration was achieved with adequate audio
and video quality over videoconferencing. However, it was noted that some challenges unique to
the online environment arose during the sessions as identified by the SLPs. As described in the
chapter, the challenges related to the: (1) occasional audio delays and difficulties rating subtle
speech features with videoconferencing; (2) frame rate and pixelated video quality with
videoconferencing; and (3) difficulties establishing eye contact with the participants in the online
environment. Despite these challenges, appropriate strategies were successfully implemented by
the SLPs and participants during the sessions which ensured that the challenges did not impact
greatly on assessment administration and outcomes.
A series of treatment studies followed in this thesis. Chapter 4 investigated the validity of
online LSVT® delivery using the PC-based telerehabilitation system. The chapter described a
laboratory-based randomised controlled non-inferiority trial where 17 participants with IPD and
mild to moderate hypokinetic dysarthria were randomised to each treatment environment. Non-
inferiority of outcomes was confirmed between online and face-to-face LSVT®. Additionally,
significant improvements with treatment in both environments were reported for the majority of
parameters and the post-treatment mean SPL values were in line with those reported for healthy
204
older adults (Fox & Ramig, 1997; Ramig, Sapir, Fox, et al., 2001). Participant satisfaction with
online LSVT® was also high overall. The online treatments were successfully delivered with
adequate audio and video quality over videoconferencing as well as without technical difficulties
for the majority of sessions. However, as noted in the assessment study, some challenges with
online delivery were encountered by the SLPs as a result of the occasional audio delays, frame rate
and pixelated video quality during videoconferencing. Additionally, the restrictive web camera
range made viewing of specific participant factors more difficult, and the SLPs were required to
perform a number of extra system operations in order to obtain information comparable to face-to-
face delivery. Again, these challenges were able to be appropriately managed without considerable
impact on the online treatment delivery. Overall, the study confirmed that treatment of PD could be
effectively achieved using the telerehabilitation system.
In Chapter 5, a laboratory-based study was undertaken to determine the feasibility of

online LSVT® delivered to four individuals with complex PD presentations. These participants
with either PSP or surgical IPD displayed greater communication, cognitive and physical
difficulties than the IPD participants in the previous study. Improvements with treatment were
generally variable for these cases across the acoustic and perceptual parameters, and the results
were similar to those reported in the literature for complex cases of PD following traditional face-
to-face LSVT® (Countryman & Ramig, 1993; Countryman, et al., 1994; Theodoros, et al., 1999;
Ward, et al., 2000). It was suggested in this chapter that the involvement of the participant factors
rather than the online treatment delivery had impacted on the treatment performance for these
individuals. Despite the variable results, it was encouraging to note that improved functional
communication and high participant satisfaction with the online modality were reported by all
participants. There were also no failed treatment sessions or technical difficulties experienced
within the online environment in this study. It was acknowledged, however, that greater challenges
with telerehabilitation were encountered by the SLP for this group of complex cases compared to
the previous treatment study. In addition to the aforementioned technical difficulties, it was evident
that the participant factors presented further challenges to the delivery of the sessions, task training
and calibration. Recommendations were made in this study that SLPs should firstly be proficient in
treating patients with more complex PD in the traditional setting as well as in delivering online
LSVT® to milder cases of PD before undertaking the combination of complex cases online. As
demonstrated in this study, such considerations would ensure the smooth delivery of treatment and
outcomes comparable to face-to-face management. Although it was acknowledged that further
research is needed in this area, the study largely demonstrated the feasibility of online LSVT® for
205
complex cases of PD, and provided insight into the use of telerehabilitation in the treatment of
people with complex PD.
The final study in this thesis investigated the feasibility of remote home-based delivery of
the LSVT®. A single participant with IPD and mild hypokinetic dysarthria participated in the study
and demonstrated substantial improvements following treatment on most of the acoustic parameters
and perceptually, on one parameter. The acoustic improvements were largely in keeping with the
treatment gains reported for participants in Chapter 4, as well as the face-to-face efficacy studies in
the literature (Ramig, et al., 1996; Ramig, Countryman, et al., 1995). The post-treatment values
were also consistent with the performances of healthy older adults (Fox & Ramig, 1997; Ramig,
Sapir, Fox, et al., 2001). There were no failed treatment sessions in this study and a high level of
participant satisfaction with online treatment was reported. Although some of the strategies utilised
in the previous chapters were also necessary in this study to minimise the occasional audio-visual
difficulties during videoconferencing, they were appropriately applied and the sessions were able to
be conducted efficiently with minimal disturbances. Other considerations specific to the home
environment ensured the smooth delivery of treatment in this setting. These included conducting
the sessions at the participant end from a quiet room in the home, not taking telephone calls during
the sessions, and maintaining a constant microphone distance from the participant‟s mouth
throughout treatment. The ability to maintain the set microphone distance via rigid microphone
arms and/or the use of externally calibrated desktop microphones was suggested as a future
consideration for home-based treatment to maximise the accuracy of the acoustic signal. Overall,
this study demonstrated the feasibility of transitioning online treatment delivery from the laboratory
to the real-world setting that is the ultimate destination for speech-language pathology
telerehabilitation services for PD.
Further research is needed to build on the findings in this thesis in order to gain a
comprehensive understanding of the benefits and challenges of this mode of service delivery for
PD, as well as the situations where telerehabilitation is most applicable. As outlined earlier
(Chapters 3 to 6), studies should involve randomised controlled trials with large participant and SLP
numbers. Assessment and treatment studies should investigate the potential impact of participant
factors such as the severity of the communication, cognitive and physical difficulties on variables
such as online assessment administration and online LSVT® delivery and outcomes, as these factors
were identified in the complex PD study as requiring further verification. Where possible,
206
comparison should be made to face-to-face management. Specifically for treatment, it would be
important to also determine the situations (e.g. complex cases) where a possible extended online
LSVT® programme or a hybrid treatment approach would best cater to patient needs.
Considerations for online maintenance sessions following online LSVT® would further assist in
determining the feasibility of telerehabilitation as a complete end-to-end mode of service delivery
for PD. In addition to increasing participant numbers, future research should also involve large
numbers of SLPs of varying degrees of proficiency in online assessment administration and
treatment delivery. These factors would assist in determining the possible effects of SLP
proficiency on outcomes in the online environment. These findings would provide valuable
information about the level of SLP training required for online delivery, the useability of
telerehabilitation systems in this environment, as well as the feature set that is essential across
systems to allow for the optimal management of PD. By establishing assessment and treatment
guidelines for telerehabilitation in speech-language pathology based on research findings, the
uptake of this mode of service delivery into mainstream clinical practice would be further assisted.
The studies in this thesis have demonstrated the useability of the custom made PC-based
telerehabilitation system for the assessment and treatment of PD in line with face-to-face
management, and provided a framework for the feature set necessary in this area. In particular, the
successful delivery over the low-bandwidth Internet connection, consistent with the minimum
bandwidth available in Queensland‟s public health system, highlighted the direct applicability of
telerehabilitation for this sector. The uptake of telerehabilitation will continue to increase in
healthcare alongside more affordable and improved telecommunication infrastructure,
communication and information technologies, project funding and reimbursement opportunities. As
the uptake continues, it would be ideal to investigate the potential benefits of utilising more
sophisticated telerehabilitation systems on the assessment and treatment outcomes of PD. As
identified in this thesis, further studies should utilise telerehabilitation systems operating on high-
bandwidth Internet connections with improved audio and video qualities, instant playback
capabilities, extended web camera range and zoom function, and improved positioning of the web
camera closer to the remote image on the PC screen. It is likely that by utilising telerehabilitation
systems that allow for a close resemblance to face-to-face services, user satisfaction would increase
as would the integration of this modality into mainstream clinical practice.
207
Finally, the next stages of research in which participants are assessed and treated within
their own homes should follow. Such studies would present a realistic indication of the validity of
telerehabilitation for PD in the real-world setting, and establish the role of telerehabilitation in
reducing the current access barriers that exist with traditional services for this population. Together
with in-depth investigations of participant and SLP satisfaction with telerehabilitation, and the cost-
benefits for all members of the healthcare system, these studies would provide a comprehensive
view of the benefits of telerehabilitation and the likelihood of its adoption into mainstream practice.
In conclusion, the studies in this thesis form the basis for the assessment and treatment of
the dysarthric speech and voice disorder associated with PD via telerehabilitation. The comparable
outcomes between the online and face-to-face environments in these studies were very encouraging
and have highlighted the potential of telerehabilitation as an additional or alternate mode of service
delivery for the rehabilitation needs of PD in speech-language pathology. The telerehabilitation
system described in this thesis provided a framework for the delivery of services to people with PD,
with wide and direct applicability to speech-language pathology centres in Australia and worldwide,
where low-bandwidth applications are still the predominant systems used. The potential of
telerehabilitation is far reaching for both adult and paediatric populations in speech-language
pathology as well as for allied health in general, where patient access to services is currently
compromised. It is an exciting time for research in telerehabilitation where a large number of
pioneering efforts continue to add to the growing knowledge base regarding the benefits of this
approach. Telerehabilitation, and more widely, telehealth, has strongly emerged as a front runner
for healthcare with the potential to address the inequity issues that are currently faced in society.
Hopefully one day, access to healthcare for all may actually be a reality.
208
References
Access Economics. (2007). Living with Parkinson's disease: Challenges and positive steps for the
future: Report for Parkinson's Australia.
Adams, S. G. (1997). Hypokinetic dysarthria in Parkinson's disease. In M. R. McNeil (Ed.), Clinical
management of sensorimotor speech disorders (pp. 261-286). New York: Thieme.
Alexander, M. (1995). Telemedicine in Australia. 1: The health-care system and the development of
telemedicine. Journal of Telemedicine and Telecare, 1, 187-195.
Australian and New Zealand Telehealth Committee (ANZTC). (2002). National telehealth plan for
Australia and New Zealand. Canberra: Public Production Unit, Commonwealth Department
of Health and Ageing.
Argue, J. (2000). Parkinson's disease and the art of moving. Oakland, CA: New Harbinger.
Aronson, A. E. (1990). Clinical voice disorders. New York: Thieme-Stratton.
Asher-McDade, C., Roberts, C., Shaw, W. C., & Gallager, C. (1991). Development of a method for
rating nasolabial appearance in patients with clefts of the lip and palate. The Cleft Palate-
Craniofacial Journal, 28(4), 385-390.
Australian Bureau of Statistics. (2009a). Year Book Australia 2009: Geographic Distribution of the
Population. Canberra: ABS Catalogue No: 1301.0.
Australian Bureau of Statistics. (2009b). Year Book Australia 2009: Geography of Australia.
Canberra: ABS Catalogue No: 1301.0.
Australian Bureau of Statistics. (2009c). Year Book Australia 2009: Population Size and Growth.
Canberra: ABS Catalogue No: 1301.0.
Averwater, N. W., & Burchfield, D. C. (2005). No place like home telemonitoring can improve
home care. Healthcare Financial Management, 59(4), 46-52.
Barlow, J., Singh, D., Bayer, S., & Curry, R. (2007). A systematic review of the benefits of home
telecare for frail elderly people and those with long-term conditions. Journal of
Telemedicine and Telecare, 13(4), 172-179.
Barlow, S., Iacono, R. P., Paseman, L. A., Biswas, A., & D'Antonio, L. (1998). The effects of
posteroventral pallidotomy on force and speech aerodynamics in Parkinson's disease. In M.
Cannito, K. M. Yorkston & D. R. Beukelman (Eds.), Neuromotor speech disorders: Nature,
assessment and management (pp. 117-155). Baltimore: Brookes.
209
Baron, M. S., Vitek, J. L., Bakay, R. A., Green, J., Kaneoke, Y., Hashimoto, T., et al. (1996).
Treatment of advanced Parkinson's disease by posterior GPi pallidotomy: One year results
of a pilot study. Annals of Neurology, 40(3), 355-366.
Bashshur, R. L. (1997). Telemedicine and the health care system. In R. L. Bashshur, J. H. Sanders
& G. W. Shannon (Eds.), Telemedicine theory and practice (pp. 5-37). Illinois: Charles C
Thomas Publisher Ltd
Bashshur, R. L., Reardon, T. G., & Shannon, G. W. (2000). Telemedicine: A new health care
delivery system. Annual Review of Public Health, 21, 613-637.
Bauer, J., & Ringel, M. (1999). Telemedicine and the reinvention of healthcare. New York:
McGraw-Hill Companies Inc.
Baumgartner, C. A., Sapir, S., & Ramig, L. O. (2001). Voice quality changes following phonatory-
respiratory effort treatment (LSVT) versus respiratory effort treatment for individuals with
Parkinson disease. Journal of Voice, 15(1), 105-114.
Becker, G., Muller, A., Braune, S., Buttner, T., Benecke, R., Greulich, W., et al. (2002). Early
diagnosis of Parkinson's disease. Journal of Neurology, 249(Suppl. 3), 40-48.
Benschoter, R. A., Wittson, C. L., & Ingham, C. G. (1965). Teaching and consultation by
television: 1. Closed-circuit collaboration. Journal of Hospital and Community Psychiatry,
16(3), 99-100.
Benzeval, M., Judge, K., & Whitehead, M. (1999). Introduction. In M. Benzeval, K. Judge & M.
Whitehead (Eds.), Tackling inequalities in health (pp. 1-10). London: King's Fund
Publishing
Berardelli, A., Rothwell, J. C., Thompson, P. D., & Hallett, M. (2001). Pathophysiology of
bradykinesia in Parkinson's disease. Brain, 124, 2131-2146.
Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet, 1(8476), 307-310.
Blignault, I. (2000). Multipoint videoconferencing in health: A review of three years‟ experience in
Queensland, Australia. Telemedicine Journal, 6(2), 269-274.
Botsis, T., & Hartvigsen, G. (2008). Current status and future perspectives in telecare for elderly
people suffering from chronic diseases. Journal of Telemedicine and Telecare, 14(4), 195-
203.
Bower, J. H. (2000). Multiple system atrophy. In C. H. Adler & J. E. Ahlskog (Eds.), Parkinson's
disease and movement disorders: Diagnosis and treatment guidelines for the practicing
physician (pp. 235-242). Totowa, NJ: Humana Press.
Brennan, D. M., Georgeadis, A., & Baron, C. (2002). Telerehabilitation tools for the provision of
remote speech-language treatment. Topics in Stroke Rehabilitation, 8(4), 71-78.
210
Brennan, D. M., Georgeadis, A. C., Baron, C. R., & Barker, L. M. (2004). The effect of
videoconference-based telerehabilitation on story retelling performance by brain-injured
subjects and its implications for remote speech-language therapy. Telemedicine Journal and
e-Health, 10(2), 147-154.
Brignell, M., Wootton, R., & Gray, L. (2007). The application of telemedicine to geriatric medicine.
Age and Ageing, 36(4), 369-374.
Burgess, L. P., Holtel, M. R., Syms, M. J., Birkmire-Peters, D. P., Peters, L. J., & Mashima, P. A.
(1999). Overview of telemedicine applications for otolaryngology. Laryngoscope, 109(9),
1433-1437.
Canter, G. (1965). Speech characteristics of patients with Parkinson's disease. III. Articulation,
diadochokinesis and overall speech adequacy. Journal of Speech and Hearing Disorders,
30, 217-224.
Chae, Y. M., Park, H. J., Cho, J. G., Hong, G. D., & Cheon, K. A. (2000). The reliability and
acceptability of telemedicine for patients with schizophrenia in Korea. Journal of
Chan, K. M., & Yiu, E. M. (2002). The effect of anchors and training on the reliability of perceptual
voice evaluation. Journal of Speech, Language and Hearing Research, 45(1), 111-126.
Chenery, H. J., Murdoch, B. E., & Ingham, J. C. (1988). Studies in Parkinson's disease. I.
Perceptual speech analysis. Australian Journal of Human Communication Disorders, 16,
17-29.
Cicchetti, D. V., & Feinstein, A. R. (1990). High agreement but low kappa: II. Resolving the
paradoxes. Journal of Clinical Epidemiology, 43(6), 551-558.
Coates, C., & Bakheit, A. M. (1997). The prevalence of verbal communication disability in patients
with Parkinson's disease. Disability and Rehabilitation, 19(3), 104-107.
Cohn, R. J., & Goodenough, B. (2002). Health professionals' attitudes to videoconferencing in
paediatric health-care. Journal of Telemedicine and Telecare, 8, 274-282.
Collins, K., & O'Cathain, A. (2003). The continuum of patient satisfaction - from satisfied to very
satisfied. Social Science and Medicine 57, 2465-2470.
Colton, R. H., Casper, J. K., & Leonard, R. (2006). Understanding voice problems: A physiological
perspective for diagnosis and treatment (3rd ed.). Philadelphia: Lippincott Williams &
Wilkins.
Countryman, S., & Ramig, L. O. (1993). Effects of intensive voice therapy on voice deficits
associated with bilateral thalamotomy in Parkinson's disease: A case study. Journal of
Medical Speech-Language Pathology, 1(4), 233-250.
211
Countryman, S., Ramig, L. O., & Pawlas, A. A. (1994). Speech and voice deficits in Parkinsonian
plus syndromes: Can they be treated? Journal of Medical Speech-Language Pathology, 2(3),
221-225.
Craig, J. (1999). Introduction. In R. Wootton & J. Craig (Eds.), Introduction to telemedicine (pp. 3-
17). London: Royal Society of Medicine Press Ltd
Craig, J., & Patterson, V. (2006). Introduction to the practice of telemedicine. In R. Wootton, J.
Craig & V. Patterson (Eds.), Introduction to telemedicine (2nd ed., pp. 3-14). London:
Royal Society of Medicine Press Ltd.
Craig, J., Russell, C., Patterson, V., & Wootton, R. (1999). User satisfaction with realtime
teleneurology. Journal of Telemedicine and Telecare, 5, 237-241.
Critchley, E. M. (1981). Speech disorders of Parkinsonism: A review. Journal of Neurology
Neurosurgery and Psychiatry, 44, 751-758.
Crowe, B. L., & McDonald, I. G. (1997). Telemedicine in Australia: Recent developments. Journal
of Telemedicine and Telecare, 3, 188-193.
Dabul, B. (2000). Apraxia battery for adults (2nd ed.). Austin, Texas: Pro-ed.
Dallal, G. E. (1997). Randomization.com. Retrieved March 1, 2005, from
http://www.randomization.com
Danish, S. F., & Baltuch, G. H. (2007). History of deep brain stimulation. In G. H. Baltuch & M. B.
Stern (Eds.), Deep brain stimulation for Parkinson's disease (pp. 1-16). New York: Informa
Healthcare USA.
Darkins, A. W. (1996). The management of clinical risk in telemedicine applications. Journal of
Telemedicine and Telecare, 2, 179-184.
Darkins, A. W., & Cary, M. A. (2000). Telemedicine and telehealth: Principles, policies,
performance, and pitfalls. London: Free Association Books.
Darley, F. L., Aronson, A. E., & Brown, J. R. (1969a). Clusters of deviant speech dimensions in the
dysarthrias. Journal of Speech and Hearing Research, 12(3), 462-496.
Darley, F. L., Aronson, A. E., & Brown, J. R. (1969b). Differential diagnostic patterns of
dysarthria. Journal of Speech and Hearing Research, 12(2), 246-269.
Darley, F. L., Aronson, A. E., & Brown, J. R. (1975). Motor speech disorders. Philadelphia: W.B.
Saunders.
de Lau, L. M., & Breteler, M. M. (2006). Epidemiology of Parkinson's disease. Lancet Neurology,
5(6), 525-535.
de Pijper, J. R. (2007). Semitone conversions. Retrieved March 1, 2007, from
http://users.utu.fi/jyrtuoma/speech/semitone.html
212
de Rijk, M. C., Tzourio, C., Breteler, M. M., Dartigues, J. F., Amaducci, L., Lopez-Pousa, S., et al.
(1997). Prevalence of parkinsonism and Parkinson's disease in Europe: The
EUROPARKINSON collaborative study. European community concerted action on the
epidemiology of Parkinson's disease. Journal of Neurology, Neurosurgery and Psychiatry,
62(1), 10-15.
DePippo, K. L. (1994). Management of communication dysfunction. In A. M. Cohen & W. J.
Weiner (Eds.), The comprehensive management of Parkinson's disease (pp. 75-88). New
York: Demos.
Dewey, R. B. (2000). Clinical features of Parkinson's disease. In C. H. Adler & J. E. Ahlskog
(Eds.), Parkinson's disease and movement disorders: Diagnosis and treatment guidelines for
the practicing physician (pp. 71-84). Totowa, NJ: Humana Press.
Dillon, E., & Loermans, J. (2003). Telehealth in Western Australia: The challenge of evaluation.
Journal of Telemedicine and Telecare, 9(Suppl. 2), 15-19.
Dreyer, N., Dreyer, K. A., Shaw, D. K., & Wittman, P. P. (2001). Efficacy of telemedicine in
occupational therapy: A pilot study. Journal of Allied Health, 30(1), 39-42.
Dromey, C., & Ramig, L. O. (1998). Intentional changes in sound pressure level and rate: Their
impact on measures of respiration, phonation, and articulation. Journal of Speech, Language
and Hearing Research, 41(5), 1003-1018.
Dromey, C., Ramig, L. O., & Johnson, A. B. (1995). Phonatory and articulatory changes associated
with increased vocal intensity in Parkinson disease: A case study. Journal of Speech and
Hearing Research, 38(4), 751-764.
Duffy, J. R. (2000). Motor speech disorders: Clues to neurologic diagnosis. In C. H. Adler & J. E.
Ahlskog (Eds.), Parkinson's disease and movement disorders: Diagnosis and treatment
guidelines for the practicing physician (pp. 35-54). Totowa, NJ: Humana Press.
Duffy, J. R. (2005). Motor speech disorders: Substrates, differential diagnosis, and management
(2nd ed.). St. Louis, MO: Elsevier Mosby.
Duffy, J. R., Werven, G. W., & Aronson, A. E. (1997). Telemedicine and the diagnosis of speech
and language disorders. Mayo Clinic Proceedings, 72(12), 1116-1122.
Duvoisin, R. C., & Sage, J. (1996). Parkinson's disease: A guide for patient and family (4th ed.).
Philadelphia: Lippincott-Raven.
Eadie, T. L., & Doyle, P. C. (2002). Direct magnitude estimation and interval scaling of
pleasantness and severity in dysphonic and normal speakers. Journal of the Acoustical
Society of America, 112(6), 3014-3021.
213
El Sharkawi, A., Ramig, L. O., Logemann, J. A., Pauloski, B. R., Rademaker, A. W., Smith, C. H.,
et al. (2002). Swallowing and voice effects of Lee Silverman Voice Treatment (LSVT): A
pilot study. Journal of Neurology, Neurosurgery and Psychiatry, 72(1), 31-36.
Enderby, P. (1983). Frenchay dysarthria assessment. San Diego, CA: College-Hill Press.
Engen, T. (1971). Psychophysics II: Scaling methods. In J. W. Kling & L. Riggs (Eds.), Woodworth
and Schlossberg's experimental psychology (pp. 47-86). New York: Holt, Rinehart, &
Winston.
Fahn, S., & Przedborski, S. (2005). Parkinsonism. In L. P. Rowland (Ed.), Merritt's neurology (11th
ed., pp. 828-845). Philadelphia: Lippincott Williams & Wilkins.
Fairbanks, G. (1960). Voice and articulation drillbook. New York: Harper and Brothers.
Farrell, A., Theodoros, D., Ward, E. C., Hall, B., & Silburn, P. (2005). Effects of neurosurgical
management of Parkinson's disease on speech characteristics and oromotor function.
Journal of Speech, Language and Hearing Research, 48, 5-20.
Ferguson, E. W., Doarn, C. R., & Scott, J. C. (1995). Survey of global telemedicine. Journal of
Medical Systems, 19, 35-46.
Findley, L., Aujla, M., Bain, P. G., Baker, M., Beech, C., Bowman, C., et al. (2003). Direct
economic impact of Parkinson's disease: A research survey in the United Kingdom.
Movement Disorders, 18(10), 1139-1145.
Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.). New York: John Wiley
and Sons.
Foote, D. R. (1977). Satellite communication for rural health care in Alaska. Journal of
Communication, 27(4), 173-182.
Forrest, K., Weismer, G., & Turner, G. S. (1989). Kinematic, acoustic, and perceptual analyses of
connected speech produced by parkinsonian and normal geriatric adults. The Journal of the
Acoustical Society of America, 85(6), 2608-2622.
Fox, C. M., Morrison, C. E., Ramig, L. O., & Sapir, S. (2002). Current perspectives on the Lee
Silverman Voice Treatment (LSVT) for people with idiopathic Parkinson's disease.
American Journal of Speech-Language Pathology, 11, 111-123.
Fox, C. M., & Ramig, L. O. (1997). Vocal sound pressure level and self-perception of speech and
voice in men and women with idiopathic Parkinson disease. American Journal of Speech-
Language Pathology, 6(2), 85-94.
Fox, C. M., Ramig, L. O., Ciucci, M. R., Sapir, S., McFarland, D. H., & Farley, B. G. (2006). The
science and practice of LSVT/LOUD: Neural plasticity-principled approach to treating
individuals with Parkinson disease and other neurological disorders. Seminars in Speech and
Language, 27(4), 283-299.
214
Fuchs, M. (1979). Provider attitudes towards STARPAHC: A telemedicine project on the Papago
Reservation. Medical Care, 17(1), 59-68.
Garshnek, V. (1991). Applications of space communications technology to critical human needs:
Rescue, disaster relief, and medical assistance. Space Communications, 8, 311-317.
Garshnek, V., & Burkle, F. (1999a). Application of telemedicine and telecommunications to
disaster medicine: Historical and future perspectives. Journal of the American Medical
Informatics Society, 6(1), 26-37.
Garshnek, V., & Burkle, F. (1999b). Telecommunication systems in support of disaster medicine:
Applications of basic information pathways. Annals of Emergency Medicine, 34, 213-218.
Garshnek, V., & Hassell, L. (1997). The telemedicine frontier: Going the extra mile. Space Policy,
13(1), 37-46.
Gelfer, M. P. (1986). The stability of total phonational frequency range. Journal of the Acoustical
Society of America, 79(Suppl. 1), 83.
Geminiani, G., Cesana, B. M., Tamma, F., Contri, P., Pacchetti, C., Carella, F., et al. (1991).
Interobserver reliability between neurologists in training of Parkinson's disease rating scales.
A multicenter study. Movement Disorders, 6(4), 330-335.
Georgeadis, A., Brennan, D., Barker, L. M., & Baron, C. (2004). Telerehabilitation and its effect on
story retelling by adults with neurogenic communication disorders. Aphasiology, 18, 639-
652.
Gerratt, B. R., Kreiman, J., Antonanzas-Barroso, N., & Berke, G. S. (1993). Comparing internal and
external standards in voice quality judgments. Journal of Speech and Hearing Research,
36(1), 14-20.
Gescheider, G. A. (1976). Psychophysics, method and theory. Hillsdale, NJ: Lawrence Erlbaum.
Ghosh, G. J., McLaren, P. M., & Watson, J. P. (1997). Evaluating the alliance in videolink
teletherapy. Journal of Telemedicine and Telecare, 3(Suppl. 1), 33-35.
Goberman, A. M., & Elmer, L. W. (2005). Acoustic analysis of clear versus conversational speech
in individuals with Parkinson disease. Journal of Communication Disorders, 38, 215-230.
Goetz, C. G., Poewe, W., Rascol, O., & Sampaio, C. (2005). Evidence-based medical review
update: Pharmacological and surgical treatments for Parkinson's disease: 2001 to 2004.
Goetz, C. G., Stebbins, G. T., Chmura, T. A., Fahn, S., Klawans, H. L., & Marsden, C. D. (1995).
Teaching tape for the motor section of the unified Parkinson's disease rating scale.
Golbe, L. I. (1991). Young-onset Parkinson's disease: A clinical review. Neurology, 41, 168-173.
215
Goodglass, H., & Kaplan, E. (1972). The assessment of aphasia and related disorders.
Philadelphia: Lea & Febiger.
Greene, P. E. (2005). Progressive supranuclear palsy. In L. P. Rowland (Ed.), Merritt's neurology
(11th ed., pp. 846-848). Philadelphia: Lippincott Williams & Wilkins.
Grigsby, B. (2004). TRC report on US telemedicine activity. Kingston, NJ: Civic Research Institute.
Grimm, N., Paul, J., & Wakeham, A. (2004). Access to speech pathology services by people with
Parkinson's disease in Queensland, Australia. Unpublished master's thesis, University of
Queensland, Australia.
Guilfoyle, C., Wootton, R., Hassall, S., Offer, J., Warren, M., & Smith, D. (2003). Preliminary
experience of allied health assessments delivered face to face and by videoconference to a
residential facility for elderly people. Journal of Telemedicine and Telecare, 9, 230-233.
Gustke, S. S., Balch, D. C., West, V. L., & Rogers, L. O. (2000). Patient satisfaction with
telemedicine. Telemedicine Journal, 6(1), 5-13.
Halpern, C. H., & Hurtig, H. (2007). Deep brain stimulation of the subthalamic nucleus: Clinical
outcome. In G. H. Baltuch & M. B. Stern (Eds.), Deep brain stimulation for Parkinson's
disease (pp. 187-198). New York: Informa Healthcare USA.
Hamani, C., Richter, E., Schwalb, J. M., & Lozano, A. M. (2005). Bilateral subthalamic nucleus
stimulation for Parkinson's disease: A systematic review of the clinical literature.
Neurosurgery, 56, 1313-1324.
Hanson, D., Gerratt, B. R., & Ward, P. H. (1984). Cinegraphic observations of laryngeal function in
Parkinson's disease. Laryngoscope, 94, 348-353.
Hanson, W., & Metter, E. J. (1980). DAF as instrumental treatment for dysarthria in progressive
supranuclear palsy: A case report. Journal of Speech and Hearing Disorders, 45(2), 268-
276.
Harding, A. E. (1993). Movement disorders. In J. Walton (Ed.), Brain's diseases of the nervous
system (10th ed., pp. 393-401). Oxford: Oxford University Press.
Hartelius, L., & Svensson, P. (1994). Speech and swallowing symptoms associated with Parkinson's
disease and multiple sclerosis: A survey. Folia Phoniatrica et Logopaedica, 46(1), 9-17.
Helm-Estabrooks, N., & Ramsberger, G. (1986). Aphasia treatment delivered by telephone.
American Journal of Physical Medicine and Rehabilitation, 67(1), 51-53.
Hely, M. A., Chey, T., Wilson, A., Williamson, P. M., O'Sullivan, D. J., Rail, D., et al. (1993).
Reliability of the Columbia scale for assessing signs of Parkinson's disease. Movement
Disorders, 8(4), 466-472.
216
Heo, J., Lee, K., Paek, S. H., Kim, M., Lee, J., Kim, J., et al. (2008). The effects of bilateral
subthalamic nucleus deep brain stimulation (STN DBS) on cognition in Parkinson disease.
Journal of the Neurological Sciences, 273, 19-24.
Hicks, L. L., Boles, K. E., Hudson, S. T., Koenig, S., Madsen, R., Kling, B., et al. (2000). An
evaluation of satisfaction with telemedicine among health-care professionals. Journal of
Hill, A., Theodoros, D., Russell, T., & Ward, E. (2009a). Using telerehabilitation to assess apraxia
of speech in adults. International Journal of Language and Communication Disorders,
44(5), 731-737.
Hill, A., Theodoros, D., Russell, T., & Ward, E. (2009b). The redesign and re-evaluation of an
Internet-based telerehabilitation system for the assessment of dysarthria in adults.
Telemedicine Journal and e-Health, 15(9), 840-850.
Hill, A., Theodoros, D., Russell, T., Ward, E., & Wootton, R. (2008). The effects of aphasia
severity on the ability to assess language disorders via telerehabilitation. Aphasiology, iFirst,
1-16.
Hill, A., Theodoros, D. G., Russell, T. G., Cahill, L. M., Ward, E. C., & Clark, K. M. (2006). An
Internet-based telerehabilitation system for the assessment of motor speech disorders: A
pilot study. American Journal of Speech-Language Pathology, 15(1), 45-56.
Hillenbrand, J., Cleveland, R. A., & Erickson, R. L. (1994). Acoustic correlates of breathy vocal
quality. Journal of Speech and Hearing Research, 37(4), 769-778.
Hjelm, N. M. (2006). Benefits and drawbacks of telemedicine. In R. Wootton, J. Craig & V.
Patterson (Eds.), Introduction to telemedicine (2nd ed., pp. 135-151). London: Royal
Society of Medicine Press Ltd.
Ho, A. K., Bradshaw, J. L., & Iansek, R. (2000). Volume perception in Parkinsonian speech.
Ho, A. K., Bradshaw, J. L., Iansek, R., & Alfredson, R. (1999). Speech volume regulation in
Parkinson's disease: Effects of implicit cues and explicit instructions. Neuropsychologia,
37(13), 1453-1460.
Ho, A. K., Iansek, R., Marigliani, C., Bradshaw, J. L., & Gates, S. (1998). Speech impairment in a
large sample of patients with Parkinson's disease. Behavioural Neurology, 11, 131-137.
Hoehn, M., & Yahr, M. (1967). Parkinsonism: Onset, progression and mortality. Neurology, 17,
427-442.
Hoodin, R. B., & Gilbert, H. R. (1989a). Nasal airflows in Parkinsonian speakers. Journal of
Communication Disorders, 22, 169-180.
217
Hoodin, R. B., & Gilbert, H. R. (1989b). Parkinsonian dysarthria: An aerodynamic and perceptual
description of velopharyngeal closure in speech. Folia Phoniatrica et Logopedia, 41, 249-
258.
Hornsby, D., & Hudson, L. (1997). Videoconferencing for service delivery and professional
education. Australian Communication Quarterly, Spring, 25-26.
Howell, S., Tripoliti, E., & Pring, T. (2009). Delivering the Lee Silverman Voice Treatment
(LSVT) by web camera: A feasibility study. International Journal of Language and
Communication Disorders, 44(3), 287-300.
Huber, J. E., Stathopoulos, E. T., Ramig, L. O., & Lancaster, S. L. (2003). Respiratory function and
variability in individuals with Parkinson disease: Pre- and post-Lee Silverman Voice
Treatment. Journal of Medical Speech-Language Pathology, 11(4), 185-201.
Hughes, E. M. (2001). Communication skills for telehealth interactions. Home Healthcare Nurse,
19(8), 469-472.
Hurting, H. (2000). What is Parkinson's disease? Neuropathology, neurochemistry and
pathophysiology. In C. H. Adler & J. E. Ahlskog (Eds.), Parkinson's disease and movement
disorders: Diagnosis and treatment guidelines for the practicing physician (pp. 57-70).
Totowa, NJ: Humana Press.
Huse, D. M., Schulman, K., Orsini, L., Castelli-Haley, J., Kennedy, S., & Lenhart, G. (2005).
Burden of illness in Parkinson's disease. Movement Disorders, 20(11), 1449-1454.
Jankovic, J. (1989). Parkinsonism-plus syndromes. Movement Disorders, 4(Suppl. 1), 95-119.
Jennett, P. A., Hall, L. A., Hailey, D., Ohinmaa, A., Anderson, C., Thomas, R., et al. (2003). The
socio-economic impact on telehealth: A systematic review. Journal of Telemedicine and
Telecare, 9, 311-320.
Jones, B., Jarvis, P., & Ebbutt, A. F. (1996). Trials to assess equivalence: The importance of
rigorous methods. British Medical Journal, 3, 36-39.
Kearns, K. P., & Simmons, N. N. (1988). Interobserver reliability and perceptual ratings: More than
meets the ear. Journal of Speech and Hearing Research, 31(1), 131-136.
Kent, R. D., & Rosenbek, J. C. (1982). Prosodic disturbance and neurological lesion. Brain and
Language, 15, 259-291.
King, J. B., Ramig, L., O., Lemke, J., H., & Horii, Y. (1994). Parkinson's disease: Longitudinal
changes in acoustic parameters of phonation. Journal of Medical Speech-Language
Pathology, 2, 29-42.
Kluin, K. J., Foster, N. L., Berent, S., & Gilman, S. (1993). Perceptual analysis of speech disorders
in progressive supranuclear palsy. Neurology, 43(3 Pt 1), 563-566.
218
Kolitsi, Z., & Iakovidis, I. (2000). Improving user acceptance of health-care telematics. Journal of
Telemedicine and Telecare, 6(Suppl. 2), 37-39.
Krauss, J. K., & Grossman, R. G. (2007). Surgery for Parkinson's disease and hyperkinetic
movement disorders. In J. Jankovic & E. Tolosa (Eds.), Parkinson's disease and movement
disorders (5 ed., pp. 620-641). Philadelphia: PA: Lippincott Williams & Wilkins.
Kreiman, J., & Gerratt, B. R. (1998). Validity of rating scale measures of voice quality. Journal of
the Acoustical Society of America, 104(3 Pt 1), 1598-1608.
Kreiman, J., Gerratt, B. R., Kempster, G. B., Erman, A., & Berke, G. S. (1993). Perceptual
evaluation of voice quality: Review, tutorial, and a framework for future research. Journal
of Speech and Hearing Research, 36(1), 21-40.
Kully, D. (2000). Telehealth in speech pathology: Applications to the treatment of stuttering.
Kuo, R. L., Delvecchio, F. C., Babayan, R. K., & Preminger, G. M. (2001). Telemedicine: Recent
developments and future applications. Journal of Endourology, 15(1), 63-66.
Lalor, E., Brown, M., & Cranfield, E. (2000). Telemedicine: Its role in speech and language
management for rural and remote patients. ACQ: Issues In Language, Speech and Hearing,
2(2), 54-55.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data.
Biometrics, 33(1), 159-174.
Leeder, S. (2000). The Australian health system. Telemedicine Journal, 6(2), 201-204.
Lewis, S. J., Dove, A., Robbins, T. W., Baker, R. A., & Owen, A. M. (2003). Cognitive
impairments in early Parkinson's disease are accompanied by reductions in activity in
frontostriatal neural circuitry. The Journal of Neuroscience, 23(15), 6351-6356.
LeWitt, P. A. (2000). Parkinson's disease: Etiologic considerations. In C. H. Adler & J. E. Ahlskog
Liotti, M., Ramig, L. O., Vogel, D., New, P., Cook, C. I., Ingham, R. J., et al. (2003). Hypophonia
in Parkinson's disease: Neural correlates of voice treatment revealed by PET. Neurology,
60(3), 432-440.
Llewellyn, C. H. (1995). The role of telemedicine in disaster medicine. Journal of Medical Systems,
19, 29-34.
Logemann, J. A., & Fisher, H. (1981). Vocal tract control in Parkinson's disease: Phonetic feature
analysis of misarticulations. Journal of Speech and Hearing Disorders, 46, 348-352.
219
Logemann, J. A., Fisher, H., Boshes, B., & Blonsky, E. (1978). Frequency and concurrence of vocal
tract dysfunctions in the speech of a large sample of Parkinson individuals. Journal of
Speech and Hearing Disorders, 42, 47-57.
Lovell, N. H., & Celler, B. G. (1999). Information technology in primary health care. International
Journal of Medical Informatics, 55, 9-22.
LSVT Global: Rehabilitative therapy for people with Parkinson's disease. (2010). Retrieved March
23, 2010, from http://www.lsvtglobal.com/index.php?action=results
Ludlow, C., & Bassich, C. (1983). The results of acoustic and perceptual assessment of two types of
dysarthria. In W. R. Berry (Ed.), Clinical Dysarthria (pp. 121-147). San Diego, CA:
College-Hill Press.
Ludlow, C., & Bassich, C. (1984). Relationships between perceptual ratings and acoustic measures
of hypokinetic speech. In M. R. McNeil, J. C. Rosenbek & A. E. Aronson (Eds.), The
dysarthrias: Physiology, acoustics, perception, management (pp. 163-195). San Diego, CA:
College-Hill Press.
Mackenzie, C., & Lowit, A. (2007). Behavioural intervention effects in dysarthria following stroke:
Communication effectiveness, intelligibility and dysarthria impact. International Journal of
Language and Communication Disorders, 42(2), 131-153.
Maheu, M. M., Whitten, P., & Allen, A. (2001). E-health, telehealth, and telemedicine: A guide to
start-up and success. San Francisco: Jossey-Bass.
Mair, F., Whitten, P., May, C., & Doolittle, G. C. (2000). Patients' perceptions of a telemedicine
specialty clinic. Journal of Telemedicine and Telecare, 6, 36-40.
Malagodi, M., & Smith, S. (1999). Prospective role for telemedicine as a communication tool for
rural rehabilitation practice. Work, 12, 245-259.
Marsden, C. D. (1989). Slowness of movement in Parkinson's disease. Movement Disorders,
4(Suppl. 1), 26-37.
Marsden, C. D., & Obeso, J. A. (1994). The functions of the basal ganglia and the paradox of
stereotaxic surgery in Parkinson's disease. Brain, 117, 877-897.
Martinez-Martin, P., Gil-Nagel, A., Gracia, L. M., Gomez, J. B., Martinez-Sarries, J., & Bermejo,
F. (1994). Unified Parkinson's Disease Rating Scale characteristics and structure. The
Cooperative Multicentric Group. Movement Disorders, 9(1), 76-83.
Mashima, P. A., Birkmire-Peters, D. P., Holtel, M. R., & Syms, M. J. (1999). Telehealth
applications in speech-language pathology. Journal of Healthcare Information Management,
13(4), 71-78.
220
Mashima, P. A., Birkmire-Peters, D. P., Syms, M. J., Holtel, M. R., Burgess, L. P., & Peters, L. J.
(2003). Telehealth: Voice therapy using telecommunications technology. American Journal
of Speech-Language Pathology, 12(4), 432-439.
Matthews, J. N. (2000). Introduction to randomized controlled clinical trials. London: Arnold.
McCullough, A. (2001). Viability and effectiveness of teletherapy for pre-school children with
special needs. International Journal of Language and Communication Disorders, 36(Suppl.
1), 321-326.
Mechanic, D. (1998). Public trust and initiatives for new health care partnerships. Milbank
Quarterly, 76, 281-302.
Mekhjian, H., Turner, J. W., Gailiun, M., & McCain, T. A. (1999). Patient satisfaction with
telemedicine in a prison environment. Journal of Telemedicine and Telecare, 5(1), 55-61.
Metter, E. J., & Hanson, W. R. (1991). Dysarthria in progressive supranuclear palsy. In C. A.
Moore & K. M. Yorkston (Eds.), Dysarthria and apraxia of speech: Perspective on
management (pp. 127-136). Baltimore: Paul H. Brookes.
Miller, E. (2001). Telemedicine and doctor-patient communication: An analytical survey of the
literature. Journal of Telemedicine and Telecare, 7, 1-17.
Miller, N., Allcock, L., Jones, D., Noble, E., Hildreth, A. J., & Burn, D. J. (2007). Prevalence and
pattern of perceived intelligibility changes in Parkinson's disease. Journal of Neurology,
Miller, N., Noble, E., Jones, D., Allcock, L., & Burn, D. J. (2008). How do I sound to me?
Perceived changes in communication in Parkinson's disease. Clinical Rehabilitation, 22(1),
14-22.
Miller, N., Noble, E., Jones, D., & Burn, D. (2006). Life with communication changes in
Parkinson's disease. Age and Ageing, 35(3), 235-239.
Mitchell, J. G. (1999). The uneven diffusion of telemedicine services in Australia. Journal of
Montani, C., Billaud, N., Couturier, P., Fluchaire, I., Lemaire, R., Malterre, C., et al. (1996).
"Telepsychometry": A remote psychometry consultation in clinical gerontology: preliminary
study. Telemed J, 2(2), 145-150.
Montani, C., Billaud, N., Tyrrel, J., Fluchaire, I., Malterre, C., Lauvernay, N., et al. (1997).
Psychological impact of a remote psychometric consultation with hospitalized elderly
people. Journal of Telemedicine and Telecare, 3, 140-145.
Morrant, D. G., & Shaw, W. C. (1996). Use of standardized video recordings to assess cleft surgery
outcome. Cleft Palate Craniofacial Journal, 33(2), 134-142.
221
Mourao, L. F., Aguiar, P. M., Ferraz, F. A., Behlau, M. S., & Ferraz, H. B. (2005). Acoustic voice
assessment in Parkinson's disease patients submitted to posteroventral pallidotomy.
Arquivos de Neuropsiquiatria, 63(1), 20-25.
Mun, S. K., & Turner, J. W. (1999). Telemedicine: Emerging e-medicine. Annual Review of
Biomedical Engineering, 1, 589-510.
Murdoch, B. E. (1998). The neuroanatomical framework of dysarthria. In B. E. Murdoch (Ed.),
Dysarthria: A physiological approach to assessment and treatment (pp. 1-36). Cheltenham,
England: Stanley Thornes.
Murdoch, B. E. (2001). Subcortical brain mechanisms in speech and language. Folia Phoniatrica et
Logopaedica, 53(5), 233-251.
Murdoch, B. E., Chenery, H. J., Bowler, S., & Ingram, J. C. (1989). Respiratory function in
Parkinson's subjects exhibiting a perceptible speech deficit: A kinematic and spirometric
analysis. Journal of Speech and Hearing Disorders, 54, 610-626.
Murdoch, B. E., Manning, C. Y., Theodoros, D. G., & Thompson, E. C. (1997). Laryngeal and
phonatory dysfunction in Parkinson's disease. Clinical Linguistics and Phonetics, 11(3),
245-266.
Murphy, R. L. H., & Bird, K. T. (1974). Telediagnosis: A new community health resource;
observations on the feasibility of telediagnosis based on 1000 patient transactions. American
Journal of Public Health, 64(2), 113-119.
Mutch, W. J., Dingwall-Fordyce, I., Downie, A. W., Paterson, J. G., & Roy, S. K. (1986).
Parkinson's disease in a Scottish city. British Medical Journal (Clinical Research Edition),
292(6519), 534-536.
Muthane, U. B., Swamy, H. S., Satishchandra, P., Subhash, M. N., Rao, S., & Subbakrishna, D.
(1994). Early onset Parkinson's disease: Are juvenile- and young-onset different? Movement
Disorders, 9(5), 539-544.
Myers, C. (2005). Telehealth applications in head and neck oncology. Journal of Speech-Language
Pathology and Audiology, 29(3), 125-129.
O'Brian, S., Packman, A., & Onslow, M. (2008). Telehealth delivery of the Camperdown program
for adults who stutter: A phase 1 trial. Journal of Speech, Language and Hearing Research,
51(1), 184-195.
O'Shannessy, L. (2000). Using the law to enhance provision of telemedicine. Journal of
Okun, M. S., Fernandes, H. H., Wu, S. S., Kirsch-Darrow, L., Bowers, D., Bova, F., et al. (2009).
Cognitive and mood in Parkinson's disease in subthalamic nucleus versus globus pallidus
interna deep brain stimulation: The COMPARE trial. Annals of Neurology, 65(5), 586-595.
222
Onor, M. L., Trevisiol, M., Urciuoli, O., Misan, S., Bertossi, F., Tirone, G., et al. (2008).
Effectiveness of telecare in elderly populations-a comparison of three settings. Telemedicine
Journal and e-Health, 14(2), 164-169.
Owen, A. M. (2004). Cognitive dysfunction in Parkinson's disease: The role of frontostriatal
circuitry. The Neuroscientist, 10(6), 525-537.
Owens, D. J., & Batchelor, C. (1996). Patient satisfaction and the elderly. Social Science and
Medicine, 42, 1483-1491.
Oxtoby, M. (1982). Parkinson's disease patients and their social needs. London: Parkinson's
Disease Society.
Palsbo, S. E. (2007). Equivalence of functional communication assessment in speech pathology
using videoconferencing. Journal of Telemedicine and Telecare, 13(1), 40-43.
Perednia, D. A., & Allen, A. (1995). Telemedicine technology and clinical applications. Journal of
the American Medical Association, 273(6), 483-488.
Perez, K. S., Ramig, L. O., Smith, M. E., & Dromey, C. (1996). The Parkinson larynx: Tremor and
videostroboscopic findings. Journal of Voice, 10(4), 354-361.
Perlman, A. L., & Witthawaskul, W. (2002). Real-time remote telefluoroscopic assessment of
patients with dysphagia. Dysphagia, 17(1), 162-167.
Peto, V., Fitzpatrick, R., & Jenkinson, C. (1997). Self-reported health status and access to health
services in a community sample with Parkinson's disease. Disability and Rehabilitation,
19(3), 97-103.
Piaggio, G., Elbourne, D. R., Atlman, D. G., Pocock, S. J., & Evans, S. J. W. (2006). Reporting of
noninferiority and equivalence randomized trials. Journal of the American Medical
Association, 295(10), 1152-1160.
Przedborski, S. (2007). Etiology and pathogenesis of Parkinson's disease. In J. Jankovic & E.
Tolosa (Eds.), Parkinson's disease and movement disorders (5th ed., pp. 77-92).
Philadelphia: Lippincott Williams & Wilkins.
Quinn, N. (1989). Multiple system atrophy - the nature of the beast. Journal of Neurology,
Neurosurgery and Psychiatry, 52(Suppl.), 78-89.
Qureshi, A. A., & Kvedar, J. C. (2003). Patient knowledge and attitude toward information
technology and teledermatology: Some tentative findings Telemedicine Journal and e-
Health, 9(3), 259-264.
Rabinov, C. R., Kreiman, J., Gerratt, B. R., & Bielamowicz, S. (1995). Comparing reliability of
perceptual ratings of roughness and acoustic measure of jitter. Journal of Speech and
Hearing Research, 38(1), 26-32.
223
Ramig, L. O. (1998). Treatment of speech and voice problems associated with Parkinson's disease.
Topics in Geriatric Rehabilitation, 14(2), 28-43.
Ramig, L. O., Bonitati, C. M., Lemke, J. H., & Horii, Y. (1994). Voice treatment for patients with
Parkinson disease: Development of an approach and preliminary efficacy data. Journal of
Ramig, L. O., Countryman, S., O'Brien, C., Hoehn, M., & Thompson, L. (1996). Intensive speech
treatment for patients with Parkinson's disease: Short-and long-term comparison of two
techniques. Neurology, 47(6), 1496-1504.
Ramig, L. O., Countryman, S., Thompson, L. L., & Horii, Y. (1995). Comparison of two forms of
intensive speech treatment for Parkinson disease. Journal of Speech and Hearing Research,
38(6), 1232-1251.
Ramig, L. O., & Dromey, C. (1996). Aerodynamic mechanisms underlying treatment-related
changes in SPL in patients with Parkinson disease. Journal of Speech and Hearing
Research, 39, 798-807.
Ramig, L. O., & Fox, C. (2004). LSVT® training and certification workshop binder. Tucson, AZ:
LSVT Foundation.
Ramig, L. O., Fox, C., & Sapir, S. (2004). Parkinson's disease: Speech and voice disorders and their
treatment with the Lee Silverman Voice Treatment. Seminars in Speech and Language,
25(2), 169-180.
Ramig, L. O., Pawlas, A. A., & Countryman, S. (1995). The Lee Silverman Voice Treatment
(LSVT®): A practical guide to treating the voice and speech disorders in Parkinson disease.
Iowa City: National Centre for Voice and Speech.
Ramig, L. O., Sapir, S., Countryman, S., Pawlas, A. A., O'Brien, C., Hoehn, M., et al. (2001).
Intensive voice treatment (LSVT) for patients with Parkinson's disease: A 2 year follow up.
Journal of Neurology, Neurosurgery and Psychiatry, 71(4), 493-498.
Ramig, L. O., Sapir, S., Fox, C., & Countryman, S. (2001). Changes in vocal loudness following
intensive voice treatment (LSVT) in individuals with Parkinson's disease: A comparison
with untreated patients and normal age-matched controls. Movement Disorders, 16(1), 79-
83.
Rey, J. M., Plapp, J. M., Stewart, G., Richards, I., & Bashir, M. (1987). Reliability of the
psychosocial axes of DSM-III in an adolescent population. The British Journal of
Psychiatry, 150, 228-234.
Ritter, K., Trotman, C. A., & Phillips, C. (2002). Validity of subjective evaluations for the
assessment of lip scarring and impairment. The Cleft Palate-Craniofacial Journal, 39(6),
587-596.
224
Roberts, R., Rigby, M., & Birch, K. (2000). Telematics in healthcare: New paradigm, new issues. In
M. Rigby, R. Roberts & M. Thick (Eds.), Taking health telematics into the 21st century (pp.
1-16). Abingdon: Radcliffe Medical Press Ltd.
Robertson, S. J. (1988). Parkinson's disease: Does speech therapy work and defining the problems.
Speech Therapy in Practice, 4, 18-19.
Rosen, M. J. (1999). Telerehabilitation. NeuroRehabilitation, 12, 11-26.
Russell, T. G. (2004). Establishing the efficacy of telemedicine as a clinical tool for
physiotherapists: From system design to randomised controlled trial. Unpublished doctoral
thesis, The University of Queensland, Brisbane, Australia.
Russell, T. G., Buttrum, P., Wootton, R., & Jull, G. A. (2003). Low-bandwidth telerehabilitation for
patients who have undergone total knee replacement: Preliminary results. Journal of
Salvemini, A. V. (1999). Challenges for user-interface designers of telemedicine systems.
Telemedicine Journal, 5(2), 163-169.
Samlan, R. A., & Weismer, G. (1995). The relationship of selected perceptual measures of
diadochokinesis to speech intelligibility in dysarthric speakers with amyotrophic lateral
sclerosis. American Journal of Speech-Language Pathology, 4(2), 9-13.
Sapir, S., Pawlas, A. A., Ramig, L. O., Countryman, S., O'Brien, C., Hoehn, M., et al. (2001). Voice
and speech abnormalities in Parkinson disease: Relation to severity of motor impairment,
duration of disease, medication, depression, gender, and age. Journal of Medical Speech-
Sapir, S., Ramig, L. O., Hoyt, P., Countryman, S., O'Brien, C., & Hoehn, M. (2002). Speech
loudness and quality 12 months after intensive voice treatment (LSVT) for Parkinson's
disease: A comparison with an alternative speech treatment. Folia Phoniatrica et
Logopaedica, 54(6), 296-303.
Sapir, S., Spielman, J. L., Ramig, L. O., Story, B. H., & Fox, C. (2007). Effects of intensive voice
treatment (the Lee Silverman Voice Treatment [LSVT]) on vowel articulation in dysarthric
individuals with idiopathic Parkinson disease: Acoustic and perceptual findings. Journal of
Speech, Language, and Hearing Research, 50(4), 899-912.
Schiavetti, N., Metz, D. E., & Sitler, R. W. (1981). Construct validity of direct magnitude
estimation and interval scaling of speech intelligibility: Evidence from a study of the hearing
impaired. Journal of Speech and Hearing Research, 24(3), 441-445.
Schiavetti, N., Sacco, P. R., Metz, D. E., & Sitler, R. W. (1983). Direct magnitude estimation and
interval scaling of stuttering severity. Journal of Speech and Hearing Research, 26(4), 568-
573.
225
Schliesser, H. F. (1985). Psychological scaling of speech by students in training compared to that by
experienced speech-language pathologists. Perceptual and Motor Skills, 61, 1299-1302.
Schrag, A., Jahanshahi, M., & Quinn, N. (2000). How does Parkinson's disease affect quality of
life? A comparison with quality of life in the general population. Movement Disorders,
15(6), 1112-1118.
Schulz, G. M., & Grant, M. K. (2000). Effects of speech therapy and pharmacological and surgical
treatments on voice and speech in Parkinson's disease: A review of the literature. Journal of
Communication Disorders, 33, 59-88.
Schulz, G. M., Greer, M., & Friedman, W. (2000). Changes in vocal intensity in Parkinson's disease
following pallidotomy surgery. Journal of Voice, 14(4), 589-606.
Schulz, G. M., Peterson, T., Sapienza, C. M., Greer, M., & Friedman, W. (1999). Voice and speech
characteristics of persons with Parkinson's disease pre- and post-pallidotomy surgery:
Preliminary findings. Journal of Speech, Language and Hearing Research, 42, 1176-1194.
Sclafani, A. P., Heneghan, C., Ginsburg, J., Sabini, P., Stern, J., & Dolitsky, J. N. (1999).
Teleconsultation in otolaryngology: Live versus store and forward consultations.
Otolaryngology - Head and Neck Surgery, 120(1), 62-72.
Scott, S. (1991). Speech therapy. In F. L. Caird (Ed.), Rehabilitation in Parkinson's disease (pp. 87-
106). London: Chapman and Hall.
Scott, S., & Caird, F. I. (1983). Speech therapy for Parkinson's disease. Journal of Neurology,
Scott, S., Caird, F. I., & Williams, B. O. (1985). Communication in Parkinson's disease. London:
Croom Helm.
Shannon, K. M. (2000). Surgical treatment of Parkinson's disease. In C. H. Adler & J. E. Ahlskog
Sheard, C., Adams, R. D., & Davis, P. J. (1991). Reliability and agreement of ratings of ataxic
dysarthric speech samples with varying intelligibility. Journal of Speech and Hearing
Research, 34(2), 285-293.
Shores, M. M., Ryan-Dykes, P., Williams, R. M., Mamerto, B., Sadak, T., Pascualy, M., et al.
(2004). Identifying undiagnosed dementia in residential care veterans: Comparing
telemedicine to in-person clinical examination. International Journal of Geriatric
Psychiatry, 19(2), 101-108.
Smith, M. C., Smith, M. K., & Ellgring, H. (1996). Spontaneous and posed facial expression in
Parkinson's disease. Journal of the International Neuropsychological Society, 2(5), 383-391.
226
Smith, M. E., Ramig, L. O., Dromey, C., Perez, K. S., & Samandari, R. (1995). Intensive voice
treatment in Parkinson disease: Laryngostroboscopic findings. Journal of Voice, 9(4), 453-
459.
Solomon, N. P., & Hixon, T. J. (1993). Speech breathing in Parkinson's disease. Journal of Speech
and Hearing Research, 36, 294-310.
Solomon, N. P., Robin, D. A., & Luschei, E. S. (2000). Strength, endurance, and stability of the
tongue and hand in Parkinson disease. Journal of Speech, Language and Hearing Research,
43(1), 256-267.
Spielman, J. L., Borod, J. C., & Ramig, L. O. (2003). The effects of intensive voice treatment on
facial expressiveness in Parkinson disease: Preliminary data. Cognitive and behavioral
neurology, 16(3), 177-188.
Spielman, J. L., Ramig, L. O., & Borod, J. C. (2001). Preliminary effects of voice therapy on facial
expression in Parkinson's disease. Journal of the International Neuropsychological Society,
7, 244.
Spielman, J. L., Ramig, L. O., Mahler, L., Halpern, A., & Gavin, W. J. (2007). Effects of an
extended version of the Lee Silverman Voice Treatment on voice and speech in Parkinson's
disease. American Journal of Speech-Language Pathology, 16(2), 95-107.
Stacy, M. (2000). Progressive Supranuclear Palsy. In C. H. Adler & J. E. Ahlskog (Eds.),
Parkinson's disease and movement disorders: Diagnosis and treatment guidelines for the
practicing physician (pp. 229-234). Totowa, NJ: Humana Press.
Stanberry, B. (1998a). The legal and ethical aspects of telemedicine. London: Royal Society of
Medicine Press Ltd.
Stanberry, B. (1998b). The legal and ethical aspects of telemedicine. 3: Telemedicine and
malpractice. Journal of Telemedicine and Telecare, 4, 72-79.
Stevens, S. S. (1974). Perceptual magnitude and its measurement. In E. C. Caterette & M. P.
Friedman (Eds.), Handbook of Perception (vol. 2, pp. 22-40). New York: Academic Press.
Stevens, S. S. (1975). Psychophysics: Introduction to its perceptual, neural, and social prospects.
New York: John Wiley and Sons.
Stewart, C. (2000). Speech-language therapy for patients with Parkinson's disease. In L. Cote, L. L.
Sprinzeles, R. Elliott & A. H. Kutscher (Eds.), Parkinson's disease and quality of life (pp.
151-156). New York: Haworth Press.
Theodoros, D. G., Hill, A., Russell, T., Ward, E., & Wootton, R. (2008). Assessing acquired
language disorders in adults via the Internet. Telemedicine Journal and e-Health, 14(6),
552-559.
227
Theodoros, D. G., & Murdoch, B. E. (1998). Hypokinetic dysarthria. In B. E. Murdoch (Ed.),
Dysarthria: A physiological approach to assessment and treatment (pp. 266-313).
Cheltenham, England: Stanley Thornes.
Theodoros, D. G., Murdoch, B. E., & Thompson, E. C. (1995). Hypernasality in Parkinson's
disease: A perceptual and physiological analysis. Journal of Medical Speech-Language
Pathology, 3(2), 73-84.
Theodoros, D. G., & Russell, T. (2008). Telerehabilitation: Current perspectives. Studies in Health
Technology and Informatics, 131, 191-209.
Theodoros, D. G., Russell, T. G., Hill, A., Cahill, L., & Clark, K. (2003). Assessment of motor
speech disorders online: A pilot study. Journal of Telemedicine and Telecare, 9(Suppl. 2),
66-68.
Theodoros, D. G., Thompson-Ward, E. C., Murdoch, B. E., Lethlean, J., & Silburn, P. (1999). The
effects of the Lee Silverman Voice Treatment program on motor speech function in
Parkinson disease following thalamotomy and pallidotomy surgery: A case study. Journal of
Tickle-Degnen, L., & Lyons, K. D. (2004). Practitioners' impressions of patients with Parkinson's
disease: The social ecology of the expressive mask. Social Science and Medicine (1982),
58(3), 603-614.
Tindall, L. R., Huebner, R. A., Stemple, J. C., & Kleinert, H. L. (2008). Videophone-delivered
voice therapy: A comparative analysis of outcomes to traditional delivery for adults with
Parkinson's disease. Telemedicine Journal and e-Health, 14(10), 1070-1077.
Tolosa, E., & Katzenschlager, R. (2007). Pharmacological management of Parkinson's disease. In J.
Jankovic & E. Tolosa (Eds.), Parkinson's disease and movement disorders (5th ed., pp. 110-
145). Philadelphia: Lippincott Williams & Wilkins.
Toner, M. A., & Emanuel, F. W. (1989). Direct magnitude estimation and equal appearing interval
scaling of vowel roughness. Journal of Speech and Hearing Research, 32(1), 78-82.
Toner, M. A., Emanuel, F. W., & Parker, D. (1990). Relationship of spectral noise levels to
psychophysical scaling of vowel roughness. Journal of Speech and Hearing Research,
33(2), 238-244.
Vasquez, M. S. (2008). Down to the fundamentals of telehealth and home healthcare nursing. Home
Healthcare Nurse, 26(5), 280-287.
Vaughn, G. R. (1976). Tel-communicology: Health-care delivery system for persons with
communicative disorders. American Speech-Language-Hearing Association, 18(1), 13-17.
228
Vestal, L., Smith-Olinde, L., Hicks, G., Hutton, T., & Hart, J. (2006). Efficacy of language
assessment in Alzheimer's disease: Comparing in-person examination and telemedicine.
Clinical Interventions in Aging, 1(4), 467-471.
Wagner, M. J. (1994). Introductory musical acoustics (3rd ed.). Raleigh, NC: Contemporary
Publishing Company of Raleigh Inc.
Waite, L. M. (2000). Accommodating Parkinson's disease: A review of the perspective of the
caregiver and the Parkinsonian. In L. Cote, L. L. Sprinzeles, R. Elliott & A. H. Kutscher
(Eds.), Parkinson's disease and quality of life (pp. 173-188). New York: Haworth Press.
Waite, M., Cahill, L. M., Theodoros, D. G., Busuttin, S., & Russell, T. G. (2006). A pilot study of
online assessment of childhood speech disorders. Journal of Telemedicine and Telecare,
12(Suppl. 3), 92-94.
Walter, B. L., & Vitek, J. L. (2004). Surgical treatment for Parkinson's disease. Lancet Neurology,
3, 719-728.
Ward, E. C., Crombie, J., Trickey, M., Hill, A., Theodoros, D., & Russell, T. (2009). Assessment of
communication and swallowing post laryngectomy: A telerehabilitation trial. Journal of
Ward, E. C., Theodoros, D. G., Murdoch, B. E., & Silburn, P. (2000). Changes in maximum
capacity tongue function following the Lee Silverman Voice Treatment Program. Journal of
Ward, E. C., White, J., Russell, T., Theodoros, D., Kuhl, M., Nelson, K., et al. (2007). Assessment
of communication and swallowing function post laryngectomy: A telerehabilitation trial.
Waters, C. W. (2005). Diagnosis and management of Parkinson's disease (4th ed.). NY:
Professional Communications.
Weismer, G., Jeng, J. Y., Laures, J. S., Kent, R. D., & Kent, J. F. (2001). Acoustic and intelligibility
characteristics of sentence production in neurogenic speech disorders. Folia Phoniatrica et
Logopaedica, 53(1), 1-18.
Weismer, G., & Laures, J. S. (2002). Direct magnitude estimates of speech intelligibility in
dysarthria: Effects of a chosen standard. Journal of Speech, Language and Hearing
Research, 45(3), 421-433.
Werner, P., & Karnieli, E. (2003). A model of the willingness to use telemedicine for routine and
specialised care. Journal of Telemedicine and Telecare, 9, 264-272.
Wertz, R. T., Dronkers, N. F., Berstein-Ellis, E., Shubitowski, R. E., Shenaut, G. K., & Knight, R.
T. (1987). Appraisal and diagnosis of neurogenic communication disorders in remote
settings. Clinical Aphasiology, 17, 117-123.
229
Wertz, R. T., Dronkers, N. F., Berstein-Ellis, E., Sterling, L. K., Shubitowski, R. E., Shenaut, G. K.,
et al. (1992). Potential of telephonic and television technology for appraising and diagnosing
neurogenic communication disorders in remote settings. Aphasiology, 6, 195-202.
Whitehill, T. L., Lee, A. S., & Chun, J. C. (2002). Direct magnitude estimation and interval scaling
of hypernasality. Journal of Speech, Language and Hearing Research, 45(1), 80-88.
Whitten, P., Frances, M., & Collins, B. (1997). Home telenursing in Kansas: Patient's perceptions of
users and benefits Journal of Telemedicine and Telecare, 3(Suppl. 1), 67-69.
Williams, T. L., May, C. R., & Esmail, A. (2001). Limitations of patient satisfaction studies in
telehealthcare: A systematic review of the literature. Telemedicine Journal and e-Health,
7(4), 293-316.
Wilson, L., Lincoln, M., & Onslow, M. (2002). Availability, access, and quality of care: Inequities
in rural speech pathology services for children and a model for redress. Advances in Speech-
Wootton, R., Blignault, I., & Cignoli, J. (2003). A national survey of telehealth activity in
Australian hospitals. Journal of Telemedicine and Telecare, 9(Suppl. 2), 73-75.
World Health Organisation (WHO). (1997a). Health-for-all policy for the 21st century. Geneva:
WHO.
World Health Organisation (WHO). (1997b). Telehealth and Telemedicine will henceforth be part
of the strategy for Health-for-All. Retrieved February 1, 2006 from
http://www.who.int/inf-pr-1997/en/pr97-98.html
Yellowlees, P. M. (1997). Successful development of telemedicine systems - seven core principles.
Journal of Telemedicine and Telecare, 3(4), 215-222.
Yellowlees, P. M. (2006). Successfully developing a telemedicine system. In R. Wootton, J. Craig
& V. Patterson (Eds.), Introduction to telemedicine (2nd ed., pp. 93-100). London: Royal
Society of Medicine Press Ltd.
Yellowlees, P. M., & Brooks, P. M. (1999). Health online: The future isn't what it used to be The
Medical Journal of Australia, 171, 522-525.
Yorkston, K. M., & Beukelman, D. R. (1980). A clinician-judged technique for quantifying
dysarthric speech based on single-word intelligibility. Journal of Communication Disorders,
13(1), 15-31.
Yorkston, K. M., & Beukelman, D. R. (1981a). Assessment of intelligibility of dysarthric speech.
Austin, TX: Pro-Ed.
Yorkston, K. M., & Beukelman, D. R. (1981b). Communication efficiency of dysarthric speakers as
measured by sentence intelligibility and speaking rate. Journal of Speech and Hearing
Disorders, 46(3), 296-301.
230
Zangaglia, R., Pacchetti, C., Pasotti, C., Mancini, F., Servello, D., Sinforiani, E., et al. (2009). Deep
brain stimulation and cognitive functions in Parkinson's disease: A three-year controlled
study. Movement Disorders, 11, 1621-1628.
Zarate, C. A., Jr., Weinstock, L., Cukor, P., Morabito, C., Leahy, L., Burns, C., et al. (1997).
Applicability of telemedicine for assessing patients with schizophrenia: acceptance and
reliability. Journal of Clinical Psychiatry, 58(1), 22-25.
Zgaljardic, D. J., Borod, J. C., Foldi, N. S., & Mattis, P. (2003). A review of the cognitive and
behavioral sequelae of Parkinson's disease: Relationship to frontostriatal circuitry. Cognitive
and Behavioral Neurology, 16(4), 193-210.
Zhang, Z., & Roman, G. C. (1993). Worldwide occurrence of Parkinson's disease: An updated
review. Neuroepidemiology, 12, 195-208.
Zraick, R. I., & Liss, J. M. (2000). A comparison of equal-appearing interval scaling and direct
magnitude estimation of nasal voice quality. Journal of Speech, Language and Hearing
Research, 43(4), 979-988.
Zundel, K. M. (1996). Telemedicine: History, applications, and impact on librarianship. Bulletin of
the Medical Library Association, 84(1), 71-79.
Zwirner, P., & Barnes, G. J. (1992). Vocal tract steadiness: A measure of phonatory and upper
airway motor control during phonation in dysarthria. Journal of Speech and Hearing
Research, 35, 761-768.
231
232
Appendix – Published Papers
1) Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. Assessing
1-15.
2) Constantinescu, G., Theodoros, D., Russell, T., Ward, E., Wilson, S., & Wootton, R. (in press).
3) Constantinescu, G.A., Theodoros, D.G., Russell, T.G., Ward, E.C., Wilson, S.J., & Wootton,
R. (2010). Home-based speech treatment for Parkinson‟s disease delivered remotely: A case
233
INT. J. LANG. COMM. DIS.
2010, Early Online Article, 1–15
Research Report
Assessing disordered speech and voice in Parkinson’s disease: a telerehabilitation
application
Gabriella Constantinescu†, Deborah Theodoros†, Trevor Russell‡, Elizabeth Ward†, Stephen Wilson§
and Richard Wootton{
†Division of Speech Pathology, University of Queensland, St Lucia, Brisbane, QLD, Australia
‡Division of Physiotherapy, University of Queensland, St Lucia, Brisbane, QLD, Australia
§School of Information Technology and Electrical Engineering, University of Queensland, St Lucia, Brisbane, QLD, Australia
{Scottish Centre for Telehealth, Aberdeen, UK
Int J Lang Commun Disord Downloaded from informahealthcare.com by University of Queensland
(Received April 2009; accepted November 2009)
Abstract
Background: Patients with Parkinson’s disease face numerous access barriers to speech pathology services for
appropriate assessment and treatment. Telerehabilitation is a possible solution to this problem, whereby
rehabilitation services may be delivered to the patient at a distance, via telecommunication and information
technologies. A number of studies have demonstrated the capacity of telerehabilitation to provide reliable and valid
assessments of speech, voice and language. However, no studies have specifically focused on assessing patients with
Parkinson’s disease.
Aims: To investigate the validity and reliability of a telerehabilitation application for assessing the speech and voice
For personal use only.
disorder associated with Parkinson’s disease.

Methods & Procedures: Sixty-one participants with Parkinson’s disease and hypokinetic dysarthria were
simultaneously assessed in an online and face-to-face environment by two speech –language pathologists. The
assessment protocol included perceptual measures of voice and oromotor function, articulatory precision, speech
intelligibility, and acoustic measures of vocal sound pressure level, phonation time and pitch range. Online
assessments were conducted via a personal computer-based videoconferencing system with store-and-forward
capabilities, operating on a 128 kbit/s Internet connection. The level of agreement between the online and face-to-
face ratings was determined using several different analyses, depending on the parameter. These included per cent
close agreement, quadratic weighted Kappa, and the Bland and Altman limits of agreement.
Outcomes & Results: Per cent close agreement between the two environments was within a predetermined clinical
criterion of 80% agreement for all voice and oromotor parameters, articulatory precision and speech intelligibility in
conversation. Levels of agreement between the environments, based on quadratic weighted Kappa, ranged from poor
to good for vocal parameters and from fair to very good for oromotor parameters. Bland and Altman limits of
agreement analyses revealed comparability between online and face-to-face environments for vocal sound pressure
level, phonation time, pitch range, sentence intelligibility and communication efficiency in reading. Intra- and inter-
rater reliability scores for all tasks were comparable between the online and face-to-face environments.
Conclusions & Implications: For the majority of parameters, comparable levels of agreement were achieved between
the two environments. Online assessment of disordered speech and voice in Parkinson’s disease appears to be valid
and reliable. The telerehabilitation application described in this study provides evidence for the delivery of online
assessment for the dysarthric speech disorder associated with Parkinson’s disease.
Keywords: Parkinson’s disease, telerehabilitation, Internet-based assessment, speech and voice disorder.
Address correspondence to: Gabriella Constantinescu, Division of Speech Pathology, University of Queensland, St Lucia, Brisbane, QLD
4067, Australia; e-mail: gabriella@hearandsaycentre.com.au
International Journal of Language & Communication Disorders
ISSN 1368-2822 print/ISSN 1460-6984 online q 2010 Royal College of Speech & Language Therapists
http://www.informahealthcare.com
DOI: 10.3109/13682820903470569
2 Gabriella Constantinescu et al.
What this paper adds

What is already known on this subject
Recent studies have indicated the benefits of using telerehabilitation to assess motor speech, voice and language
disorders in the adult neurological population. However, no studies have focused specifically on the online
assessment of hypokinetic dysarthria associated with Parkinson’s disease.
What this study adds

The study has demonstrated the validity and reliability of online assessment for evaluating the speech and voice
disorder specifically associated with hypokinetic dysarthria and Parkinson’s disease. The study provides a basis for
the delivery of online services for people with Parkinson’s disease.
Introduction Reduced services for people with Parkinson’s

disease have also been noted in studies conducted in
Parkinson’s disease is a progressive neurodegenerative

the United Kingdom and Sweden, where relatively few
disorder associated with significant motor disturbance
participants surveyed (as little as 2.0 to 14.2%) had
and speech difficulties. Worldwide, Parkinson’s disease
been able to access a SLP (Hartelius and Svensson 1994,
is present in approximately 1% of the population aged
Mutch et al. 1986, Oxtoby 1982, Peto et al. 1997).
65 years and above (De Rijk et al. 1997). In 2005, the
These findings are in contrast to the 49% of people
prevalence rate of Parkinson’s disease in Australia was
with Parkinson’s disease who were identified as
estimated to be 290 cases per 100 000 population aged having speech difficulties (Oxtoby 1982) and 33%
55– 64 years, with a total cost to the health system of of people with Parkinson’s disease presenting with
AUD$343.9 million per year (Access Economics 2007). speech difficulties who were dissatisfied with their level
Hypokinetic dysarthria, the motor speech impair- of service access (Peto et al. 1997).
ment associated with Parkinson’s disease, is character- The disparity between the supply and demand of
ized by monotony of pitch and loudness, reduced speech pathology services for people with Parkinson’s
loudness and stress, imprecise articulation, variable rate disease suggests the need for an additional or alternate
and short rushes of speech, inappropriate silences, and a mode of service delivery for this population. One
harsh and breathy voice (Darley et al. 1969). The possible solution is the use of telerehabilitation,
incidence of the speech disorder occurs in as many as whereby telecommunication and information technol-
50– 90% of individuals during the course of their ogies are used in the delivery of healthcare at a distance.
disease (Ramig et al. 2004), with the severity of the Telerehabilitation is an emerging field in speech
dysarthria increasing with disease progression (Hartelius pathology and research has included various communi-
and Svensson 1994). Impaired speech intelligibility may cation technologies for the assessment and treatment of
result in decreased involvement in communicative motor speech, fluency, voice and language disorders.
exchanges, isolation within the family and community, The earlier use of the telephone, closed-circuit
and a subsequent degradation in the person’s quality of television and satellite-based videoconferencing, are
life (Oxtoby 1982). gradually being replaced by Internet-based videocon-
Patient access to speech-language pathology services ferencing via a personal computer which are now
for appropriate assessment and treatment of this accessible to many individuals.
condition is limited. Grimm et al. (2004) in a survey Regardless of the technology used, valid and reliable
of 250 people with Parkinson’s disease across Queens- assessment procedures need to be established to ensure
land, Australia, identified certain barriers impacting on effective telerehabilitation services. Recent studies using
service access. The barriers included: low priority placed Internet-based videoconferencing via a personal com-
on speech pathology services in the public health puter have highlighted the benefits of telerehabilitation
system; limited availability of speech – language path- for the assessment of motor speech performance, story
ologists (SLPs) trained to administer the effective Lee retell, language comprehension and expression for
Silverman Voice Treatmentw (LSVT) for Parkinson’s adults with neurological impairments (Brennan et al.
disease; physical incapacity of the individuals; difficul- 2004, Georgeadis et al. 2004, Hill et al. 2006, 2008a,
ties with transport and travel, and the large distances to 2008b, Palsbo 2007, Theodoros et al. 2008). The
the health service facilities. Of the people with studies commonly reported high levels of agreement
Parkinson’s disease who were surveyed, only 36.7% between the online and face-to-face ratings for the
reported having access to speech pathology services. majority of parameters investigated on the informal and
Online speech assessment of Parkinson’s disease 3
standardized assessments. High participant satisfaction Fifty-seven of the participants had been diagnosed with
with the online environment was also obtained Idiopathic Parkinson’s disease, of which nine participants
(Brennan et al. 2004, Georgeadis et al. 2004, Hill had undergone surgical treatment for Parkinson’s disease
et al. 2008a, 2008b, Theodoros et al. 2008). including deep brain stimulation (seven participants) and
Furthermore, it was encouraging to note that severity pallidotomy (two participants). The remaining four
of aphasia and apraxia did not significantly impact on participants in the cohort had been diagnosed with
the accuracy of the online assessments. However, the Parkinson-plus syndromes, including progressive supra-
occasional audio-visual disturbances online caused by nuclear palsy (three participants) and multiple system
heavy traffic on the network did make it more difficult atrophy (one participant). Stages of Parkinson’s disease as
for the SLPs to conduct the apraxia assessment for per the Hoehn and Yahr (1967) scale for the participants
the more severe participants, and rate the aphasia ranged from I to IV with 48 participants rated at Stages I
parameters of naming and paraphasia (Hill et al. 2008a, and II, and 13 participants rated at Stages III and IV. For
2008b). The ratings of some motor speech parameters all participants, an overall severity level for hypokinetic
(palatal movement in speech, laryngeal volume, tongue dysarthria was determined by the investigators from
elevation and lateral tongue movements), were also clinical judgement. The dysarthria levels ranged from
more difficult online for all participants involved, mild to severe, with 41% of participants considered as
and this was a result of audio-visual difficulties, mild, 48% of participants as moderate and 11% of
lighting, camera positioning and lack of zoom focus participants as severe. Participants were recruited from
(Hill et al. 2006). various support groups of Parkinson’s Queensland
As previous studies have demonstrated the capacity Incorporated, public hospitals and from private neurol-
of telerehabilitation to provide reliable and valid ogists in Brisbane, Australia. Proficiency in the use of
assessment of speech, voice and language, it was computers was not a requirement for inclusion in the
proposed that an Internet-based application to provide study as all aspects of the online assessment delivery were
services to people with Parkinson’s disease might lessen performed by the online assessing SLP. Exclusion criteria
the access issues that exist for this population. In order included a speech and/or language disturbance or a co-
to ensure that valid assessments underpin treatment existing neurological disorder inconsistent with Parkin-
programs, the current study aimed to investigate the son’s disease, a severe uncorrected auditory and/or
validity and reliability of an Internet-based assess- visual disturbance, a cognitive disturbance inconsistent
ment protocol specifically designed to evaluate the with the capacity to provide informed consent, a
speech and voice disturbances associated with Parkin- respiratory dysfunction unrelated to the neurological
son’s disease, by comparison with clinical face-to-face disorder and a positive history of alcohol abuse. The
assessment. It was hypothesized that online assessment primary mode of assessment (online or face-to-face led)
of the speech and voice disturbances in Parkinson’s was randomly selected for each participant and 31
disease can be achieved to a level comparable with assessments were led face-to-face and 30 led online.
standard face-to-face assessment. The current study A computerized random-number generator was used
forms the first validation stage to determine the for the randomization.
feasibility of online delivery as a complete assessment
and treatment service delivery model for Parkinson’s
disease. Assessors
Three SLPs experienced in the assessment of motor
speech disorders and Parkinson’s disease took part in the
Method study. Assessments were conducted at The University of
Queensland. Prior to the commencement of the study,
Participants
SLP training was conducted in a 3-hour session which
Before commencement of the study, ethical clearance covered the administration of all assessments in both the
was obtained from the Behavioural and Social Sciences online and face-to-face assessment environments. The
Ethical Review Committee of The University of SLPs were deemed competent with online adminis-
Queensland, Brisbane. Sixty-one participants with tration when they could adequately deliver a mock
Parkinson’s disease and hypokinetic dysarthria (42 session within a 1-hour time frame and also agree on the
males, 19 females) aged between 52 and 89 years level of severity of five dysarthric speakers who were not
(mean ¼ 69.23 years; standard deviation (SD) ¼ 8.60) involved in the study. The speakers were judged on the
volunteered for the study. Participants were diagnosed as perceptual measures of voice, overall articulatory
having Parkinson’s disease by a neurologist experienced in precision, overall speech intelligibility in conversation
movement disorders. Time post-diagnosis ranged from and oromotor function. During the study, two of the
6 months to 30 years (mean ¼ 6.52 years; SD ¼ 6.53). three SLPs took part in each assessment session, where
one SLP led the session, while the second SLP acted as a Oromotor function
silent rater and did not interact with the participant.
An informal assessment of non-speech oromotor
One SLP assessed the participant in the face-to-face
function was developed to evaluate specific parameters
environment (within the same room as the participant),
using a five-point rating scale (1 ¼ normal, 5 ¼
while the second SLP conducted the assessment in
severely impaired). The parameters included masked
the online environment, through a videoconferencing
facial expression, lip movement (retraction, pucker,
link via the Internet. The SLPs were also randomized
seal, alternate movement), tongue movement (sym-
to the assessment environments and were blind to
metry, protrusion, elevation/depression, lateral and
the participant and their level of hypokinetic
alternate movement), breath support and diadochoki-
dysarthria prior to assessment. In total, SLP 1 took
netic (DDK) rates (alternate motion rate [AMR] /p^
part in 25 of the online assessments (16 as leader and 9
p^/and sequential motion rate [SMR] /p^t^k^/).
as silent rater) and 20 of the face-to-face assessments
(9 as leader and 11 as silent rater); SLP 2 took part in 21
online sessions (7 as leader and 14 as silent rater) and 24 Overall articulatory precision
face-to-face assessments (11 as leader and 13 as silent
rater); and SLP 3 took part in 14 online assessments A perceptual rating of each participant’s articulatory
(7 as leader and 7 as silent rater) and 18 face-to-face precision was made from the speech sample obtained
assessments (12 as leader and 6 as silent rater). during the reading of The Grandfather Passage.
Articulatory precision was rated on a five-point scale
(1 ¼ normal, precise production of sounds, 5 ¼ severe
distortion or imprecision that interferes with speech
Assessment battery
intelligibility).
Each participant underwent a 1-hour assessment on one
occasion on a battery of perceptual and acoustic
measures specifically designed for this study. The Measures of speech intelligibility
battery consisted of perceptual ratings of voice and The Assessment of Intelligibility of Dysarthric Speech
oromotor parameters, overall articulatory precision, (ASSIDS) (Yorkston and Beukelman 1981) was used to
speech intelligibility in reading and conversation, and measure speech intelligibility at the single word and
an instrumental evaluation of sound pressure levels, sentence level, as well as communication efficiency.
duration of vowel prolongation and pitch range. For this task, participants read or repeated a series of
These measures were chosen for the study as they are 50 words and 22 sentences of increasing length. The
commonly used to diagnose and define the level of words and sentences had been randomly generated prior
severity of hypokinetic dysarthria associated with to the assessment, in accordance with test procedure. The
Parkinson’s disease. Furthermore, as this study forms reading material was displayed on the participant’s screen
part of a larger validation trial that also evaluates online or presented as per the test booklet, depending on the
treatment, the measures were chosen as they have been assessment environment. Copyright approval was
used in the LSVTw literature as sensitive predictors of obtained from the publishers (PRO-ED, Austin, TX,
treatment change (Ramig et al. 1995a). USA) to enable conversion of test materials to an online
format. Audio recordings of participant speech samples
were made in both environments. Following assessments
Perceptual measures of all participants, the speech samples obtained face-to-
face and online were numerically coded, randomized and
Perceptual voice parameters
saved to CD for analysis and scoring. Two independent
In the absence of a standardized voice assessment SLPs who did not participate in the study and who were
available at the time of the study, vocal parameters were blinded to the assessment environment and participants
evaluated using a five-point rating scale developed for transcribed the speech samples obtained in each
the study (1 ¼ normal, 5 ¼ severely impaired). The environment. Following the ASSIDS ratings, the values
reading of a standard passage, The Grandfather Passage given by the two SLPs from the online and face-to-face
(Darley et al. 1975) was used for perceptual ratings of recordings were averaged to express a single mean value for
breathiness, roughness (lack of clarity), strain-strangled each sample obtained in that environment. Scores for the
vocal quality, vocal tremor, pitch and phonation breaks. word and sentence intelligibility tasks were expressed as
Modal pitch and loudness levels and pitch and loudness per cent correct. The communication efficiency ratio was
variability were rated on a 30 s conversational determined by dividing the participant’s rate of
monologue about a topic of interest such as family, intelligible speech (intelligible words per minute) by
hobbies or a recent holiday trip. the mean rate of intelligible words per minute for
normal speakers (190 words per minute) (Yorkston and Assessment environment
Beukelman 1981).
An additional rating of the participant’s overall speech Online environment
intelligibility in conversation was made from the 30 s Two personal computer-based videoconferencing
monologue sample using a five-point scale (1 ¼ normal, systems developed at The University of Queensland
completely intelligible speech, 5 ¼ severely unintelligible were used for the online assessment. The applications
speech with difficulties deciphering many words). For this operated on a 128 kbit/s Internet connection which was
task, the rating was made by the two SLPs who took part the minimum connection speed available in Queens-
in the online and face-to-face assessment. land’s public health systems at the time of the study.
Videoconferencing at 320 £ 240 pixel resolution was
Acoustic measures conducted between the online SLP’s computer and that
of the participant. Additional features of the system
The LSVTw Evaluation Protocol (Ramig et al. 1995b) which were used in this study included the ability: to
was used to assess the participant’s sound pressure levels display printed material and instructional video clips on
(SPL), duration of vowel prolongation and pitch range the participant’s screen; to control the remote camera
during several speaking tasks. This protocol has been with the use of a robotic arm and adjust its alignment
widely used in the LSVTw literature as a routine for optimal viewing of the participant’s head and upper
assessment. torso; to capture high-quality video (640 £ 480 pixel
resolution compressed with the windows media video
CODEC Version 8 at 384 kbit/s) and audio record-
Sound pressure levels and duration of vowel prolongation
ings (windows media audio CODEC Version 8 at
The SPLs (dB-C) of the participant’s speech were 368 kbit/s) independent of videoconferencing for the
recorded during six maximum sustained vowel oromotor tasks, and then to store-and-forward these
phonation of /a/, readings of the Rainbow Passage audio and video files back to the online SLP for later
(Fairbanks 1960) and The Grandfather Passage, and review.
during a 30 s conversational monologue. The duration For tasks requiring acoustic measures, both the
of each vowel phonation was also measured in seconds. online and face-to-face SLPs were able to view and
For all tasks, the participants were instructed to speak in sample real-time calibrated average recordings of SPL
a comfortable voice and no reference was made to their (dB-C), peak frequency (Hz) and duration (s) data via
loudness level. Following the assessment, the SPL and the system’s acoustic speech processor specifically
duration levels were then averaged to provide mean developed for this study. The validity of the speech
levels for each participant. processor as an acoustic measurement device was
examined in a series of calibration trials. These trials
involved the generation of pure-tones by a Function
Pitch range
Generator (Topward Electronic Instruments Model
Each participant performed a series of six vocal glides, TFG-462) at varying levels of SPL (55 – 95 dB) and
reaching their highest and lowest pitch levels respect- pitch (100– 975 Hz), and comparing measures from
ively. No reference was made to their loudness level. the acoustic speech processor with those of the
The average highest and lowest frequency levels (Hz) commercially available Visi-Pitch II (Kay Elemetrics
obtained for each participant were then converted to a Model No. 3300). Statistical analyses using paired
maximum range in semitones (ST) (de Pijper 2007). t-tests revealed no significant differences ( p . 0.05) in
SPL and pitch measures for pure-tones between the two
devices. Furthermore, verification trials comparing SPL
Participant satisfaction questionnaire
measures using the speech processor with those from a
The 30 participants in the online-led assessments Digital Sound Level Meter (Radio Shackw Model
completed a brief questionnaire. On a five-point scale, No. 23-553) using voice samples (66 – 91 dB) also
the questionnaire evaluated the level of participant revealed no significant differences ( p . 0.05) between
satisfaction with: (1) the online assessment sessions the two devices (table 1). To standardize the acoustic
(possible responses ranging from would not participate measures across the two assessment environments, the
again to would prefer these types of sessions to face-to-face system’s acoustic speech processor was used as the
sessions); (2) the audio and video quality during the objective measurement tool in both the online and face-
sessions (responses ranging from poor to excellent); and to-face environments for all acoustic measures.
(3) overall satisfaction with the online modality (ranging During the online assessment, the online SLP wore
from not at all satisfied to very satisfied). Please refer to the a headset microphone attached to the telerehabilitation
Appendix. system for communication with the participant.
Table 1. Calibration trial for measures of pitch and sound pressure levels
Speech MAD speech MAD speech

processor, mean Visi-Pitch II, SLM, mean processor and Visi- processor and T-value p-value
Task variable (SD) mean (SD) (SD) Pitch (SD) SLM (SD) (t-test) (t-test)
Pure-tone pitch 535.94 (263.32) 536.85 (263.86) – 1.46 (3.66) – 20.015 0.988
(Hz)
Pure-tone SPL 77.95 (9.70) 77.71 (9.77) – 0.36 (0.19) – 0.077 0.939
(dB-C)
Voice samples 79.36 (7.67) – 79.43 (7.53) – 0.5 (0.51) 20.025 0.980
SPL (dB-C)
Note: Dashes ( – ) correspond to data not obtained. Speech processor is the system’s online acoustic speech processor. Visi-Pitch II (Kay Elemetrics Model No. 3300). SLM is a Digital
Sound-Level Meter (Radio Shackw Model No. 23-553). MAD is maximum average difference. Pure-tone pitch is the pitch calibration trial of the speech processor with the Visi-Pitch II
using pure-tones. Pure-tone SPL is the sound pressure level calibration trial of the speech processor with the Visi-Pitch II using pure-tones. Voice samples SPL is the sound pressure level
verification trial comparing sound pressure level measures using the speech processor with the Digital Sound Level Meter for voice samples. Measurements are in Hertz. dB-C is
measurements in decibels-C weighted. SD, standard deviation.
The SLP controlled all displays on the participant’s displayed on the participant’s screen by the online SLP.
screen, without the need for the participant to operate Throughout the online-led assessment, the face-to-face
the system. For standardization purposes, the participant SLP acted as the silent rater at the participant site. The
was seated in front of the system at a distance of face-to-face SLP wore headphones and was able to
approximately 50 cm from the monitor and wore a follow the assessment instructions given to the
headset microphone to enable interaction with the participant. The online assessment environment is
online SLP during videoconferencing. The microphone represented in figures 1a and 1b.
distance was set at 5 cm from the corner of the
participant’s mouth in order to reduce sound distortion,
maximize visibility of the participant’s face, and allow for Face-to-face environment
accurate recordings of pitch and SPL.

During the online-led assessment, the online SLP During the face-to-face led assessment, the participant
administered the various tasks and interacted with the was seated in front of the telerehabilitation system with
the monitor turned off. Standard face-to-face test
participant over the 128 kbit/s Internet videoconferen-
administration procedures were used. The online SLP
cing link. At this bandwidth, live ratings of pitch and
became the silent rater and viewed, listened, and
SPL were possible, however, judgements of fine
recorded tasks while the face-to-face SLP interacted
movements and precision on the oromotor assessment
with the participant. In keeping with the online
were more difficult due to a low picture frame rate and
procedure, the face-to-face SLP obtained real-time
resolution picture quality. In addition, the real-time
measures of SPL, duration and pitch via the system’s
detection of subtle features of speech production for
acoustic speech processor. The SLP also used a video
perceptual ratings of voice, overall articulatory precision
camera and minidisk recorder to collect video and
and speech intelligibility was also more difficult on
audio data respectively for later analyses, as per standard
occasion due to the degradation of audio quality.
clinical practice. The video camera was positioned as
Therefore, to improve the video and audio quality for
close as possible behind the web cameras and a
rating, the online SLP used the store-and-forward
microphone on a stand was connected to a minidisk
features of the system to record the task and store the
recorder and placed 30 cm from the participant. The
video and audio files for later viewing and analyses. The
face-to-face SLP wore headphones and was able to hear
store-and-forward feature was used routinely by all the online SLP if there was a need for a task repetition
online SLPs (leading and silent assessors). To or further online recording. A summary of the face-to-
standardize the perceptual measures, SLPs in both face assessment procedure is also displayed in table 2.
environments rated the assessments live (where
possible) and then reviewed the sessions off-line using
the equipment available in that environment.
Statistical analyses
A summary of the online assessment procedure is
displayed in table 2. Online and face-to-face ratings and measurements for
In addition, the effects of the audio-visual difficulties all participants were compared on each assessment task
with videoconferencing on the participants’ ability to to determine the level of agreement between the two
follow task instructions were minimized with the use of environments. For those parameters consisting of
pre-recorded task demonstrations of the oromotor ordinal data (perceptual ratings of voice and oromotor
assessment. Where necessary, these demonstrations were parameters, overall articulatory precision and speech
Table 2. Assessment procedure for online and face-to-face environment
Online Online scoring FTF scoring

Assessment measure instrument procedure FTF instrument procedure
Perceptual Measures
Perceptual voice parameters Online store and Real-time where possible and Minidisk recorder Real-time where possible
forward audio reviewed off-line and reviewed off-line
Oromotor parameters Online store and Real-time where possible and Video camera Real-time where possible
forward video reviewed off-line and reviewed off-line
Overall articulatory precision Online store and Real-time where possible and Minidisk recorder Real-time where possible
forward audio reviewed off-line and reviewed off-line
Overall speech intelligibility Online store and Real-time where possible and Minidisk recorder Real-time where possible
in conversation forward audio reviewed off-line and reviewed off-line
ASSIDS Online store and Off-line Minidisk recorder Offline
forward audio
Acoustic measures
Sound pressure levels Online acoustic Real-time Online acoustic speech Real-time
speech processor processor

Duration of phonation Online acoustic Real-time Online acoustic speech Real-time
Pitch range Online acoustic Real-time Online acoustic speech Real-time
Participant satisfaction Paper based End of online-led assessment n.a. n.a.
questionnaire session
Note: ASSIDS ¼ assessment of intelligibility of dysarthric speech; FTF, face-to-face assessment environment; n.a., not applicable.
intelligibility in conversation), percent close agreement beyond chance between the online and face-to-face
(PCA) and the quadratic weighted Kappa (kw) statistic measures. The kw assigned weights to the observed and
(Landis and Koch 1977) were calculated. Analysis of chance agreement and presented levels of agreement
the ASSIDS and acoustic parameters (SPL tasks, where kw less than 0.20 is interpreted as poor;
duration of vowel phonation and pitch range) were 0.21– 0.40 is fair; 0.41– 0.60 is moderate; 0.61– 0.80 is
performed using the Bland and Altman (1986) ‘limits good, and 0.81– 1.00 indicates very good agreement
of agreement’ method for continuous data. (Landis and Koch 1977). For this study, the clinical
criterion for an acceptable level of agreement was set at
kw . 0.6 (good agreement).
Percent close agreement
PCA was chosen as it is commonly used to quantify
agreement in perceptual ratings of dysarthria. PCA was Bland and Altman (1986) limits of agreement
also selected for the present study to further verify kw This statistic establishes the limits of agreement (LA)
as it has been reported in some instances that non- within which 95% of differences between the two
linear distribution of data can negatively impact on kw environments are predicted to lie. If the LA are found to
creating a paradox (Cicchetti and Feinstein 1990). PCA be within a predetermined clinical criterion, the new
expressed the percentage of ratings where differences method can be considered an acceptable measurement
were within ^1 scale point on the perceptual rating tool and the two methods can be used interchangeably
scales (Kearns and Simmons 1988). In keeping with (Bland and Altman 1986). In the absence of a reported
previous studies that examined dysarthric speech using minimal, clinically important difference for word and
perceptual rating scales, the clinical criterion for an sentence intelligibility of the ASSIDS assessment, the
acceptable level of agreement in the present study was clinical criterion was established on the test – retest
considered to be equal to or greater than 80% agreement variability reported in the manual for dysarthric
within ^1 scale point (Kearns and Simmons 1988). speakers assessed in the face-to-face environment
(Yorkston and Beukelman 1981). Values of ^3.2%
and ^8.6% were set for these respective measures. In
Quadratic weighted Kappa statistic
addition, the clinical criterion for the communication
The kw is widely used in telerehabilitation studies for efficiency ratio was set at ^0.27 which was consistent
ordinal data and provides an indication of agreement with the criterion determined by Hill et al. (2006) for
between raters (Landis and Koch 1977). In the present dysarthric speakers. Furthermore, as the assessment
study, the statistic provided a measure of agreement battery used in the study was designed ultimately to
Figure 1a. Online-led assessment by online SLP and equipment at site. Note (1) the videoconferencing system displaying the participant;
(2) pitch (Hz) and SPL (dB-C) data via the system’s acoustic speech processor; and (3) the web camera.
determine treatment outcome, the clinical criteria for (Ramig et al. 1995a). The clinical criterion for the
SPL, vowel duration and pitch range tasks were set at duration of vowel phonation was set at ^ 3 s, as the
levels below the minimal improvement expected minimal change in phonation time expected with
following the LSVTw. For all SPL tasks (maximum the LSVTw has been reported to be a mean of 3.72 s
vowel phonation, reading and monologue loudness), (Ramig et al. 1995a). For measures of pitch range, the
the clinical criterion was set at ^4 dB difference clinical criterion was set at ^3 ST, which was below the
between the two raters, a level below the 4.5 –14.03 dB 4 ST minimum improvement in fundamental frequency
mean level of improvement reported with the LSVTw range post-LSVTw (Ramig et al. 1994). This clinical
Figure 1b. Online-led assessment at participant site with face-to-face SLP as the silent rater (left). Note (1) the videoconferencing system
displaying the online SLP; (2) the web cameras; (3) the video camera; and (4) pitch (Hz) and SPL (dB-C) data via the system’s acoustic speech
processor displayed for the face-to-face clinician.
criterion was also in keeping with the level of subject breaks, phonation breaks, modal pitch and loudness
variability in healthy adults that can range from 2 to 4 ST variability) were below the clinical criterion of good
(Gelfer 1986). agreement (kw . 0.6).
Reliability Oromotor function parameters

Reliability between the online and face-to-face Analyses revealed that all individual oromotor par-
environments was conducted for all the perceptually ameters reached the clinical criterion for PCA (table 4).
based assessments and the acoustic pitch measure. The kw indicated that only two parameters (masked
Although the pitch data was objectively obtained via the facial expression and lip retraction) fell outside of the
acoustic speech processor, the SLPs were required to clinical criterion of good agreement.
select a sample pitch level from a section of the vocal
glide thus introducing a subjective element to this task.
The SLPs used the audio, video and pitch files captured Overall articulatory precision
during the assessment session in the respective
For ratings of overall articulatory precision between the

environments to rate and score the various parameters. online and face-to-face environments, PCA (100) and
Intra- and inter-rater reliability between the online and kw (0.67 good agreement) were within the clinical
face-to-face assessors was calculated using two-way, criteria.
random effect intra-class correlations (ICC(2,1)) for
20% (n ¼ 13) of participants in each environment. For
inter-rater reliability, the third SLP who did not take Measures of intelligibility
part in a particular assessment session became the
additional rater and was randomly assigned to the For the ASSIDS assessment, the Bland and Altman
online or face-to-face ratings. Intra- and inter-rater (1986) LA at the 95% confidence interval are dis-
reliability was calculated collectively using the ratings played in figures 2a– 2c for word (LA ¼ 210.27% to
from each of the three SLPs in the particular 8.77%) and sentence intelligibility reading tasks
environment. ICC values below 0.40 corresponded to (LA ¼ 25.59% to 6.16%), and the communication
poor-to-fair reliability; between 0.40 and 0.75 to efficiency ratio (LA ¼ 20.12 to 0.10). The LA for
moderate-to-good reliability; and values above 0.75 sentence intelligibility and communication efficiency
represented very good reliability (Fleiss 1981). ratio were within the respective clinical criterion
(^8.6% and ^0.27), while the word intelligibility
LA fell outside of the clinical criterion of ^3.2%. In
Results addition, perceptual ratings of overall speech intellig-
ibility in conversation were within the clinical criteria
Perceptual measures
for PCA (98.36) and kw (0.79 good agreement).
Perceptual voice parameters
Table 4. Perceptual oromotor parameters between face-to-face
All individual voice parameters met the predetermined and online environments
clinical criterion of 80% agreement for PCA (table 3).
However, when using kw, seven of the ten parameters Oromotor parameters PCA kw
(breathiness, roughness, strained-strangled, pitch Breath support 96.72 0.83 (very good)
Masked facial expression 86.89 0.31 (fair)*
Table 3. Perceptual voice parameters between face-to-face and Lip movement
online environments Retraction 98.36 0.56 (moderate)*
Pucker 98.36 0.77 (good)
Voice parameters PCA kw
Seal 95.08 0.66 (good)
Breathiness 91.80 0.36 (fair)* Alternate 100 0.95 (very good)
Roughness 95.08 0.33 (fair)* Tongue movement
Strained-strangled 95.08 0.41 (moderate)* Symmetry 100 0.66 (good)
Vocal tremor 100 0.69 (good) Protrusion 100 0.94 (very good)
Pitch breaks 96.66 0.11 (poor)* Elevation/depression 98.36 0.93 (very good)
Phonation breaks 95.00 0.37 (fair)* Lateral 100 0.89 (very good)
Modal pitch 96.66 0.38 (fair)* Alternate 100 0.85 (very good)
Pitch variability 96.72 0.63 (good) Diadochokinetic (DDK)
Loudness level 100 0.69 (good) AMR /p^p^/ 100 0.75 (good)
Loudness variability 98.36 0.49 (moderate)* SMR /p^t^k^/ 100 0.87 (very good)
Note: kw, quadratic weighted Kappa statistic: *achieved lower than the clinical criterion Note: kw ¼ quadratic weighted Kappa statistic: *achieved lower than the clinical
of kw . 0.6; PCA, per cent close agreement. criterion of kw . 0.6; PCA, per cent close agreement.
(a)
–1.97 vowel 1.35
–1.05 rainbow 1.04

–10.27 8.77
–1.18 grandfather 1.11
–1.07 monologue 0.81
– 15 – 10 –5 0 5 10 15
CC CC –5 –3 –1 1 3 5
CC CC
Limits of Agreement (%)
Limits of Agreement (dB)
(b)
Figure 3. Bland and Altman (1986) limits of agreement for sustained
vowel phonation, reading of the Rainbow and Grandfather Passages
and monologue task relative to the clinical criteria. Clinical criteria

(CC) for all SPL tasks ¼ ^4 dB.
–5.59 6.16
(LA ¼ 21.07 to 0.81 dB). For all SPL tasks, the LA
were within the predetermined clinical criterion of
^4 dB. Similarly, the LA for the duration of sustained
vowel phonation task (LA ¼ 22.74 to 2.70 s) were
– 10 –5 0 5 10 within the clinical criterion of ^3 s for online and face-
CC CC to-face ratings (figure 4a).
Limits of Agreement (%)
(c)
Pitch range
Figure 4b represents the LA for the pitch range (LA ¼
22.03 to 2.19 ST) that were within the clinical
– 0.12 0.10 criterion of ^3 ST.
Reliability
Intra-class correlations ranged from moderate to very
– 0.3 – 0.2 – 0.1 0 0.1 0.2 0.3 good intra-rater reliability in both assessment environ-
CC CC ments (ICC ¼ 0.43–0.99 face-to-face; ICC ¼ 0.48–
Limits of Agreement
0.99 online), indicating comparable intra-rater reliability
Figures 2a– c. Bland and Altman (1986) limits of agreement for between environments. Inter-rater reliability was also
word intelligibility, sentence intelligibility, and communication found to be comparable between environments, with
efficiency ratio relative to the clinical criteria: (a) word intelligibility: reliability values between moderate to very good for the
clinical criterion (CC) ¼ ^ 3.2 percentage points; (b) sentence
intelligibility: clinical criterion (CC) ¼ ^ 8.6 percentage points;
majority of face-to-face (ICC ¼ 0.43–0.99) and online
and (c) communication efficiency ratio: clinical criterion (ICC ¼ 0.48–0.99) assessments (table 5).
(CC) ¼ ^0.27.

Acoustic measures On the participant satisfaction questionnaire, the majority
of participants in the online-led assessments felt comfort-
Sound pressure levels
able while participating in the online session (56.67%) or
The Bland and Altman (1986) LA are displayed in very happy with the session (36.67%). The audio quality
figure 3 for sustained vowel phonation (LA ¼ 21.97 to during videoconferencing was largely rated as excellent
1.35 dB), the reading of the Rainbow Passage (40%), adequate (30%) and more than adequate
(LA ¼ 21.05 to 1.04 dB) and The Grandfather Passage (23.33%), while the video quality was primarily found
(LA ¼ 21.18 to 1.11 dB), and monologue loudness to be adequate (33.33%) or more than adequate (30%).
(a) system were sensitive enough for the SLP to determine

the presence and level of severity of specific voice
parameters online. However, at the more stringent level
of analysis (kw), the online and face-to-face assessments
–2.74 2.70 of several of the individual voice parameters failed to
reach acceptable levels of agreement. The kw revealed
greater variability than PCA for seven of the ten voice
parameters (breathiness, roughness, strained-strangled
vocal quality, pitch breaks, phonation breaks, modal
–4 –3 –2 –1 0 1 2 3 4 pitch and loudness variability) that fell below
CC CC clinical criterion (achieving poor to moderate agreement).
Limits of Agreement (s) These lower values may reflect rater variability
commonly seen in perceptual rating of voice and/or the
(b) nature of kw.
Findings of lower agreement for vocal parameters in
an online assessment were similarly reported by Hill

et al. (2006) in their study, where consensus agreement
– 2.03 2.19 was below 80% for ratings of strained-strangled vocal
quality (73.68%), breathiness (68.42%) and roughness
parameters (63.16%). These authors attributed aspects
of the lower agreement to possible inter-rater variability
that is inherent in perceptual ratings of voice. It has
–5 –3 –1 1 3 5
been acknowledged that naturally occurring variability
CC CC is commonly associated with traditional face-to-face
Limits of Agreement (ST) evaluations and reflects the subjective nature of
perceptual rating scales. The listener often applies

Figure 4a, b. Bland and Altman (1986) limits of agreement
for the duration of sustained phonation and pitch range relative to variable internal standards of pathological voice
the clinical criterion: (a) duration of sustained phonation: clinical qualities from their own experiences to the evaluation
criterion (CC) ¼ ^3 s; and (b) pitch range: clinical criterion (Kreiman and Gerratt 1998). Consequently, achieving
(CC) ¼ ^4 ST. high agreement between different raters in relation to
voice qualities is often problematic. A further explanation
for the lower kw relates to the statistic per se. It has been
Participant satisfaction with the online modality overall found that kw may be negatively influenced by the data
ranged from very satisfied (50%) to more than satisfied distribution, where despite high inter-rater agreement,
(30%) and satisfied (20%). calculations using non-linear data can result in low kw
(Cicchetti and Feinstein 1990). For example, in the
Discussion present study, strain-strangled vocal quality (non-linear
data), where although receiving comparable PCA to the
The results of the present study indicated that an vocal tremor parameter (linear data), obtained only
Internet-based assessment of the disordered speech and moderate agreement according to kw, in contrast to good
voice associated with Parkinson’s disease was generally agreement for vocal tremor (table 3).
reliable and valid. For the majority of the perceptual Further support for the comparability of the online
and acoustic parameters, the face-to-face and online and face-to-face assessment environments in the
ratings were within the clinical criteria as reported in evaluation of all vocal parameters was provided by the
previous face-to-face studies. reliability measures obtained for these ratings. Intra-
and inter-rater reliability were found to be similar in
Perceptual measures each assessment environment (table 5), suggesting that
the online assessment was as reliable as the traditional
Perceptual voice parameters face-to-face assessment on this task.
Analysis of individual voice parameters showed PCA
between the online and face-to-face ratings to be within
Oromotor parameters
the clinical criterion. This suggested that an Internet-
based videoconferencing application is a valid tool for For the purpose of assessing oromotor function, the two
administering a voice evaluation for Parkinson’s disease assessment environments were found to be clinically
and the audio store-and-forward capabilities of the comparable (PCA between 86.89% and 100%) (table 4).
Table 5. Intra-class correlation (ICC) values for intra-rater and inter-rater reliability for online and face-to-face ratings
FTF intra-rater Online intra-rater FTF inter-rater Online inter-rater

Assessment task reliability reliability reliability reliability
Oromotor parameters 0.85 0.81 0.74 0.76
Perceptual voice parameters 0.60 0.68 0.44 0.56
OAP 0.43 0.48 0.43 0.48
OIC 0.63 0.69 0.75 0.82
Pitch range 0.99 0.99 0.99 0.99
ASSIDS
WI 0.93 0.99 0.89 0.94
SI 0.97 0.77 0.94 0.82
CER 0.99 0.97 0.97 0.98
Note: CER, communication efficiency ratio; FTF, face-to-face environment; OAP, overall articulatory precision; OIC, overall speech intelligibility in conversation; ASSIDS, assessment
of intelligibility of dysarthric speech; SI, percentage sentence intelligibility; WI, percentage word intelligibility; ICC values below 0.40 ¼ poor-to-fair; from 0.40 to 0.75 ¼ moderate-
to-good, above 0.75 ¼ very good reliability. Reliability was calculated for three assessors.
Two of the 14 variables (masked facial expression and consistent with a previous study by Hill et al. (2006)
lip retraction), however, were below the clinical criterion where 89.47% consensus agreement was achieved
using kw. As noted with the vocal parameters, a certain between online and face-to-face ratings for consonant
level of variability inherent in perceptual ratings may have precision. Direct comparison of outcomes between the
contributed to the lower ratings for these oromotor two studies is not possible, however, due to the different
parameters. Previous face-to-face evaluations of facial assessment procedures used by Hill et al. (2006)
expression in Parkinson’s disease using a range of rating including the use of a four-point rating scale, different
scales and statistical analyses have shown varying levels of evaluation criteria and the non-simultaneous assess-
intra- and inter-rater reliability including fair, moderate ment of participants in the online and face-to-face
and substantial agreement (Goetz et al. 1995, Martinez- environments. In the present study, the high level of
Martin et al. 1994). The authors of these studies agreement between online and face-to-face ratings
attributed some level of the variability to the subjective and the comparable intra- and inter-rater reliability
interpretation of severity levels, the possible inexperience values (moderate agreement) between environments
of a few of the raters, and some variability in consensus lends further support to the validity of an online
prior to rating. application.
Similarly, labial judgements which have been
investigated predominantly in the cleft palate literature,
have been associated with rater variability (Morrant and Measures of speech intelligibility
Shaw 1996, Ritter et al. 2002). Face-to-face evaluations The Bland and Altman (1986) LA were used for online
of lip retraction in participants with repaired unilateral and face-to-face scores of the ASSIDS assessment
cleft lip have shown poor (Morrant and Shaw 1996) and (reading tasks). For the sentence tasks, both the LA for
moderate levels of inter-rater agreement (Ritter et al. sentence intelligibility and communication efficiency
2002). The subjective interpretation of severity levels has ratio were within the clinical criterion, indicating that
also been reported to affect rater agreement in these comparable measures of speech intelligibility can be
studies (Morrant and Shaw 1996, Ritter et al. 2002). achieved between the online store-and-forward method
Analyses of the oromotor parameters including and traditional face-to-face audio recordings using this
masked facial expression and lip retraction using kw assessment. Hill et al. (2006) similarly reported
may also require a level of cautious interpretation due to comparable values for the communication efficiency
the non-linear distribution of the data, and results may ratio in their study, while sentence intelligibility was
need to be interpreted alongside PCA. On the whole, just outside the clinical criterion. In the present study,
the high intra- and inter-rater reliability obtained for the LA for word intelligibility were outside the clinical
the oromotor parameters collectively, and the compar- criterion of ^3.2 percentage points between the
able levels between the two environments were very environments. It is possible that speaker severity may
encouraging. have influenced these results. Differences of three or
more words between raters, which were outside the
clinical criterion, occurred predominantly for partici-
Overall articulatory precision
pants with moderate and severely reduced intelligibility
The complete agreement obtained between the two (66.66% of the time), as identified on the overall speech
assessment environments for articulatory precision is intelligibility in conversation rating scale. The reduced
speaker intelligibility may have contributed to the within a clinically acceptable level, and the subjective
differences in ratings between the two environments. element of selecting a sample from a section of the vocal
Yorkston and Beukelman (1981) acknowledge that glide for analysis did not impact greatly on the results.
transcription of word tasks (as used in this study) in the Moreover, reliability measures revealed very good intra-
traditional face-to-face environment is difficult with and inter-rater reliability for the pitch task between the
more severely dysarthric speakers. For all the ASSIDS two environments. Collectively, the comparable values
tasks, intra- and inter-rater reliability was largely obtained for all acoustic measures suggested that the
comparable between the two environments, and showed acoustic speech processor used in the online environ-
very good agreement overall. These values are in keeping ment is a sensitive assessment tool that can be used to
with the very good reliability measures reported for the detect minimal changes in SPL, duration and pitch in a
ASSIDS assessment (Yorkston and Beukelman 1981). Parkinson’s disease assessment battery.
Comparable levels between assessment environments and
previous literature lend further support to the use of this
assessment in an online application. Audio-visual challenges and the online environment
For the additional ratings of overall speech intellig- Although the current trials largely support the feasi-
ibility in conversation, PCA within the clinical criterion bility of an online assessment, it is acknowledged that a
was achieved. This finding is in keeping with previous number of challenges were experienced with this
reports of high inter-rater agreement within one scale modality. Firstly, the assessments were conducted over a
point for overall speech intelligibility in traditional face- 128 kbit/s Internet videoconferencing connection,
to-face ratings (Sheard et al. 1991). The kw further which at this bandwidth, made the real-time evaluation
reflected good agreement within the clinical criterion. of a number of assessment items difficult. This included
Together with the comparable reliability measures, these the detection of fine motor movements and precision
findings suggest that ratings of overall speech intelligibility on the informal oromotor assessment due to the frame
in conversation can be made reliably online. rate and pixelated image, especially with movement.
The more subtle features of speech production on the
overall articulatory precision and speech intelligibility

Acoustic measures
tasks were also difficult to rate due to the less than optimal
Objective measures of SPL, vowel duration and pitch audio quality that occurred intermittently. As mentioned
range were obtained in real-time via the system’s previously, this was especially evident for participants
acoustic speech processor in both environments. with more severe dysarthria. The use of the store-and-
Although the calibration trials demonstrated the forward capabilities of the online system did allow for
validity of the speech processor as an objective tool high quality audio and video recordings and helped to
(table 1), it was important to assess the performance of minimize the audio-visual difficulties associated with real-
the speech processor during each of the different time videoconferencing.
assessment modes to determine if there was an effect of Despite the advantages of the store-and-forward
transmission across the Internet. Acoustic measures method, other difficulties were encountered that
were analysed using the Bland and Altman (1986) LA. affected the ratings. Factors such as shadowing on the
For the SPL tasks (sustained vowel phonation, reading participant’s face or reduced contrast of facial features
and conversational loudness), all measures were within due to background lighting and/or lack of webcam
the clinical criterion of ^4 dB. This finding is not zoom function, and considerable pixelation of a few
surprising as the acoustic speech processor provided video recordings on occasion did impact upon the
objective measures of SPL which were consistent online ratings of lip and tongue symmetry, tongue
between environments. The minor differences in values deviation and general facial features. Such difficulties
most likely reflected the slightly unsynchronized start of have also been reported in other online studies and it
SPL sampling by the two SLPs in each environment, has been proposed that web cameras with greater zoom
while transmission across the Internet appeared to have and focus capabilities and higher Internet bandwidth
little effect on the task (figure 3). Measures of sustained would possibly enhance the online ratings in real-time
vowel duration were also within the clinical criterion of (Hill et al. 2006). Additional audio disturbances such as
acceptable differences (^3 s), and the minor differences intermittent static in the recordings of the ASSIDS and
in duration may also have reflected the subjective reading passages were occasionally present within the
element in initiating the sampling (figure 4a). store-and-forward modality, making some judgements
For the pitch range task, the LA were also within the of these parameters more difficult. Furthermore,
predetermined clinical criterion of ^3 ST (figure 4b). participant factors such as head and body dystonias
This further demonstrates that SLPs in both environ- and stooped forward posture also made it more difficult
ments were able to obtain comparable pitch values on occasions to view the participant’s entire face and
judge aspects of lip retraction, pucker, and tongue Cowles, and Christina Iezzi for their contribution to the research.
movements. However, these difficulties in judgement Declaration of interest: The authors report no conflicts of interest.
The authors alone are responsible for the content and writing of
occurred independently of the assessment environment. the paper.
Overall, the audio-visual disturbances that occurred
with the store-and-forward method were infrequent
and did not impact significantly on rater-agreement and References
reliability between the two environments.
ACCESS ECONOMICS , 2007, Living with Parkinson’s Disease:
In relation to the participant–clinician rapport, the Challenges and Positive Steps for the Future. Report for
occasional audio delays of up to three seconds over the Parkinson’s Australia (Access Economics).
videoconferencing link had the potential to affect general BLAND, J. M. and ALTMAN, D. G., 1986, Statistical methods for
communicative interactions and turn-taking in the online assessing agreement between two methods of clinical
environment. However, the SLPs and participants were measurement. Lancet, i, 307– 310.
BRENNAN, D. M., GEORGEADIS, A. C., BARON, C. R. and BARKER,
able to quickly and effectively compensate for any L. M., 2004, The effect of videoconference-based telereh-
disturbances by actively waiting until the other had clearly abilitation on story retelling performance by brain-injured
finished speaking before replying. Thus, participants did subjects and its implications for remote speech – language
not report that this delay impacted greatly on the therapy. Telemedicine Journal and e-Health, 10, 147– 154.
assessment. Overall, the online assessments were rated CICCHETTI, D. V. and FEINSTEIN, A. R., 1990, High agreement but
low kappa: II. Resolving the paradoxes. Journal of Clinical
favourably by the participants, the application was user- Epidemiology, 43, 551– 558.
friendly, and the features of the application were DARLEY, F. L., ARONSON, A. E. and BROWN, J. R., 1969, Clusters of
conducive to assessment. deviant speech dimensions in the dysarthrias. Journal of
speech and hearing research, 12, 462– 496.
DARLEY, F. L., ARONSON, A. E. and BROWN, J. R., 1975, Motor
Conclusion Speech Disorders (Philadelphia, PA: W. B. Saunders).
DE PIJPER, J. R., 2007, Semitone conversions (available at: http://
The comparable ratings achieved for the majority of users.utu.fi/jyrtuoma/speech/semitone.html) (accessed 1
parameters between the online and face-to-face environ- March 2007).
ments and high rater reliability have demonstrated the DE RIJK, M. C., TZOURIO, C., BRETELER, M. M., DARTIGUES, J. F.,
AMADUCCI, L., LOPEZ-POUSA, S., MANUBENS-BERTRAN, J. M.,
validity and reliability of the online assessment tool for ALPEROVITCH, A. and ROCCA, W. A., 1997, Prevalence of
evaluating the speech and voice parameters associated parkinsonism and Parkinson’s disease in Europe: The EUR-
with hypokinetic dysarthria and Parkinson’s disease. The OPARKINSON collaborative study. European community
telerehabilitation application described in this study concerted action on the epidemiology of Parkinson’s disease.
provides a basis for the delivery of online assessment for Journal of Neurology Neurosurgery and Psychiatry, 62, 10–15.
FAIRBANKS, G., 1960, Voice and Articulation Drill Book (New York,
the dysarthric speech and voice disorder associated with NY: Harper & Brothers).
Parkinson’s disease. Online service delivery may prove to FLEISS, J. L., 1981, Statistical Methods for Rates and Proportions
be a necessary alternative or addition for people with (New York, NY: Wiley).
Parkinson’s disease, whereby lessening the difficult access GELFER, M. P., 1986, The stability of total phonational frequency
issues that exist for this population. Further research range. Journal of the Acoustical Society of America, 79, 83.
GEORGEADIS, A., BRENNAN, D., BARKER, L. M. and BARON, C.,
should involve analyses of online rater variability 2004, Telerehabilitation and its effect on story retelling by
involving a larger group of assessors and possible effects adults with neurogenic communication disorders. Aphasiology,
on assessment outcomes; provide a comprehensive 18, 639–652.
analyses of participant and SLP satisfaction with the GOETZ, C. G., STEBBINS, G. T., CHMURA, T. A., FAHN, S.,
online modality; examine the effects of dysarthria KLAWANS, H. L. and MARSDEN, C. D., 1995, Teaching tape
for the motor section of the unified Parkinson’s disease rating
severity and variance of Parkinson’s disease on the scale. Movement Disorders, 10, 263– 266.
online assessments; and provide an in-depth cost analysis GRIMM, N., PAUL, J. and WAKEHAM, A., 2004, Access to speech
of the use of telerehabilitation applications in the pathology services by people with Parkinson’s disease in
assessment and treatment of people with Parkinson’s Queensland, Australia (Unpublished report: University of
disease. Queensland, Australia).
HARTELIUS, L. and SVENSSON, P., 1994, Speech and swallowing
symptoms associated with Parkinson’s disease and multiple
sclerosis: a survey. Folia Phoniatrica et Logopaedica, 46,
Acknowledgements 9 –17.
HILL, A., THEODOROS, D. G., RUSSELL, T. G., CAHILL, L. M.,
This research was funded by a National Health and Medical WARD, E. C. and CLARK, K. M., 2006, An Internet-based
Research Council Project (Grant Number 301029). The research telerehabilitation system for the assessment of motor speech
forms part of the doctoral thesis of Gabriella Constantinescu at disorders: a pilot study. American Journal of Speech –
The University of Queensland, Australia. The authors would like Language Pathology, 15, 45 – 56.
to acknowledge Parkinson’s Queensland Incorporated, all HILL, A., THEODOROS, D., RUSSELL, T. and WARD, E., 2008a,
participants, Roy Anderson, Anne Hill, Monique Waite, Jasmin Using telerehabilitation to assess apraxia of speech in adults.
International Journal of Language and Communication RITTER, K., TROTMAN, C. A. and PHILLIPS, C., 2002, Validity
Disorders, 1 –17. of subjective evaluations for the assessment of lip scarring
HILL, A., THEODOROS, D., RUSSELL, T., WARD, E. and WOOTTON, and impairment. Cleft Palate-Craniofacial Journal, 39,
R., 2008b, The effects of aphasia severity on the ability to 587– 596.
assess language disorders via telerehabilitation. Aphasiology, SHEARD, C., ADAMS, R. D. and DAVIS, P. J., 1991, Reliability and
[epub ahead of print] 1– 16. agreement of ratings of ataxic dysarthric speech samples with
HOEHN, M. and YAHR, M., 1967, Parkinsonism: onset, progression varying intelligibility. Journal of Speech and Hearing Research,
and mortality. Neurology, 17, 427– 442. 34, 285– 293.
KEARNS, K. P. and SIMMONS, N. N., 1988, Interobserver reliability THEODOROS, D., HILL, A., RUSSELL, T., WARD, E. and WOOTTON,
and perceptual ratings: more than meets the ear. Journal of R., 2008, Assessing acquired language disorders in adults via
Speech and Hearing Research, 31, 131– 136. the Internet. Telemedicine Journal and e-Health,, 14,
KREIMAN, J. and GERRATT, B. R., 1998, Validity of rating scale 552– 559.
measures of voice quality. Journal of the Acoustical Society of YORKSTON, K. M. and BEUKELMAN, D. R., 1981, Assessment of
America, 104, 1598– 1608. Intelligibility of Dysarthric Speech (Austin, TX: PRO-ED).
LANDIS, J. R. and KOCH, G. G., 1977, The measurement of
observer agreement for categorical data. Biometrics, 33,
159– 174.
MARTINEZ-MARTIN, P., GIL-NAGEL, A., GRACIA, L. M., GOMEZ,
J. B., MARTINEZ-SARRIES, J. and BERMEJO, F., 1994, Unified

Parkinson’s Disease Rating Scale characteristics and Appendix: Online Participant Satisfaction Questionnaire
structure. The Cooperative Multicentric Group. Movement This questionnaire has been developed to determine your
Disorders, 9, 76 – 83. satisfaction with your assessment across the Internet. Please
MORRANT, D. G. and SHAW, W. C., 1996, Use of standardized CIRCLE the answer that you feel is most appropriate based on
video recordings to assess cleft surgery outcome. Cleft Palate your experience.
Craniofacial Journal, 33, 134– 142.
MUTCH, W. J., DINGWALL-FORDYCE, I., DOWNIE, A. W., 1. How did you feel while participating in this session on the
PATERSON, J. G. and ROY, S. K., 1986, Parkinson’s disease Internet?
in a Scottish city. British Medical Journal (Clinical Research a) Would prefer this type of session to face-to-face session
Edn), 292, 534– 536. b) Very happy with this session
OXTOBY, M., 1982, Parkinson’s Disease Patients and Their Social c) Comfortable
Needs (London: Parkinson’s Disease Society). d) Uneasy

PALSBO, S. E., 2007, Equivalence of functional communication e) Would not participate again
assessment in speech pathology using videoconferencing.
Journal of Telemedicine and Telecare, 13, 40 – 43. 2. What is your opinion of the audio quality (what you were able
PETO, V., FITZPATRICK, R. and JENKINSON, C., 1997, Self-reported to hear) during the session?
health status and access to health services in a community a) Excellent
sample with Parkinson’s disease. Disability and Rehabilita- b) More than adequate
tion, 19, 97 – 103. c) Adequate
RAMIG, L. O., BONITATI, C. M., LEMKE, J. H. and HORII, Y., 1994, d) Inadequate
Voice treatment for patients with Parkinson disease: develop- e) Poor
ment of an approach and preliminary efficacy data. Journal of
Medical Speech–Language Pathology, 2, 191–209. 3. What is your opinion of the visual quality (what you were able
RAMIG, L. O., COUNTRYMAN, S., THOMPSON, L. L. and HORII, Y., to see) during the session?
1995a, Comparison of two forms of intensive speech a) Excellent
treatment for Parkinson disease. Journal of Speech and b) More than adequate
Hearing Research, 38, 1232– 1251. c) Adequate
RAMIG, L. O., FOX, C. and SAPIR, S., 2004, Parkinson’s disease: d) Inadequate
speech and voice disorders and their treatment with the Lee e) Poor
Silverman Voice Treatment. Seminars in Speech and Language,
25, 169–180. 4. Please rate your overall satisfaction with the Internet session.
RAMIG, L. O., PAWLAS, A. A. and COUNTRYMAN, S., 1995b, The a) Very satisfied
Lee Silverman Voice Treatment (LSVTw): A Practical b) More than satisfied
Guide to Treating the Voice and Speech Disorders in Parkinson c) Satisfied
Disease (Iowa City, IA: National Centre for Voice and d) Less than satisfied
Speech). e) Not at all satisfied
INT. J. LANG. COMM. DIS., 2010,
VOL. 00, NO. 0, 1–16
Research Report
Treating disordered speech and voice in Parkinson’s disease online:
a randomized controlled non-inferiority trial
Gabriella Constantinescu†, Deborah Theodoros†, Trevor Russell†, Elizabeth Ward†, Stephen Wilson‡
and Richard Wootton§
†School of Health and Rehabilitation Sciences University of Queensland, Brisbane, Australia
‡School of Information Technology and Electrical Engineering, University of Queensland, Brisbane, Australia
§Scottish Centre for Telehealth, Aberdeen, UK
(Received 28 September 2009; accepted 8 April 2010)
Abstract
Background: Telerehabilitation may be a feasible solution to the current problems faced by people with Parkinson’s
disease in accessing speech pathology services.
Aim: To investigate the validity and reliability of online delivery of the Lee Silverman Voice Treatment (LSVTw) for
the speech and voice disorder associated with Parkinson’s disease.
Method & Procedures: Thirty-four participants with Parkinson’s disease and mild-to-moderate hypokinetic
dysarthria took part in the randomized controlled non-inferiority laboratory trial and received the LSVTw in either
the online or the face-to-face environment. Online sessions were conducted via two personal computer-based
videoconferencing systems with real-time and store-and-forward capabilities operating on a 128 kbit/s Internet
connection. Participants were assessed pre- and post-treatment on acoustic measures of mean vocal sound pressure
level, phonation time, maximum fundamental frequency range, and perceptual measures of voice, articulatory
precision and speech intelligibility.
Outcomes & Results: Non-inferiority of the online LSVTw modality was confirmed for the primary outcome
measure of mean change in sound pressure level on a monologue task. Additionally, non-significant main
effects for the LSVTw environment, dysarthria severity, and interaction effects were obtained for all outcomes
measures. Significant improvements following the LSVTw were also noted on the majority of measures. The LSVTw
was successfully delivered online, although some networking difficulties were encountered on a few occasions.
High participant satisfaction was reported overall.
Conclusions & Implications: Online treatment for hypokinetic dysarthria associated with Parkinson’s disease appears
to be clinically valid and reliable. Suggestions for future research are outlined.
Keywords: Parkinson’s disease, telerehabilitation, Internet-based treatment, Lee Silverman Voice Treatment
(LSVTw), voice therapy.
What this paper adds

What is already known on this subject
Recent proof-of-concept studies have touched on the feasibility of online delivery of the LSVTw to people with
Parkinson’s disease. However, further large-scale studies are needed to demonstrate the full potential of online
treatment.
What this study adds

This study has demonstrated the validity and reliability of the online delivery of the LSVTw to people with
Parkinson’s disease. It lends further support to the use of online services for people with Parkinson’s disease in order
to overcome barriers to treatment access.
Address correspondence to: Gabriella Constantinescu, School of Health and Rehabilitation Sciences, University of Queensland, Brisbane 4072,
Australia; e-mail: gabriella@hearandsaycentre.com.au
International Journal of Language & Communication Disorders
ISSN 1368-2822 print/ISSN 1460-6984 online q 2010 Royal College of Speech & Language Therapists
http://www.informahealthcare.com
DOI: 10.3109/13682822.2010.484848
Introduction the delivery of the LSVTw via information technologies

and telecommunication. To date, only three proof-of-
Parkinson’s disease (PD) affects approximately 1% of
concept studies have investigated the feasibility of the
the population aged 65 years and above (de Rijk et al.
online delivery of the LSVTw for people with PD.
1997). The motor speech impairment associated with
PD is known as hypokinetic dysarthria, which manifests An earlier study conducted by the present research
in as high as 50 – 90% of people with PD at some stage group involved ten participants with mild to moderate-
of the disease (Ramig et al. 2004). The features of severe PD who received the online delivery of the
hypokinetic dysarthria include reduced loudness and LSVTw via PC-based videoconferencing (Theodoros
stress, monotony of pitch and loudness, imprecise et al. 2006). The results included significant improve-
articulation, variable rate and short rushes of speech, ments in mean sound pressure level (SPL) on sustained
inappropriate silences, and a harsh and breathy voice vowel phonation, reading and monologue tasks.
(Darley et al. 1969). For many people with PD the Significant improvements were also noted in maximum
overall effects of the hypokinetic dysarthria may have a fundamental frequency range and on the perceptual
significant and negative impact on their quality of life, measures of pitch and loudness variability, loudness
with many experiencing fewer communicative inter- level and breathiness. Additionally, high participant
actions and greater social isolation (Oxtoby 1982). satisfaction was reported with the online modality
To date, the Lee Silverman Voice Treatment (Theodoros et al. 2006).
(LSVTw) has been proven to be the most effective The feasibility of the online delivery of the LSVTw
treatment for PD in reducing the impact on via PC-based videoconferencing was also investigated
hypokinetic dysarthria on functional communication. in a smaller study by Howell et al. (2009). Three
The LSVTw is an intensive program that is delivered participants with PD and mild-to-moderate hypo-
over 4 weeks (1-h sessions, 4 days per week), with a kinetic dysarthria were treated online via a broadband
direct focus on improving vocal loudness by performing connection over Skype. Post-treatment improvements
all treatment tasks at high intensity and with maximum in mean SPL were noted on the sustained vowel
effort. Training loud phonation helps increase phonation, reading and monologue tasks.
respiratory and phonatory muscle drive, and it has Thirdly, Tindall et al. (2008) used videophones
been hypothesized that this approach may result in a to deliver the LSVTw remotely to the homes of
reduction of hypokinesia, bradykinesia and muscle 24 participants with PD and hypokinetic dysarthria
rigidity associated with PD (Ramig et al. 1994, 1995a). (severity level not specified). The authors reported
The additional focus on sensory awareness training or significant increases in SPL with treatment for measures
‘calibration’ with the LSVTw assists people with PD to of sustained vowel phonation, reading, monologue and
overcome the sensory processing deficit associated with picture description. High participant satisfaction with
their disease, through improved self monitoring and the use of videophones was also reported.
consistent use of the louder voice in daily communi- Although these studies have demonstrated the
cation (Ramig et al. 1994, 1995a, 1995b). Improve- potential of the online delivery of the LSVTw, a
ments in vocal loudness, quality and speech number of study limitations make it difficult to
intelligibility have been reported following the generalize the findings and to determine whether the
LSVTw delivery in the traditional face-to-face benefits obtained via the online delivery of the LSVTw
modality, with these results maintained up to 2 years are comparable with those obtained face-to-face. These
following treatment (Ramig et al. 1994, 1996, 2001a). limitations include the small sample sizes in the studies
Although the long-term results of the LSVTw have by Howell et al. (2009) and Theodoros et al. (2006), as
been promising, specific access barriers to speech well as the lack of a comparison face-to-face treatment
treatment for people with PD impact on the number of group within each study design. Furthermore, unlike
individuals that actually receive the LSVTw in Australia the study by Theodoros et al. (2006), the online systems
and worldwide. These barriers include: an inadequate of the remaining studies did not allow the SLPs to
number of LSVTw certified speech – language pathol- incorporate all essential elements of the LSVTw, in line
ogists (SLPs) to deliver the treatment; low caseload with face-to-face delivery. For the LSVTw ultimately to
priority for people with PD; and the physical difficulties be successful, the SLP must be able to monitor the
of people with PD which affect their ability to travel to participant’s vocal loudness (SPL) during all treatment
the services that may be at a distance from their place of tasks, in order to aid calibration (Ramig et al. 1995b).
residence (Grimm et al. 2004). In Tindall et al.’s (2008) study, an attempt at
The use of telerehabilitation as an additional or monitoring SPL was made, where participants
alternate service delivery model may help to improve themselves were required to position the SPL meter
access to speech treatment for people with PD, through during the session, so that it could be viewed by the SLP
Online voice treatment of Parkinson’s disease 3
over the videophone. However, this practice may have a good indication of the level of carryover of the louder
introduced a level of variability to treatment, as the voice to daily life, and a minimum improvement of
positioning of the SPL meter and its distance relative to 4.5 dB on this parameter following the LSVTw is
the participant’s mouth may have varied between considered clinically relevant (Ramig et al. 1995a,
sessions, potentially affecting the accuracy of the SPL 1996). The non-inferiority margin was set at 2.25 dB,
readings. In Howell et al.’s (2009) study, the which is half of the minimum improvement of 4.5 dB,
monitoring of SPL was not possible with videoconfer- as recommended by Jones et al. (1996). Due to the lack
encing and therefore, one treatment session each week of previously published studies comparing performance
was conducted face-to-face in order to monitor SPL, with the LSVTw between the online and face-to-face
build participant –clinician rapport and provide treat- environment at the time of the study, the sample size
ment materials. This mixed treatment modality was was calculated using preliminary data collected in this
therefore not truly representative of face-to-face study. Analysis of the first eight participants to receive
delivery. Both studies were also unable to transmit the online and face-to-face LSVTw in the current study
treatment materials to the participants via the online revealed a standard deviation (SD) of pre- to post-
systems, highlighting the potential difficulties of treatment difference scores of 2.48 dB. This value was
applying these methods to the home setting. included in the power calculation. Using an alpha level
There is a recognized need for remote delivery of of 0.05 and a statistical power of 80%, a minimum of
the LSVTw due to the intensive nature of the treatment 15 participants were required for each LSVTw
and the difficulties that arise for people with PD in environment (online and face-to-face). Allowing a
accessing this treatment. To help overcome these access 10% loss to follow-up, a total of 34 participants with
barriers and ensure best practice, there is a need for idiopathic PD and hypokinetic dysarthria took part in
online technology to closely replicate face-to-face the study. The participants (27 males, seven females)
treatment delivery and for research designs to were aged between 54 and 85 years (mean ¼ 70.12
incorporate comparison groups of face-to-face delivery. years, SD ¼ 8.56). Descriptive characteristics of the
The current study aimed to investigate the validity and participants are summarized in table 1. Eight of the
reliability of a PC-based treatment protocol for the participants included in this study have been previously
delivery of the LSVTw across the Internet with a large described in the proof-of-concept study by Theodoros
group of people with PD, and to compare it with et al. (2006). These participants have been highlighted
face-to-face treatment. The online application was in table 1.
specifically designed to replicate face-to-face delivery of All participants had been diagnosed with PD by a
the LSVTw as closely as possible. It was hypothesized neurologist experienced in movement disorders, with
that the LSVTw can be effectively delivered online and time post-diagnosis ranging from 8 months to 30 years
that treatment outcomes will be comparable with those (mean ¼ 6.16 years, SD ¼ 6.48). Stages of PD as per
obtained face-to-face. the Hoehn and Yahr (1987) Scale ranged from I to IV
with 27 participants rated at Stages I– II and seven
Method participants rated at Stages III– IV. The participants
were recruited from private neurologists in Brisbane,
Study design Queensland, public hospitals, and support groups of
This study was a single-blinded, prospective, random- Parkinson’s Queensland Incorporated. The criteria for
ized controlled non-inferiority trial. The study inclusion in the study were: the presence of
was conducted according to the extension of the hypokinetic dysarthria (ranging from mild to severe)
CONSORT guidelines for non-inferiority trials associated with diagnosed PD that was impacting on
(Piaggio et al. 2006). communication; a videolaryngoscopic evaluation of the
vocal fold structure and movement consistent with PD;
stimulability to increased loudness; and a consistent
Participants
drug regimen for PD. Proficiency in the use of
Ethical clearance was obtained prior to commencement computers was not a requirement for study inclusion as
of the study from the Behavioural and Social Sciences all aspects of the online treatment were delivered
Ethical Review Committee of the University of remotely by the online treating SLPs. The exclusion
Queensland, Brisbane. Prior to participant recruitment, criteria consisted of: speech and/or language difficulties
the sample size for a non-inferiority trial (Jones et al. unrelated to PD; an additional co-existing neurological
1996) was calculated using the study’s primary outcome disorder; respiratory difficulties inconsistent with PD;
measure of mean change in SPL (dB-C) following the cognitive difficulties resulting in the inability to
LSVTw on a 30s monologue task. An improvement in provide informed consent; a severe uncorrected visual
monologue loudness following the LSVTw provides and/or auditory disturbance; a history of alcohol abuse;
Table 1. Descriptive characteristics of participants who received the LSVTw in the online and face-to-face environments.
Participant Age (years) Sex Time post-diagnosis (years) Hoehn and Yahr (1967) Stage Dysarthria severity level
Participants who received the LSVT w online (n ¼ 17)
1 80 Male 2 1 Mild
2 56 Male 6 1 Mild
3 69 Male 2 1.5 Mild
4a 59 Male 4 1 Mild
5a 77 Male 2 1 Mild
6a 66 Male 1.5 1 Mild
7 71 Female 1 1 Mild
8a 54 Male 4 1 Mild
9 58 Male 0.7 1.5 Mild
10a 81 Male 7 2.5 Moderate
11 61 Male 2 1.5 Moderate
12 74 Male - 2.5 Moderate
13a 75 Female 14 3 Moderate
14 69 Male 22 2 Moderate
15a 85 Male 9 3.5 Moderate
16a 84 Male 1 1 Moderate
17 82 Female 8 1 Moderate
Group mean (SD) 70.65 (10.26) 5.39 (5.76) 1.59 (0.81)
Participants who received the LSVT w face-to-face (n ¼ 17)
18 68 Male 3 1 Mild
19 59 Male 11 1 Mild
20 67 Male 2.5 1 Mild
21 68 Male 1 1 Mild
22 66 Male 5.5 1 Mild
23 74 Female 1 1.5 Mild
24 75 Male 7 1 Mild
25 76 Female 5 1 Mild
26 69 Male 9 2 Mild
29 62 Female 4 2.5 Moderate
33 68 Female 3 1.5 Moderate
Group mean (SD) 69.59 (6.71) 6.88 (7.19) 1.53 (0.82)
Notes: aParticipants described by Theodoros et al. (2006).

– , Unspecified; SD, standard deviation.
and participation in the LSVTw within 12 months of treatment from the 30s monologue, which included
the present study. mean SPL (dB), and perceptual ratings of overall speech
Before the commencement of treatment, the intelligibility in conversation on a five-point scale
participants underwent a videolaryngoscopic examin- (1 ¼ normal, completely intelligible speech, 5 ¼ severely
ation by an Ear Nose and Throat specialist to examine unintelligible speech with difficulties deciphering many
the vocal fold structure and movement. The examin- words). Sound pressure levels above 65 dB in monologue
ations excluded any laryngeal pathology inconsistent loudness, together with a mild reduction in overall speech
with the effects of PD. Evaluations identified bowed intelligibility were classified as mild hypokinetic
vocal folds in 14 participants, vocal tremor in one dysarthria. Sound pressure levels ranging from 60 to
participant and slight supraglottal constriction in 65 dB with an accompanying mild-to-moderate
another. For the remaining 18 individuals, the reduction in speech intelligibility corresponded to
examinations were unremarkable. All participants were moderate hypokinetic dysarthria. A severe hypokinetic
cleared for inclusion in the treatment phase. dysarthria level was considered to be consistent with SPL
Before treatment, all participants were classified by values below 60 dB and a moderate or severe reduction
the principal investigator on the basis of an overall in speech intelligibility. Using this classification, 18
severity level for hypokinetic dysarthria. This classifi- participants in the study demonstrated mild hypo-
cation was made using baseline data collected prior to kinetic dysarthria, and the remaining 16 participants
Figure 1. Schematic flow chart of participants through the study.

demonstrated moderate dysarthria (table 1). Based on online environment and seven face-to-face; SLP 2
their dysarthria severity classification, the participants delivered five treatments online and seven face-to-face;
were then stratified, and randomly assigned by the SLP 3 delivered two treatments online and face-to-face;
principal investigator to receive the LSVTw in either the and SLP 4 delivered a single treatment online and
traditional face-to-face or online treatment environment. face-to-face.
Randomization occurred using a computerized random-
number generator (Dallal 1997). As the participants
entered the study, randomization codes were generated The LSVTw program
and assigned by the principal investigator to the All online and face-to-face laboratory treatment sessions
participants in their order of recruitment to the study. were conducted within the Telerehabilitation Research
This process resulted in 17 participants in each treatment Unit at the University of Queensland. The treatment
environment. Each group was comprised of nine was delivered intensively, 1-h per day, 4 days per week
participants with mild dysarthria and eight participants over a 4-week period, in accordance with the LSVTw
with moderate dysarthria (table 1). Figure 1 is a schematic program (Ramig et al. 1994, 1995a, 1995b). Daily
flow chart of the participants through the study. sessions consisted of repetitive drills including maxi-
mum sustained vowel phonation, maximum funda-
mental frequency range, functional speech loudness
Speech – language pathologists
drills and hierarchical speech loudness tasks that were
Four experienced and LSVTw certified SLPs from the performed at high intensity. The LSVTw protocol aims
University of Queensland conducted the treatment to promote increased respiratory drive, vocal fold
sessions. The SLPs were randomized to both the adduction and carryover of the louder voice into
treatment environments and the participants within functional communication. During the treatment tasks,
each environment using a computerized random- the SLP monitored the participant’s vocal loudness (dB)
number generator (Dallal 1997). Due to other work and quality, and aided calibration (that is, the
commitments, the SLPs were randomly allocated to the participant’s ability to self-monitor and consistently
study within bounds of their availability. The same SLP use their louder voice in everyday communication).
delivered the entire treatment program for a partici- Participants also practised the daily tasks and
pant. In total, SLP 1 delivered nine treatments in the hierarchical speech loudness drills at home on a daily
basis. This homework encouraged further carryover of link at the SLP site. At the other site, the participant was
the loud voice into the participant’s everyday seated comfortably at approximately 50 cm from the
environment. PC monitor and wore a headset microphone at a set
distance of 5 cm from the corner of the mouth.
Treatment environment To maintain the accuracy of the online acoustic data
(SPL, fundamental frequency and duration), the
Online LSVTw environment microphone distance was measured at the beginning
A PC-based videoconferencing application developed of each session by the SLP, and the online SPL values
at the University of Queensland was used for the obtained on a sample of at least two sustained /a/
online treatments. Videoconferencing was established phonations were verified against the values obtained on
over a 128 kbit/s Internet connection, which at the a conventional SPL meter (Radio Shackw Model No.
time of the study was consistent with the minimum 23-553). All aspects of treatment were delivered
bandwidth used in the public health systems of remotely by the SLP. Clinician training on the use of
Queensland. The specific features of the application the online application for the LSVTw delivery was
designed to deliver the LSVTw included the ability to: conducted in a 3-h session before the commencement
(1) conduct videoconferencing in real-time; (2) of the study by the principal investigator. The clinicians
present functional phrases and reading material for were familiarized with the specific features of the online
the hierarchical speech tasks on the participant’s system that were necessary for the delivery of each
screen; (3) adjust the remote web cameras for LSVTw daily task. Clinicians were deemed competent
maximum viewing of the participant; (4) capture with online delivery of the LSVTw when they could
high quality audio (windows media audio CODEC adequately deliver a mock treatment session to face-to-
Version 8 at 368 kbit/s) and video recordings face standard, within a 1-h time frame. The online
(640 £ 480 pixel resolution compressed with the treatment environment is represented in figures 2 and 3.
windows media video CODEC Version 8 at
384 kbit/s) as a separate function to videoconferen- Face-to-face LSVTw environment
cing, and the ability to review the files when needed
via the store-and-forward capabilities of the system; The face-to-face LSVTw sessions were conducted in
and (5) view and sample calibrated average measures accordance with standard practice (Ramig et al. 1994).
of SPL (all measured as dB-C), fundamental frequency In this environment, the participant and clinician were
(Hz) and duration (s) during the daily tasks and seated at a table, facing each other. A SPL meter (Radio
speech loudness drills in real-time using the system’s Shackw Model No. 23-553) was kept at a set distance of
acoustic speech processor that was specifically devel- 30 cm from the participant’s mouth and was used by
oped for this study. the SLP to monitor the SPL throughout the session. For
Validation trials using the acoustic speech processor the duration of sustained vowel phonation and
were conducted prior to the commencement of the fundamental frequency measures, a generic stopwatch
study using pure tones generated by a Function Generator and a chromatic tuner (KORG Model DT-7) were
(Topward Electronic Instruments Model TFG-462). used, respectively. For the hierarchical speech loudness
The measures obtained using the acoustic speech
processor at SPLs of 55 dB to 95 dB and fundamental
frequency levels of 100 to 975 Hz were compared with
the Visi-Pitch II (Kay Elemetrics Model No. 3300).
Paired t-tests analyses showed no significant differences
in the pure-tone SPL ( p ¼ 0.939) and pitch data ( p ¼
0.988) between the two instruments. Further validation
trials of the acoustic speech processor against a Digital
Sound Level Meter (Radio Shackw Model No. 23-553)
were conducted for SPL measures using voice samples
(66–91 dB). Paired t-tests analyses also showed no
significant differences ( p ¼ 0.980) between the two
instruments.
For online treatment, two videoconferencing
systems utilizing the customized application were
located in separate rooms and on different floors of Figure 2. Online LSVTw environment at a participant site: (1)
the building at the University of Queensland, with the videoconferencing system displaying the clinician; (2) web cameras;
SLP delivering the sessions over the videoconferencing and (3) system’s acoustic speech processor.
treating clinician for a particular participant, they were
not allocated to the post-LSVTw assessment. During the
post-treatment assessments, the participants were asked
not to divulge their treatment environment to the
assessors.
Acoustic measures
The LSVTw Evaluation Protocol (Ramig et al. 1995b)
was used to assess SPL, phonation time and maximum
fundamental frequency range during a number of
tasks. The protocol has been routinely used in the
LSVTw literature for pre- and post-treatment assess-
ment. For the acoustic measures, the online system’s
acoustic speech processor was used to obtain measures
Figure 3. Online LSVTw environment at SLP site: (1)
of mean SPL (dB), fundamental frequency (Hz) and
videoconferencing system displaying the participant; (2) pitch duration (s) during the pre- and post-treatment
(Hz) and sound pressure level (SPL) (dB-C) data via the system’s assessments. In order to do so, each participant wore
acoustic speech processor; and (3) the web camera. the headset microphone used in the online treatment
sessions, with the microphone distance measured at
5 cm from their mouth. The acoustic measures were
displayed on a PC in the assessment room and were
sampled by the assessing SLP. During the assessments,
the participants were instructed to perform all tasks in a
comfortable voice and were not prompted by the SLPs
to monitor their loudness level, phonation time or
frequency range.
Mean SPLs (dB) were recorded on three assessment
tasks. These included six maximum phonations of /a/,
reading of The Grandfather Passage (Darley et al. 1975)
and a 30s monologue. The duration of the sustained
vowel phonations was also recorded in seconds.
The maximum fundamental frequency range was
calculated for each participant from the difference
between the highest and lowest fundamental frequency
Figure 4. Face-to-face LSVTw environment: (1) use of the sound mean levels (Hz) obtained during the performance of
pressure level (SPL) meter; (2) reading material; and (3) homework six vocal glides to the highest and lowest levels,
exercises.
respectively. The range in hertz (Hz) was then converted
drills, the same reading material used online was to a maximum semitone range (ST) (de Pijper 2007).
provided in hard copy. The face-to-face treatment
environment is represented in figure 4. Perceptual speech and voice parameters
Pre- and post-LSVTw readings of the Rainbow Passage
(Fairbanks 1960) were rated for each participant on
Assessment of outcomes
the parameters of breathiness, roughness and overall
The participants were assessed pre- and post-LSVTw in articulatory precision. In addition, the 30s monologue
the face-to-face environment on the primary outcome performed by each participant for the acoustic loudness
measure of mean change in SPL during the 30s measure was also rated according to loudness level,
monologue, and secondary measures on a number of pitch and loudness variability, and overall speech
acoustic and perceptual parameters that have been used intelligibility in conversation. Perceptual ratings of the
in the LSVTw literature as sensitive predictors of speech and voice parameters were made using Direct
treatment change (Ramig et al. 1996, 1995a). Three Magnitude Estimation (DME), a method of scaling
SLPs were involved in the assessments and were that allows the listener to rate a speech or voice
randomized to conduct the pre- and post-treatment parameter numerically, with no limits to the endpoints
assessments. Only one SLP was required for each of the scale. Together with the use of a ‘standard’
assessment. If, however, one of the SLPs had been the or an external reference of the speech or voice
parameter in question to rate against, DME scaling The reliability of the DME method was also
allows for improved validity and reliability (Schiavetti determined for the perceptual voice parameters as a
et al. 1981, Stevens 1975, Weismer and Laures 2002). group (breathiness, roughness, loudness level, loudness
To obtain the speech samples for DME rating, pre- variability and pitch variability), overall articulatory
and post-LSVTw recordings of the Rainbow Passage precision and overall speech intelligibility in conversa-
and 30s monologues were made using a minidisk tion. Intra- and inter-rater reliability of the logarithmic
recorder. The middle three sentences of the Rainbow values for these parameters was calculated using intra-
Passage were used for the ratings. The pre- and post- class correlation coefficients (ICC(2,1)) for 47% (n ¼ 16)
LSVTw recordings were numerically coded and of participants. Half of the data for analyses were taken
randomized to CD using a computerized random- from the participants’ pre- and post-LSVTw assess-
number generator (Dallal 1997). The standards for ments, respectively. ICC values below 0.40 corre-
each parameter were also added to the CD and these sponded to poor-to-fair reliability; between 0.40 and
were readings and monologues of additional speakers 0.75 to moderate-to-good reliability; and values above
with PD who had not taken part in treatment. 0.75 represented very good reliability (Fleiss 1981).
A different standard was used for each speech feature In addition to the DME ratings for overall speech
and represented a moderate degree of impairment for intelligibility in conversation, the Assessment of
each parameter (Stevens 1975, Weismer and Laures Intelligibility of Dysarthric Speech (ASSIDS) (Yorkston
2002), as judged by the investigators. and Beukelman 1981) was used to measure word and
The DME ratings were made by two independent sentence intelligibility and communication efficiency.
SLPs who had not taken part in the assessments or The participants were required to read or repeated
delivery of the LSVTw, and who were blinded to the 50 words and 22 sentences of increasing length that had
intent of the study. Rater training was conducted for been randomized prior to the assessment. The pre- and
30 min before the ratings using additional stimuli that post-treatment audio recordings for each participant
were then randomized and rated by the two independent
were not included in the data analyses. A definition of
and blinded SLPs that had performed the DME ratings.
each parameter was provided to the raters to assist them
The two ratings made by the SLPs were then averaged to
in identifying the speech feature. Following standard
express a single pre- and post-treatment mean value.
DME procedures, one parameter was rated at a time
These values were represented as percent correct for the
and the raters first listened to the standard, which was
word and sentence intelligibility tasks. To calculate the
then repeated after every four speech samples (Whitehill communication efficiency ratio, the participant’s
et al. 2002). If a sample needed to be heard a second number of intelligible words per minute was divided
time, the rater would again listen to the standard and by that expected for normal speakers (190 words per
then the four speech samples in that set. The volume minute) (Yorkston and Beukelman 1981).
was also kept at a consistent level for all speech samples.
For rating, a set value of 100 was assigned to each Participant satisfaction questionnaire
standard (Schiavetti et al. 1981, Whitehill et al. 2002),
and raters assigned values greater than 100 to denote In addition to the perceptual and acoustic measures
improvement for measures of loudness level, pitch and obtained, the participants who had received the online
loudness variability, overall articulatory precision and delivery of the LSVTw completed a satisfaction
overall speech intelligibility in conversation. For these questionnaire post-treatment. A five-point rating scale
parameters, a value of 200 indicated that the sample was was used to assess satisfaction with the online treatment
rated as twice as improved in quality as the standard, sessions, the audio and video quality and overall
while a value of 50 denoted that the sample was only satisfaction with online treatment. For the questionnaire,
half as clear in quality compared with the standard see the appendix.
(Weismer and Laures 2002). In contrast, for the
breathiness and roughness parameters, values greater Statistical analyses
than 100 denoted increased impairment in vocal The primary analysis (non-inferiority of the online
quality. Following the DME ratings, the values given by LSVTw modality) was evaluated by firstly computing
the two SLPs were averaged to express a single mean the within-group differences from baseline on the
value for each sample. As the DME ratings are primary outcome measure of mean SPL on the
considered to have a log normal distribution (Engen monologue task, and the 95% confidence interval (CI)
1971, Stevens 1975), the means were then converted to of these differences. It was then determined whether the
logarithmic values for inclusion in the statistical CI, centred on the observed difference between the
analyses, and represented as geometric means and online and face-to-face LSVTw groups, did lie entirely
standard deviations. between the non-inferiority margin (^2.25 dB) and
zero. Statistical analyses of the pre- and post-LSVTw
0.894 10.29 0.266 197.17 , 0.001*
0.867 94.15 , 0.001*

0.599 65.17 , 0.001*
0.046*
data were then performed individually for each
0.598
parameter using repeated-measures general linear
Notes: pre ¼ pre-LSVTw, post ¼ post-LSVTw; " ¼ improvement with treatment; * ¼ significant effect; severity ¼ hypokinetic dysarthria severity; time ¼ pre- to post-LSVTw; Env ¼ environment; change ¼ treatment change.
Time
model (GLM). Time (pre- to post-LSVTw) was the
within-subjects variable, while the mode of the LSVTw
0.814 0.285
4.31
delivery (online or face-to-face environment) and
F
participant severity level (mild or moderate dysarthria)
0.877
LSVT Env £
were the between-subjects variables. The interactions
Severity £
p
between the mode of the LSVTw delivery, participant
Time
severity level and time were also calculated using GLM.
0.03
0.28
0.06
0.03
F
For the analyses, the level of significance was set at
p , 0.05.
0.003 0.956 10.51 0.228

" 3.88 (3.24) , 0.001 0.984 20.52 0.123
0.60 0.810 , 0.001 0.998
0.871
p
Severity
Results
0.03
0.02
For the primary outcome measure of mean change in
F
Acoustic parameters: means (standard deviations), F and p values.
SPL with LSVTw on the monologue task, the upper
95% CI of the within-group differences from baseline
0.940
0.880
LSVT Env
was 1.41. This value was found to lie entirely between
p
the non-inferiority margin of ^2.25 dB and zero.
0.02
Therefore, non-inferiority of the online LSVTw
0.01
modality was confirmed.
F
" 10.3 (4.42)
" 4.97 (3.07)
" 0.24 (4.84)
" 2.29 (4.94)

Acoustic measures
Change
Table 2 displays the pre- and post-LSVTw values,
standard deviations, changes from baseline, F and p
Face-to-face LSVT
values on the acoustic measures. For all acoustic

" 10.29 (4.01) 73.45 (5.49) 83.75 (4.38)
67.84 (3.23) 72.81 (3.65)

66.74 (3.08) 70.62 (3.68)
11.08 (5.22) 11.32 (5.15)
9.69 (5.01) 11.98 (4.92)

measures (SPL parameters, duration of phonation and
Post
maximum fundamental frequency range), statistic

analyses using repeated-measures GLM showed non-
significant main effects for the LSVTw environment,
dysarthria severity and interaction effects ( p . 0.05).
Statistically significant increases in mean SPL on all
Pre
parameters and maximum fundamental frequency

range ( p , 0.05) were evident with time (pre- to
post-LSVTw). However, for the duration of phonation
" 5.04 (2.85)
" 3.87 (2.27)
" 0.61 (4.09)
" 2.68 (8.11)
parameter, there was no significant increase evident

Change
Table 2.
with time for either treatment environment.
Perceptual measures
Online LSVT
73.49 (4.55) 83.78 (3.88)
68.81 (2.86) 73.84 (2.52)

67.39 (3.01) 71.25 (2.75)
10.65 (4.57) 11.26 (4.71)
11.04 (5.91) 13.71 (4.82)
The pre- to post-LSVTw values, standard deviations, F

Post
and p values on the perceptual measures are displayed in

table 3. Analyses using repeated-measures GLM
revealed non-significant main effects ( p . 0.05) for
the LSVTw environment, participant severity level and
interaction effects on the LSVTw outcomes for all
Pre
perceptual measures (perceptual voice parameters,

overall articulatory precision and measures of intellig-
ibility). Statistical analyses did reveal a significant effect
Monologue (dB)
of phonation (s)
of time (pre- to post-LSVTw) for the majority of

Sustained vowel
phonation (dB)
Reading (dB)
fundamental
perceptual measures. This effect included improve-

range (ST)
Maximum
parameter
frequency
Duration
ments pre- to post-treatment on all the perceptual voice

Acoustic
parameters (breathiness, roughness, loudness level,

loudness variability and pitch variability), the word
and sentence intelligibility tasks of the ASSIDS
, 0.001*
0.019*
0.002*
, 0.001*
, 0.001*
0.001*
0.001*
0.001*
Notes: pre ¼ pre-LSVTw; post ¼ post-LSVTw; * ¼ significant effect; severity ¼ hypokinetic dysarthria severity; time ¼ pre- to post-LSVTw; Env ¼ environment; OAP ¼ overall articulatory precision; OIC ¼ overall speech intelligibility in
0.243
0.067
assessment, and overall speech intelligibility in
p
conversation. Overall articulatory precision and the
Time
communication efficiency ratio determined from the
ASSIDS were the only perceptual parameters that did
26.21
6.14
11.18
73.70
92.87
1.42
13.05
13.92
13.69
3.63
F
not improve significantly with time ( p , 0.05).
For the perceptual voice parameters using DME
scaling, intra-class correlations revealed very good intra-
0.473
0.101
0.653
0.429
0.139
0.051
0.877
0.849
0.750
0.492
Severity £ Time
LSVT Env £
rater (ICC ¼ 0.96 Rater 1; ICC ¼ 0.86 Rater 2) and
p
inter-rater reliability (ICC ¼ 0.86). Moderate intra-
rater (ICC ¼ 0.55 Rater 1; ICC ¼ 0.50 Rater 2) and
0.024
0.53
2.88
0.21
0.65
2.32
4.14
0.04
0.10
0.49
very good inter-rater reliability (ICC ¼ 0.82) was
F
achieved for overall articulatory precision. Additionally,
for overall intelligibility in conversation, very good
0.737
0.557
0.894
0.106
0.378
0.703
0.131
0.243
0.157
0.451
intra-rater (ICC ¼ 0.89 Rater 1; ICC ¼ 0.96 Rater 2)
p
Severity
and moderate inter-rater reliability (ICC ¼ 0.44) was
Perceptual parameters: means (standard deviations), F and p values.

obtained.
0.12
0.35
0.02
2.79
0.80
0.15
2.42
1.42
2.11
0.58
F
0.523
0.166
0.522
0.303
0.636
0.216
0.119
0.189
0.182
0.972
p
On the participant satisfaction questionnaire, the
LSVT En
majority of participants who received the online
delivery of the LSVTw were equally very happy or
0.001
0.42
2.02
0.42
1.10
0.23
1.60
2.58
1.81
1.87
F
comfortable (47.07% for each) while participating in
the online sessions. The audio quality during
conversation; WI ¼ percentage word intelligibility; SI ¼ percentage sentence intelligibility; CER ¼ communication efficiency ratio.
videoconferencing was largely rated as adequate
63.53 (11.24)
84.84 (10.90)
201.23 (21.78)
146.25 (14.13)
110.48 (14.20)
128.03 (18.72)
123.82 (21.01)
92.53 (9.68)
98.31 (3.38)
0.84 (0.26)
(41.18%) or more than adequate (35.29%), while
Post
there was greater variability in the video quality ratings,
Face-to-face LSVT
with responses ranging from more than adequate

(17.65%), to adequate and inadequate (32.29% for
each). Overall, participant satisfaction with online
treatment ranged from very satisfied (29.41%), to more
84.12 (17.14)
97.10 (12.48)
157.07 (10.70)
125.40 (14.89)
101.20 (16.42)
127.91 (12.36)
108.49 (19.20)
88.47 (12.45)
95.73 (5.87)
0.89 (0.22)
than satisfied (52.94%) and satisfied (17.65%).
Pre
Online LSVTw delivery

For all 17 participants, the LSVT w program
85.33 (11.74)
195.16 (21.12)
115.77 (11.19)
142.00 (11.86)
123.85 (11.97)
was successfully delivered online using PC-based

Table 3.
62.14 (9.59)
153.43 (7.37)
96.31 (2.47)
99.23 (1.22)
0.96 (0.21)
videoconferencing operating on a 128 kbit/s Internet

Post
connection. Out of the 272 sessions delivered during

Online LSVT
the study, none failed for any of the participants.

Additionally, for seven of the 17 participants, all of their
sessions ran very smoothly, without technical difficulties
77.20 (12.52)
88.39 (10.50)
158.74 (10.81)
127.20 (10.63)
100.00 (11.88)
132.74 (12.83)
117.60 (12.73)
and with adequate audio and video quality for treatment

94.38 (4.65)
98.02 (2.38)
1.01 (0.22)
delivery. The remaining ten participants encountered

Pre
some difficulties during one of their sessions. These

difficulties included: some network congestion that
considerably compromised the audio and video quality
(five sessions); a faulty headset microphone which made
Perceptual parameter
Loudness variability
it difficult for the clinician to hear the participant

(one session); and the acoustic speech processor
Pitchvariability
Loudness level
malfunctioned and the SPL and fundamental frequency

Breathiness
Roughness
data could not be obtained (four sessions). However, all

difficulties were appropriately addressed and the
OAP
CER
OIC
WI
sessions were able to continue.

SI
Discussion mild-to-moderate hypokinetic dysarthria (Ramig et al.
1995a, 1996). Efficacy studies by Ramig et al. (1995a
For the primary outcome measure of mean change in
1996) have reported mean improvements in SPL
SPL on the monologue task with the LSVTw, the 95%
following the LSVTw of 13.00 and 14.03 dB for
CI of the within-group differences from baseline were
sustained vowel phonation, 8.03 and 8.75 dB for
within the non-inferiority margin, thus confirming
reading loudness, and 4.50 and 4.68 dB conversational
non-inferiority of the online LSVTw modality.
loudness, respectively. However, the clinical relevance
Furthermore, the results obtained in the present
of the treatment gains in the present study should be
study on the secondary outcome measures supported
interpreted alongside the pre- and post-treatment mean
the study hypothesis that the online LSVTw is valid
SPL values. On the majority of tasks, the participants in
and reliable. For the participants in both the online and
both the online and face-to-face LSVTw groups
face-to-face LSVTw environments, non-significant
achieved higher pre- and post-treatment mean SPL
effects of treatment environment, participant severity
values than those participants in previous studies.
level, and the interaction of these factors were found for
Additionally, the post-treatment mean SPL values in
all the acoustic and perceptual measures post-LSVTw.
the current study were comparable with the values
The treatment gains for the majority of the parameters
reported for a group of healthy older adults speaking at
were statistically significant with time and comparable
comfortable loudness levels in the face-to-face setting
between the two LSVTw environments. While findings
(Fox and Ramig 1997, Ramig et al. 2001a). It is
indicated that comparable treatment outcomes could
acknowledged that comparisons were made to control
be achieved in the online and face-to-face environ-
data involving smaller samples sizes (n ¼ 14 healthy
ments, the study also revealed that the online LSVTw
participants in each study). However, these studies were
was clinically effective in improving the speech and
similar in their assessment procedures to the current
voice of people with PD. Indeed, significant improve-
study. It was therefore felt that comparisons with the
ments were identified across the majority of acoustic
control data were appropriate and provided a good
and perceptual parameters.
indication of treatment success and clinical significance
in the present study.
Acoustic measures
SPL tasks
Sustained vowel duration
For the sustained vowel phonation, reading and
monologue tasks, statistical analyses disclosed signifi- For sustained vowel duration, statistically significant
cant improvements in mean SPL with the LSVTw for main effects of time (pre- to post-LSVTw) were not
both the online and face-to-face PD participants. observed for either the online or face-to-face PD
These findings are consistent with the statistically participants. Overall, the treatment changes for both
significant improvements reported on such SPL tasks environments (mean ¼ 0.61 s, SD ¼ 4.09 online;
in previous LSVTw studies conducted face-to-face mean ¼ 0.24 s, SD ¼ 4.84 face-to-face) were markedly
with PD participants of similar dysarthria severity lower than the 3.72 s and 4.90 s mean improvement
levels (El Sharkawi et al. 2002, Ramig et al. 1995a, reported by Ramig et al. (1994 1995a) for PD
1996, 2001a, 2001b). It has also been suggested that participants with mild-to-moderate hypokinetic dysar-
these changes in SPL contribute to the perceived thria. The post-treatment mean duration values
improvements in voice parameters and speech (mean ¼ 11.26 s, SD ¼ 4.71 online; mean ¼ 11.32
intelligibility (Baumgartner et al. 2001, Ramig et al. s, SD ¼ 5.15 face-to-face) in the present study were
1994, 1995a Sapir et al. 2002), which were observed also lower than the 17.94 s (SD ¼ 5.01) mean
in the present study. recording for healthy older adults, which was the
The treatment gains in mean SPL in the present longest duration of their three trials (Fox and Ramig
study were comparable between the online and face-to- 1997). It is possible that the absence of change in the
face LSVTw groups for sustained vowel phonation, duration of sustained phonation is related to the
reading and monologue loudness. However, these participants’ focusing mainly on vocal quality during
changes (sustained vowel phonation mean ¼ 10.29 dB the task rather than duration, and as a result, did not
online, mean ¼ 10.3 dB face-to-face; reading loudness demonstrate a significant change in duration in
mean ¼ 5.04 dB online, mean ¼ 4.97 face-to-face; conjunction with an increase in loudness. The clear
conversational loudness mean ¼ 3.87 dB online, quality of the post-treatment phonations for partici-
mean ¼ 3.88 dB face-to-face) were lower than the pants in both groups, compared with the pitch breaks,
minimal improvements in mean SPL reported with roughness and breathy vocal qualities perceived pre-
the face-to-face LSVTw for PD participants with treatment, further suggests that participants learned to
self-monitor the quality of their phonations as part of strength and endurance, and tongue-base swallowing
calibration, which was an important part of treatment. function (Dromey et al. 1995, El Sharkawi et al. 2002,
Sapir et al. 2007, Ward et al. 2000). In the present
Maximum fundamental frequency range study, only minor increases in articulatory precision
were perceived for the participants. This finding may
On this task, statistically significant improvements with
suggest that the perceptual rating process for this
the LSVTw were achieved for the PD participants in
parameter was unable to detect changes in articulatory
both treatment environments. This finding is in keeping
precision with treatment. This may be due to the fact
with the significant gains reported perceptually for pitch
that pre-LSVTw DME values were already high and
variability in conversation in the present study, and face-
above the standard. It is possible that participants in
to-face studies for maximum fundamental frequency
general, did not present with obvious impairments in
range and variability on reading, monologue tasks and
articulatory precision per se prior to treatment, and as
vocal glides (Ramig et al. 1994, 1995a, 1996, 2001a).
such, substantial post-treatment changes were not
Although significant gains were achieved on this
identified. In future studies, physiological investigations
parameter, the post-treatment mean increases
of tongue function with instrumentation such as
in maximum fundamental frequency range (mean ¼
electromagnetic articulography may provide more
2.67 ST, SD ¼ 8.11 online; mean ¼ 2.29 ST,
conclusive evidence of treatment specific changes in
SD ¼ 4.94 face-to-face) in the present study were
articulatory precision.
lower than the 4 ST minimum improvements reported
by Ramig et al. (1994).
Speech intelligibility
Perceptual measures
On the evaluation of overall speech intelligibility in
Perceptual voice parameters conversation using DME scaling, statistically significant
improvements with treatment were found for the PD
Participants in the online and face-to-face treatment
participants in both treatment environments. These
environments showed significant improvements on all
results are in keeping with the improvements in
of the perceptual voice parameters following treatment
conversational speech intelligibility in previous face-to-
(table 3). These changes included perceived reductions
face LSVTw studies, which were perceived by assessing
in the levels of breathiness and roughness in reading,
clinicians, participants and their families using visual
and increases in pitch variability and loudness level and
analogue scales (El Sharkawi et al. 2002, Ramig et al.
variability in conversational speech. The results are in
1994, 1995a). The improvement in speech intellig-
keeping with the literature that has reported significant
ibility is also consistent with the significant improve-
post-LSVTw improvements in breathiness, hoarseness,
ments noted on the word and sentence intelligibility
loudness level and monotonicity for PD participants
tasks of the ASSIDS.
with mild-to-moderate dysarthria who were treated in
On the ASSIDS assessment, participants in both
the conventional manner. These improvements were
treatment environments made statistically significant
perceived by the participants themselves, their families,
gains in word and sentence intelligibility with
and assessing clinicians (Baumgartner et al. 2001,
treatment. Moreover, the post-LSVTw sentence scores,
Ramig et al. 1994, 1995a, Sapir et al. 2002). As the
which have been suggested by Yorkston and Beukel-
previous studies have used visual analogue scales rather
man (1981) to be the most representative of functional
than DME to determine treatment changes, as well as
speech out of all ASSIDS tasks, demonstrated that
participant and family ratings, direct comparisons of
participants in both treatment environments had
results with the current study are difficult.
improved in speech intelligibility from relatively mild
impairment to near normal levels. Such findings of
Overall articulatory precision
improved speech intelligibility with the LSVTw are
Significant gains in overall articulatory precision with consistent with a face-to-face study of a group of 18 PD
treatment were not achieved for the participants in participants with mild to severe hypokinetic dysarthria
either environment. This finding is inconsistent with who made significant gains in sentence intelligibility
previous studies of PD participants with predominantly post-LSVT w, approaching the normal range
mild-to-moderate dysarthria which have highlighted (Ward et al. 2000). Improved intelligibility with the
changes in articulatory function following the LSVTw. LSVTw has been associated with the global effects
These post-LSVTw changes were measured as improve- of loud phonation, where enhanced motor drive
ments in: extent and rate of movement; vowel and and coordination of movement throughout the
whole word duration; transition duration; rise time; speech subsystems are perceived at this overarching
vowel formants; frication duration; enhanced tongue level (Dromey and Ramig 1998, Ward et al. 2000).
Although the treatment gains for word and sentence Benefits and challenges of online LSVTw delivery
intelligibility were significant in the present study, the The features of the online application were conducive
levels of improvement fell within the levels of test– to treatment delivery. The 128 kbit/s videoconferencing
retest variability reported in the ASSIDS manual for rate provided sufficient audio and video quality for the
dysarthric speakers assessed in the face-to-face treatment sessions. Additionally, the features of the
environment (Yorkston and Beukelman 1981). This online application including the real-time display and
is likely due to the ceiling effect of the ASSIDS where it sampling of average recordings of SPL (dB), duration
has been suggested that ongoing recovery for speakers (s) and fundamental frequency data (Hz), and the
with mild dysarthria and those with sentence store-and-forward facility, incorporated all the essential
intelligibility scores above 90% (as seen pre-treatment components for successful treatment delivery, to
in both online and face-to-face groups), may not be face-to-face standard. Furthermore, the application
adequately represented on this assessment (Yorkston was user-friendly and easy to operate and allowed
and Beukelman 1981). Furthermore, if participants in the SLPs to focus their attention appropriately on
the present study were to demonstrate treatment gains the LSVTw goals of training loud phonation and
above the speaker variability on the sentence intellig- calibration. The successful delivery of all treatment
ibility task, the post-LSVTw values online and face-to- sessions and the high participant satisfaction with the
face would have exceeded the 100% upper limit of the online treatment overall was encouraging.
assessment range. In the absence of more representative It is acknowledged, however, that a number of
standardized assessments, the current treatment gains challenges unique to the online environment arose for
the SLPs and participants. Firstly, occasional audio
on the various measures of intelligibility were very
delays of up to 3 s were encountered during some of the
promising.
online sessions. These delays could potentially affect the
On the final measure of the ASSIDS, significant
communicative interactions. However, the participants
post-treatment improvements in communication effi- and SLPs were able to accommodate quickly for any
ciency were not achieved for the PD participants in either disturbances caused by the audio delays even within the
treatment environment. The post-LSVTw communi- first few treatment sessions. A useful strategy that aided
cation efficiency ratios were in fact lower than this process was waiting until the speaker had clearly
pre-treatment (mean ¼ 4.95% lower online; mean ¼ finished before responding. Where necessary, the SLPs
5.62% lower face-to-face). This finding can be explained also used shorter and more precise instructions during
in relation to the calculation of the communication tasks to facilitate the flow of the conversation and to
efficiency ratio. For the majority of the participants in minimize the likelihood of both individuals talking on
both treatment environments (64.71% of online top of each other. On occasion, the participants with
participants; 70.59% face-to-face), the reading time moderate dysarthria and a level of cognitive difficulty
was longer post-LSVTw (mean ¼ 0.20 min longer, would require greater prompting from the SLPs for
SD ¼ 0.16 online; mean ¼ 0.32 min longer, shaping the desired vocal responses, especially during
SD ¼ 0.26 face-to-face). This increase in time con- the first few treatment sessions. For these participants,
tributed to lower efficiency values in the calculations, the SLPs also used easy to detect hand-cues for
despite the significant gains obtained in sentence immediate feedback while the participant was perform-
intelligibility. A longer reading time is an anticipated ing the task, in order to further avoid the audio delays.
positive effect of the LSVTw that contributes to Examples of the hand-cues included ‘hand raising’ to
improved speech intelligibility. The louder phonation indicate the need to increase loudness, ‘thumbs-up’ to
with treatment has been found to decrease speech rate by denote appropriate performance and to continue with
the task, and ‘stop’ to terminate the task.
increasing pause length and reducing the duration of
Secondly, in comparison with the face-to-face
utterances during reading and conversational talks modality where viewing of the participant was optimal,
(Ramig et al. 1995a). Goberman and Elmer (2005) the video quality with videoconferencing at 128 kbit/s
also identified slower articulation rates in reading and was relatively poor. The frame rate and pixelated image
conversational tasks for participants with PD in their especially during movement, made it more difficult for
study following prompting to use clear speech. With the participants and SLPs to view each other clearly
these factors in mind, the communication efficiency during the sessions. However, the participants and SLPs
parameter alone may not clearly represent the were also able to compensate for the reduced video
post-LSVTw changes and results should be interpreted quality. Indirectly, the structured nature of the LSVTw
alongside those of word, sentence and conversational and reliance primarily on verbal communication helped
intelligibility. to promote the most optimal viewing possible with
videoconferencing. As the LSVTw does not require some unique challenges for treatment delivery, the
whole body movements, both the participant and SLP participants and speech –language pathologists (SLPs)
were able to sit relatively still in front of the PC during were able to adapt to this modality quickly, and a high
the sessions, which allowed for the least image level of participant satisfaction was achieved overall.
pixelation. Additionally, where clearer viewing of the Further studies are needed to investigate the feasibility
participant was needed, such as for monitoring their and validity of the online delivery of the LSVTw for
breath support, open mouth posture, positioning and participants at more advanced stages of PD and with
posture, the SLP was able to record the desired task moderate to severe hypokinetic dysarthria, to identify
using the store-and-forward function and review it any challenges to treatment that may exist for this
before providing feedback. In future, the population in this environment. Moreover, online
use of applications with higher Internet bandwidth treatment delivery in the ‘real-world’ remote setting at
may assist in further reducing any audio-visual either the participant’s home or within a community
difficulties with videoconferencing, and the need for health centre should also be the next stage of research.
compensatory strategies. In-depth analyses of participant and SLP satisfaction
Thirdly, compared with the face-to-face sessions, with the online modality and cost analyses are also
the online SLPs found it more challenging to give needed. Together, this research would outline the total
the impression of maintaining eye contact with the benefits of online delivery as an alternative or additional
participant by looking directly at the web camera on mode of service provision for people with PD.
top of the monitor rather than at the participant’s
face on the PC screen. In future, the repositioning of the
web camera closer to the participant image would create a Acknowledgements
more natural environment where the SLP could This research was funded by a National Health and Medical
maintain eye contact at the same time as monitoring Research Council Project Grant (Number 301029). This research
forms part of the doctoral thesis of Gabriella Constantinescu at the
participant performance like in face-to-face. Overall, the University of Queensland, Brisbane, Australia. The authors would
aforementioned challenges associated with the online like to acknowledge Parkinson’s Queensland Incorporated, all
delivery of the LSVTw were able to be addressed easily participants, Roy Anderson, Anne Hill, Monique Waite, Chiara
and did not impact significantly on treatment outcomes. Wall, Jasmin Cowles, and Christina Iezzi for their contribution to
Such challenges were therefore not seen as major the research. Declaration of interest: The authors report no
conflicts of interest. The authors alone are responsible for the
limitations of online treatment delivery. content and writing of the paper.
Finally, although the online application incorpor-
ated all the essential components for successful
treatment delivery, it did not enable audio feedback References
to participants during the sessions, to assist calibration. BAUMGARTNER, C. A., SAPIR, S. and RAMIG, L. O., 2001, Voice
In the face-to-face setting, this was possible via tape quality changes following phonatory – respiratory effort
recordings. By incorporating this additional feature into treatment (LSVT) versus respiratory effort treatment for
future applications, the online delivery of the LSVTw individuals with Parkinson disease. Journal of Voice, 15,
would more closely resemble the face-to-face treatment 105– 114.
DALLAL, G. E., 1997, Randomization.com (available at: http://www.
format. randomization.com) (accessed on 1 March 2005).
DARLEY, F. L., ARONSON, A. E. and BROWN, J. R., 1969, Clusters of
deviant speech dimensions in the dysarthrias. Journal of
Conclusion Speech and Hearing Research, 12, 462– 496.
This study has demonstrated the validity and reliability DARLEY, F. L., ARONSON, A. E. and BROWN, J. R., 1975,
Motor Speech Disorders (Philadelphia, PA: W. B. Saunders).
of the online laboratory Lee Silverman Voice Treat- DE PIJPER, J. R., 2007, Semitone Conversions (available at: http://
ment (LSVTw) delivery via PC-based videoconferen- users.utu.fi/jyrtuoma/speech/semitone.html) (accessed on
cing for participants with Parkinson’s disease (PD) and 1 March 2007).
mild-to-moderate hypokinetic dysarthria. The results DE RIJK, M. C., TZOURIO, C., BRETELER, M. M., DARTIGUES, J. F.,
verified the non-inferiority of the online modality and AMADUCCI, L., LOPEZ-POUSA, S., MANUBENS-BERTRAN, J.
M., ALPEROVITCH, A. and ROCCA, W. A., 1997, Prevalence
supported the hypothesis that participants treated in the of parkinsonism and Parkinson’s disease in Europe: The
online environment would achieve comparable treat- EUROPARKINSON collaborative study. European com-
ment outcomes on the acoustic and perceptual munity concerted action on the epidemiology of Parkinson’s
measures to those treated in the face-to-face setting. disease. Journal of Neurology Neurosurgery and Psychiatry, 62,
On the majority of the parameters, the treatment 10 –15.
DROMEY, C. and RAMIG, L. O., 1998, Intentional changes in
outcomes for the participants in both the LSVTw sound pressure level and rate: their impact on measures
environments were also statistically and clinically of respiration, phonation, and articulation. Journal of Speech
significant. Although the online modality presented Language and Hearing Research, 41, 1003– 1018.
DROMEY, C., RAMIG, L. O. and JOHNSON, A. B., 1995, Phonatory RAMIG, L. O., FOX, C. and SAPIR, S., 2004, Parkinson’s disease:
and articulatory changes associated with increased vocal speech and voice disorders and their treatment with the Lee
intensity in Parkinson disease: a case study. Journal of Speech Silverman Voice Treatment. Seminars in Speech and
and Hearing Research, 38, 751– 764. Language, 25, 169– 180.
EL SHARKAWI, A., RAMIG, L. O., LOGEMANN, J. A., PAULOSKI, B. R., RAMIG, L. O., PAWLAS, A. A. and COUNTRYMAN, S., 1995b, The Lee
RADEMAKER, A. W., SMITH, C. H., PAWLAS, A., BAUM, S. Silverman Voice Treatment (LSVTw): A Practical Guide to
and WERNER, C., 2002, Swallowing and voice effects of Lee Treating the Voice and Speech Disorders in Parkinson Disease
Silverman Voice Treatment (LSVT): a pilot study. Journal of (Iowa City, IA: National Centre for Voice and Speech).
Neurology Neurosurgery and Psychiatry, 72, 31 – 36. RAMIG, L. O., SAPIR, S., COUNTRYMAN, S., PAWLAS, A. A.,
ENGEN, T., 1971, Psychophysics II: scaling methods. In J. W. Kling O’BRIEN, C., HOEHN, M. and THOMPSON, L. L., 2001a,
and L. Riggs (eds), Woodworth and Schlossberg’s Experimental Intensive voice treatment (LSVT) for patients with
Psychology (New York, NY: Holt, Rinehart, & Winston), Parkinson’s disease: a 2 year follow up. Journal of Neurology
pp. 47 – 86. Neurosurgery and Psychiatry, 71, 493– 498.
FAIRBANKS, G., 1960, Voice and Articulation Drillbook (New York, RAMIG, L. O., SAPIR, S., FOX, C. and COUNTRYMAN, S., 2001b,
NY: Harper & Brothers). Changes in vocal loudness following intensive voice
FLEISS, J. L., 1981, Statistical Methods for Rates and Proportions, 2nd treatment (LSVT) in individuals with Parkinson’s disease:
edn (New York, NY: Wiley). a comparison with untreated patients and normal age-
FOX, C. M. and RAMIG, L. O., 1997, Vocal sound pressure level and matched controls. Movement Disorders, 16, 79 – 83.
self-perception of speech and voice in men and women with SAPIR, S., RAMIG, L. O., HOYT, P., COUNTRYMAN, S., O’BRIEN, C.
idiopathic Parkinson disease. American Journal of Speech – and HOEHN, M., 2002, Speech loudness and quality
Language Pathology, 6, 85 – 94. 12 months after intensive voice treatment (LSVT) for
GOBERMAN, A. M. and ELMER, L. W., 2005, Acoustic analysis of Parkinson’s disease: a comparison with an alternative speech
clear versus conversational speech in individuals with treatment. Folia Phoniatrica et Logopaedica, 54, 296– 303.
Parkinson disease. Journal of Communication Disorders, 38, SAPIR, S., SPIELMAN, J. L., RAMIG, L. O., STORY, B. H. and FOX, C.,
215– 230. 2007, Effects of intensive voice treatment (the Lee Silverman
GRMM, N., PAUL, J. and WAKEHAM, A., 2004, Access to Voice Treatment [LSVT]) on vowel articulation in dysarthric
individuals with idiopathic Parkinson disease: acoustic and
speech pathology services by people with Parkinson’s disease
perceptual findings. Journal of Speech, Language, and Hearing
in Queensland, Australia. Unpublished Master’s thesis,
Research, 50, 899– 912.
University of Queensland, Brisbane.
SCHIAVETTI, N., METZ, D. E. and SITLER, R. W., 1981, Construct
HOEHN, M. and YAHR, M., 1967, Parkinsonism: onset, progression
validity of direct magnitude estimation and interval scaling
and mortality. Neurology, 17, 427– 442.
of speech intelligibility: evidence from a study of the hearing
HOWELL, S., TRIPOLITI, E. and PRING, T., 2009, Delivering the Lee
impaired. Journal of Speech and Hearing Research, 24,
Silverman Voice Treatment (LSVT) by web camera: a 441– 445.
feasibility study. International Journal of Language and STEVENS, S. S., 1975, Psychophysics: Introduction to its Perceptual,
Communication Disorders, 44, 287– 300. Neural, and Social Prospects (New York, NY: Wiley).
JONES, B., JARVIS, P. and EBBUTT, A. F., 1996, Trials to THEODOROS, D., CONSTANTINESCU, G., RUSSELL, T. G., WARD,
assess equivalence: the importance of rigorous methods. E. C., WILSON, S. J. and WOOTTON, R., 2006, Treating the
British Medical Journal, 3, 36 – 39. speech disorder in Parkinson’s disease online. Journal of
OXTOBY, M., 1982, Parkinson’s Disease Patients and Their Social Telemedicine and Telecare, 12 (Suppl. 3), S3:88 – 91.
Needs (London: Parkinson’s Disease Society). TINDALL, L. R., HUEBNER, R. A., STEMPLE, J. C. and KLEINERT, H.
PIAGGIO, G., ELBOURNE, D. R., ALTMAN, D. G., POCOCK, S. J. L., 2008, Videophone-delivered voice therapy: a comparative
and EVANS, S. J. W., 2006, Reporting of noninferiority analysis of outcomes to traditional delivery for adults with
and equivalence randomised trials: an extension of the Parkinson’s disease. Telemedicine Journal and e-H Health, 14,
CONSORT statement. Journal of the American Medical 1070– 1077.
Association, 295, 1152– 1160. WARD, E. C., THEODOROS, D. G., MURDOCH, B. E. and SILBURN,
RAMIG, L. O., BONITATI, C. M., LEMKE, J. H. and HORII, Y., 1994, P., 2000, Changes in maximum capacity tongue function
Voice treatment for patients with Parkinson disease: following the Lee Silverman Voice Treatment Program.
development of an approach and preliminary efficacy data. Journal of Medical Speech – Language Pathology, 8, 331– 335.
Journal of Medical Speech –Language Pathology, 2, 191–209. WEISMER, G. and LAURES, J. S., 2002, Direct magnitude estimates
RAMIG, L. O., COUNTRYMAN, S., O’BRIEN, C., HOEHN, M. and of speech intelligibility in dysarthria: effects of a chosen
THOMPSON, L., 1996, Intensive speech treatment for standard. Journal of Speech Language and Hearing Research,
patients with Parkinson’s disease: short- and long-term 45, 421– 433.
comparison of two techniques. Neurology, 47, 1496– 1504. WHITEHILL, T. L., LEE, A. S. and CHUN, J. C., 2002, Direct
RAMIG, L. O., COUNTRYMAN, S., THOMPSON, L. L. and HORII, Y., magnitude estimation and interval scaling of hypernasality.
1995a, Comparison of two forms of intensive speech Journal of Speech Language and Hearing Research, 45, 80 – 88.
treatment for Parkinson disease. Journal of Speech and YORKSTON, K. M. and BEUKELMAN, D. R., 1981, Assessment of
Hearing Research, 38, 1232– 1251. Intelligibility of Dysarthric Speech (Austin, TX: PRO-ED).
Appendix: Online Participant Satisfaction Questionnaire
This questionnaire has been developed to determine you satisfaction with the treatment that you had across the Internet. Please CIRCLE the
answer that you feel is most appropriate based on your experience.
(1) How did you feel while participating in the treatment sessions on the Internet?
(a) Would prefer these types of sessions to face-to-face sessions
(b) Very happy with this session
(c) Comfortable
(d) Uneasy
(e) Would not participate again
(2) What is your opinion of the audio quality (what you were able to hear) during the majority of the sessions?
(a) Excellent
(b) More than adequate
(c) Adequate
(d) Inadequate
(e) Poor
(3) What is your opinion of the visual quality (what you were able to see) during the majority of the sessions?
(a) Excellent
(b) More than adequate
(c) Adequate
(d) Inadequate
(e) Poor
(4) Please rate your overall satisfaction with the treatment across the Internet.
(a) Very satisfied
(b) More than satisfied
(c) Satisfied
(d) Less than satisfied
(e) Not at all satisfied
RESEARCH Case report
.................................................................................................................................
Q Home-based speech treatment for Parkinson’s

disease delivered remotely: a case report
Gabriella A Constantinescu*, Deborah G Theodoros*, Trevor G Russell*,
Elizabeth C Ward*, Stephen J Wilson† and Richard Wootton‡§
*School of Health and Rehabilitation Sciences, University of Queensland, Brisbane; †School of Information Technology and Electrical
Engineering, University of Queensland, Brisbane; ‡University of Queensland, Brisbane, Australia; §Scottish Centre for Telehealth,
Aberdeen, UK
Summary
We investigated the validity and feasibility of online delivery of the Lee Silverman Voice Treatment (LSVT) for the treatment
of the speech disorder of a patient with idiopathic Parkinson’s disease. The treatment was delivered in 16 sessions to
the participant’s home, 90 km from the speech language pathologist. A PC-based videoconferencing system was used,
operating at 128 kbit/s over the public telecommunications network. The patient achieved substantial improvements
in vocal sound pressure levels during sustained vowel phonation (6.13 dB), reading (12.28 dB) and conversational
monologue (11.32 dB). There were improvements in the duration of sustained vowel phonation (4 s). Improvements were
also perceived in the degree of breathiness and roughness in the voice, and in overall speech intelligibility in conversation.
The patient was very satisfied with the audio and video quality of the conferencing, and with the online treatment overall.
He reported a preference for online sessions for the future management of his condition, rather than face-to-face
treatment. Remote LSVT delivery was found to be feasible and effective.
Introduction Treatment (LSVT). This is a structured programme which is

.............................................................. delivered intensively in 16 treatment sessions and promotes
increased respiratory drive, vocal fold adduction and
Home telecare is increasingly being used to help elderly and increased vocal loudness in everyday communication.10,11
infirm people remain independent within their homes by Improvements following traditional face-to-face delivery of
improving their access to health care.1,2 Home telecare is the LSVT have been reported in vocal loudness, quality and
ideal for people with Parkinson’s disease (PD) as it can speech intelligibility, with these positive effects maintained
reduce barriers to service access such as the travel difficulties up to 1212 and 24 months post-treatment.13
associated with the large distances to specialised health-care Despite the effectiveness of the LSVT in the management
facilities that are frequently encountered in Australia and in of dysarthria in PD, patient access to face-to-face speech
other countries.3 Idiopathic Parkinson’s disease (IPD) is a pathology services for this treatment remains limited due
chronic degenerative disease. The motor speech disorder to restricted availability of trained clinicians, conflicting
associated with IPD is known as hypokinetic dysarthria and caseload priorities and patient access barriers.3 The aim of
is characterised by reduced loudness, monotony of pitch the present study was to investigate the validity and feasibility
and loudness, inappropriate silences, variable rate and short of remote online LSVT delivery for a single case of IPD
rushes of speech, imprecise articulation, and a harsh and using PC-based videoconferencing over the Internet.
breathy voice.4,5 Hypokinetic dysarthria can affect a high Our hypothesis was that LSVT can be effectively delivered
proportion of people with PD and progresses in severity over remotely, with treatment outcomes similar to those for
time.6 – 8 As a result of the reduced speech intelligibility, traditional face-to-face LSVT.
people with PD often experience communication
difficulties which may lead to isolation and may reduce
their quality of life.8,9
At present, the most effective, evidence-based treatment Methods
..............................................................
for PD and hypokinetic dysarthria is the Lee Silverman Voice
Mr B, a 65-year-old retired man living in a regional city,
Accepted 10 June 2009 90 km north of Brisbane, was recruited to the study. Ethics
Correspondence: Gabriella Constantinescu, Division of Speech Pathology,
University of Queensland, St Lucia 4072, Australia (Fax: þ61 7 3870 3998; approval was obtained from the appropriate committee.
Email: gabriella@hearandsaycentre.com.au) Mr B had been diagnosed with IPD six years prior to the
Journal of Telemedicine and Telecare 2010; 16: 100– 104 DOI: 10.1258/jtt.2009.090306
G A Constantinescu et al. Home-based speech treatment for Parkinson’s disease
study and was classified as Stage I on the Hoehn and Yahr took approximately 10 minutes to complete. The
Scale.14 In the previous few years, the participant and his participant was shown how to perform the basic functions
wife had noted a reduction in his vocal loudness and speech required for treatment, including turning on/off the PC,
intelligibility, as well as a breathy speech quality. As a result, activating/closing the application and positioning the
Mr B was finding it more difficult to be understood in group headset microphone. Once the set-up was completed, the
situations, where there was background noise and microphone distance was adjusted to 5 cm from the corner
occasionally while speaking by telephone. His speech and of the participant’s mouth. This helped to reduce sound
voice difficulties were beginning to affect his duties as a distortion, maximize visibility of the participant’s face,
coordinator of a community group. In daily life, Mr B began and allow for accurate recordings of pitch and SPL. To verify
to avoid speaking by telephone and would often spend the the positioning of the microphone, the SPL (dB-C) data
greater part of his day at home, where he did not have to generated by the system’s acoustic speech processor was
talk to anyone. Mr B had not previously received any speech then confirmed against a conventional Digital Sound Level
pathology services. There were also no public or private Meter (Model No. 23 –553, Radio Shack) during three
speech pathology services offering the LSVT in his local sustained phonations of the vowel /a/. The participant was
community at the time of the study. The intensive nature instructed to avoid adjusting the microphone arm over
and commitment of the LSVT programme, and participant the course of treatment and to take care where he placed
fatigue when driving, made it difficult for Mr B to seek the headset microphone when not in use. During the
treatment for his speech and voice difficulties outside his treatment, the SLP communicated with the participant via a
local area. headset microphone during videoconferencing. The SLP
Prior to treatment, the participant’s speech was classified controlled all displays on the participant’s screen, without
by the principal investigator as demonstrating mild the need for the participant to operate the system.
hypokinetic dysarthria. This classification was based on
Mr B’s pre-LSVT assessment results on acoustic measures of
vocal volume for sustained vowel phonation (81.96 dB), LSVT programme
a monologue (68.13 dB), and a perceptual rating of mildly The LSVT sessions were delivered according to standard
reduced overall speech intelligibility in conversation (using clinical practice, for one hour per day, four days a week,
a 5-point rating scale). A videolaryngoscopic examination, for four weeks.10,11 The daily sessions included maximum
conducted by an ear, nose and throat specialist, revealed duration of sustained vowel phonation, fundamental
some vocal fold bowing, a feature consistent with IPD. Mr B frequency range, functional speech loudness drills and
wore bilateral hearing aids for his mild-moderate bilateral hierarchical speech loudness tasks. The tasks were
hearing loss. He remained on a constant drug regime for IPD performed at high intensity and with maximum effort to
throughout the study. promote increased respiratory drive, vocal fold adduction
and carryover of the louder voice into functional
communication. Homework reading materials were emailed
Procedure to the participant at the end of each session.
The online treatment was delivered by a LSVT-certified
speech language pathologist (SLP). The SLP was located in
Brisbane, and delivered the treatment to the participant’s Outcome measures
home via a PC-based telerehabilitation system. One PC was Pre- and post-LSVT assessments were conducted in the
located at the SLP site, while the other was a laptop traditional face-to-face manner in Brisbane by two SLPs who
computer located at the participant’s home. had not taken part in the treatment. A battery of acoustic
The telerehabilitation system was able to: (1) provide and perceptual measures was used in these assessments.
videoconferencing; (2) present reading material for the The acoustic measures were obtained using the LSVT
participant; (3) manipulate the web cameras at the participant Evaluation Protocol10 and included: (1) average SPLs of the
site via a robot arm to maintain a clear view of the participant participant’s speech during six phonations of the vowel /a/,
throughout the session; (4) obtain average measures of reading of the standard ‘grandfather passage’,15 and during
sound-pressure level (SPL), duration and peak frequency via a 30 s monologue about a topic of interest; (2) the average
the system’s acoustic speech processor; and (5) capture high duration of six sustained vowel phonations; and (3) the
quality video (640 480 pixel resolution) and audio, pitch range obtained from the average of the highest and
compressed at 384 kbit/s for later examination. The acoustic lowest frequencies from a series of six vocal glides, that was
speech processor was connected to the laptop computer via a then converted to a maximum range in semitones.16
USB port. A 128 kbit/s Internet connection was established The acoustic measures were obtained using the online
between the two videoconferencing systems using the public system’s acoustic speech processor.
telecommunications network (ADSL). The connection Perceptual measures of voice and speech included ratings
enabled videoconferencing at 320 240 pixel resolution of voice variables (breathiness, roughness, loudness level,
between the two systems. loudness and pitch variability), overall articulatory
Prior to treatment, the participant’s system was set up in a precision and overall speech intelligibility in conversation.
quiet room of his home by the principal investigator. This The variables of breathiness, roughness and overall
Journal of Telemedicine and Telecare Volume 16 Number 2 2010 101

articulatory precision were determined from pre- and Table 1 Pre- and post-LSVT values and treatment changes
post-LSVT readings of the standard ‘rainbow passage’.17
Task Pre-LSVT Post-LSVT Change
A 30 s monologue was used to determine the loudness level,
pitch, loudness variability and overall intelligibility in Sustained vowel phonation (dB) 81.96 88.09 6.13
Reading (dB) 71.42 83.70 12.28
conversation.
Monologue (dB) 68.13 79.45 11.32
The ratings on all of the perceptual variables were made Duration of phonation (s) 9.67 13.67 4.00
using Direct Magnitude Estimation (DME), a scaling Pitch range (semitones) 12.87 9.01 23.86
method that allows the rater to assign any numerical value
Improvement
to the variable and rate the sample against a ‘standard’ or a
representation of a midrange impairment of that variable
for improved reliability.18 – 20 A set value of 100 was used to Table 2 Pre- and post-values and treatment changes (values shown are
represent the standard.18,21 For the variables of loudness DME ratings)
level and variability, pitch variability, overall speech Task Pre-LSVT Post-LSVT Change
intelligibility in conversation and overall articulatory
Breathiness 75.00 44.67 230.33
precision, a rating of 50 suggested that the sample was only
Roughness 79.43 64.57 214.86
half as clear as the standard, while a rating of 200 indicated Loudness level 186.21 151.36 234.85
that the sample was twice as clear as the standard.20 For the Loudness variability 120.23 114.82 25.41
Pitch variability 104.71 95.50 29.21
variables of roughness and breathiness, however, a sample
Articulatory precision 162.18 162.18 0
that was rated lower than 100 represented an improvement Overall speech intelligibility in conversation 128.82 141.25 12.43
in quality. DME ratings were made by two experienced SLPs
Improvement
who were blinded to the study intent. For each variable, the
raters listened to the standard followed by the pre- and
post-treatment speech samples in random order.21 A performance with treatment was evident on the remaining
different standard was used for each variable.20 An average variables of loudness level (a 34.85 DME reduction),
pre- and post-treatment value was calculated for each loudness variability (5.41 DME reduction) and pitch
variable, which was then converted to a logarithmic value variability (9.21 DME reduction). The pre- and post-LSVT
and represented as a geometric mean.19,22 values and treatment changes are summarised in Table 2.
The participant also completed a satisfaction On the participant satisfaction questionnaire, Mr B rated
questionnaire (5-point scale) relating to the online the audio and video quality as excellent. He was very
treatment modality. The questionnaire determined his satisfied with the online treatment overall and indicated
satisfaction with: (1) the online treatment sessions that he would prefer online sessions to face-to-face for
(responses ranging from would not participate again to future management of his condition. Overall, there were no
would prefer these types of sessions to face-to-face sessions failed treatment sessions and the majority (n ¼ 13) of the
for future management of PD); (2) the audio and video sessions ran very smoothly, without technical difficulties
quality during the sessions ( poor to excellent); and (3) and with adequate audio and video quality for treatment
overall satisfaction with online treatment (not at all satisfied delivery. During the remaining three sessions, some
to very satisfied). networking difficulties considerably reduced the audio and
video quality. These problems were solved by disconnecting
and re-establishing the videoconferencing connection
between the two telerehabilitation systems.
Results
..............................................................
Descriptive comparisons and degree of change were
Discussion
determined between the pre- and post-LSVT acoustic and ..............................................................
perceptual measures. For the acoustic variables, substantial
treatment gains were made on all SPL tasks (6.13 dB The present case report demonstrates the feasibility of
improvement on sustained vowel phonation; 12.28 dB on remote online delivery of the LSVT via a PC-based
reading and 11.32 dB on monologue loudness), and on the videoconferencing system operating on a 128 kbit/s
duration of sustained vowel phonation variable (4 s Internet connection and supports the study hypothesis.
improvement). Pitch range failed to show an improvement On the whole, Mr B showed substantial improvements with
with treatment. The results are summarised in Table 1. remote LSVT for most of the acoustic and perceptual
For the perceptual variables, treatment gains were variables, and these were similar to the treatment outcomes
observed for measures of breathiness (a 30.33 DME value reported in the literature for IPD participants with mild to
reduction), roughness (14.86 DME reduction) and overall moderate hypokinetic dysarthria following face-to-face
speech intelligibility in conversation (12.43 DME LSVT.11,12,23 – 26 The post-treatment values for the SPL
improvement). No treatment changes were evident for variables were also consistent with the average values
overall articulatory precision, with this variable remaining reported for two groups of healthy older adults speaking at a
at the high pre-treatment DME value of 162.18. Lower comfortable loudness level, in studies with similar
102 Journal of Telemedicine and Telecare Volume 16 Number 2 2010

assessment procedures.27,28 It is possible that the lack of of remote home-based LSVT via the public network.
perceived improvement in some speech and voice variables There were no failed treatment sessions and the majority
( pitch range, loudness level and variability, and overall of sessions ran very smoothly, with adequate audio and
articulatory precision), was related to the participant’s mild video quality for treatment delivery. On only three
degree of hypokinetic dysarthria which made it difficult for occasions was the SLP required to disconnect and
the raters to perceive changes in these variables following re-establish the videoconference connection due to
treatment. It is recognised that the perception of speech and networking difficulties.
vocal changes is more difficult in milder degrees of speech Although occasional audio delays were also encountered
impairment29 and may not be as accurate as objective during treatment, they were effectively managed by the
measurement. Despite the lack of substantial improvements participant and clinician who waited until the other had
in some specific speech and voice variables, the participant clearly finished speaking before replying, and by the
demonstrated improved speech intelligibility in clinician using shorter and more precise instructions.
conversation and reported improved performance in Strategies, however, were needed to maximise the video
everyday speech activities following treatment. quality, as the frame rate and pixelated video image
especially during movement made it more difficult for the
clinician and participant to clearly view each other during
Participant’s perspectives the session. Useful techniques to improve the video quality
Overall, Mr B benefited from remote LSVT and found the included: sitting relatively still in front of the PC during the
treatment useful for increasing his loudness to pre-morbid sessions; the SLP using easy to detect hand-cues for quick
levels, and for integrating and maintaining the treatment input; relying more heavily on specific verbal directions and
gains of improved loudness level, vocal quality and speech participant feedback rather than on visual information; and
intelligibility in daily life. As a result of these positive using the store-and-forward modality to record the desired
treatment changes, Mr B felt that his speech sounded task when necessary. These strategies were easily adopted in
natural and close to how he remembered it prior to PD. the sessions and aided the smooth delivery of treatment.
He regained his confidence in talking to his family and The use of higher Internet bandwidth would improve the
friends, gave a speech at his daughter’s wedding, and began interface and thus more closely resemble the face-to-face
speaking again by telephone and in group meetings. modality, reduce the audio and video difficulties and lessen
Overall, Mr B was very satisfied with the online treatment the need for compensatory strategies.
and rated the video and audio quality of the online Other procedures necessary to ensure the smooth delivery
application as excellent. His level of hearing loss, which was of treatment in the home environment were identified
corrected by hearing aids, did not interfere with his ability during the study. The sessions were conducted in a quiet
to hear instructions over the videoconferencing link. room of the participant’s home, which reduced any
Interestingly, on the satisfaction survey, Mr B reported that household distractions and noise, and telephone calls
he would prefer online sessions to face-to-face for the future were not taken during the sessions. Furthermore, the
management of his condition. He felt that the online microphone distance from the participant’s mouth was kept
method provided: ease of access to treatment without the constant, to ensure that the acoustic levels were constant
nuisance of travelling and the need to leave his own home; between sessions. The participant also took great care to
time-savings from not having to travel; and a friendly ensure that this distance was maintained. It would be useful
technical interface for treatment using reliable technology. in future for the SLP to be able to provide audio-recordings
The convenience of remote online LSVT proved to be a of the participant’s performance as a form of feedback for
highly motivating factor for Mr B. Similar findings have also them. This additional capability would make online LSVT
been reported in other telehealth studies where patients even more similar to face-to-face treatment.
were motivated and accepting of the technology used when In conclusion, the treatment gains, the high participant
they could be treated in their natural or least restrictive satisfaction and motivation with remote LSVT illustrate the
environment.30,31 potential of this type of service delivery for people with PD.
Remote treatment may assist in reducing the effects of
physical disability, transport and travel difficulties, and
Clinician’s perspectives distance for those with PD, which at present represent
The real-time videoconferencing feature of the application substantial barriers to service access. Remote treatment may
and the ability to capture and display SPL and frequency also facilitate earlier access to intervention, thus allowing
data and display therapy materials online were important to individuals to remain independent and active within their
the successful delivery of the treatment, allowing the own homes and communities. Because the present report
clinician to: provide timely instructions to the participant concerns a single case, the findings cannot be generalised to
and assist with shaping correct voice productions and the wider PD population. Future large-scale studies are
overall calibration; monitor the loudness level, pitch level needed to investigate the effectiveness of remote online
and vocal quality; and maintain good rapport with the LSVT with a larger number of participants with PD and
participant. From the clinician’s perspective, the online dysarthria severity levels and greater numbers of treating
application was user-friendly and allowed effective delivery clinicians.
Journal of Telemedicine and Telecare Volume 16 Number 2 2010 103

Acknowledgements: This research was funded by a National 14 Hoehn MM, Yahr MD. Parkinsonism: onset, progression and mortality.
Neurology 1967;17:427 –42
Health and Medical Research Council Project Grant 301029.
15 Darley FL, Aronson AE, Brown JR. Motor Speech Disorders. Philadelphia:
The work forms part of the doctoral thesis of Gabriella Saunders, 1975
Constantinescu at the University of Queensland. We thank 16 de Pijper JR. Semitone conversions. See http://users.utu.fi/jyrtuoma/
Parkinson’s Queensland Incorporated, the participant, speech/semitone.html (last checked 18 June 2009)
17 Fairbanks G. Voice and Articulation Drillbook. New York: Harper, 1960
Roy Anderson, Anne Hill, Monique Waite, Jasmin Cowles 18 Schiavetti N, Metz DE, Sitler RW. Construct validity of direct magnitude
and Christina Iezzi for their contribution to the research. estimation and interval scaling of speech intelligibility: evidence from a
LSVT is a registered trademark. study of the hearing impaired. J Speech Hear Res 1981;24:441 –5
19 Stevens SS. Psychophysics: Introduction to its Perceptual, Neural, and Social
Prospects. New York: John Wiley, 1975
20 Weismer G, Laures JS. Direct magnitude estimates of speech intelligibility
References in dysarthria: effects of a chosen standard. J Speech Lang Hear Res
1 Barlow J, Singh D, Bayer S, Curry R. A systematic review of the benefits of 2002;45:421 –33
home telecare for frail elderly people and those with long-term 21 Whitehill TL, Lee AS, Chun JC. Direct magnitude estimation and interval
conditions. J Telemed Telecare 2007;13:172 –9 scaling of hypernasality. J Speech Lang Hear Res 2002;45:80 –8
2 Onor ML, Trevisiol M, Urciuoli O, et al. Effectiveness of telecare in elderly 22 Engen T. Psychophysics II: Scaling methods. In: Kling JW, Riggs L, eds.
populations – a comparison of three settings. Telemed J E Health Woodworth and Schlossberg’s Experimental Psychology. New York: Holt,
2008;14:164 –9 Rinehart & Winston, 1971:47 –86
3 Grimm N, Paul J, Wakeham A. Access to Speech Pathology Services by People 23 Ramig LO, Countryman S, Thompson LL, Horii Y. Comparison of two
with Parkinson’s Disease in Queensland, Australia. (Unpublished report) forms of intensive speech treatment for Parkinson disease. J Speech Hear
Brisbane: University of Queensland, 2004 Res 1995;38:1232 – 51
4 Darley FL, Aronson AE, Brown JR. Differential diagnostic patterns of 24 El Sharkawi A, Ramig LO, Logemann JA, et al. Swallowing and voice
dysarthria. J Speech Hear Res 1969;12:246 – 69 effects of Lee Silverman Voice Treatment (LSVT): a pilot study. J Neurol
5 Darley FL, Aronson AE, Brown JR. Clusters of deviant speech dimensions Neurosurg Psychiatry 2002;72:31 –6
in the dysarthrias. J Speech Hear Res 1969;12:462 –96 25 Baumgartner CA, Sapir S, Ramig TO. Voice quality changes following
6 Ramig LO, Fox C, Sapir S. Parkinson’s disease: speech and voice disorders phonatory-respiratory effort treatment (LSVT) versus respiratory
and their treatment with the Lee Silverman Voice Treatment. Semin Speech effort treatment for individuals with Parkinson disease. J Voice 2001;
Lang 2004;25:169 –80 15:105 – 14
7 Hartelius L, Svensson P. Speech and swallowing symptoms associated with 26 Sapir S, Ramig LO, Hoyt P, Countryman S, O’Brien C, Hoehn M. Speech
Parkinson’s disease and multiple sclerosis: a survey. Folia Phoniatr Logop loudness and quality 12 months after intensive voice treatment (LSVT)
1994;46:9 –17
for Parkinson’s disease: a comparison with an alternative speech
8 Stewart C. Speech-language therapy for patients with Parkinson’s disease.
treatment. Folia Phoniatr Logop 2002;54:296 –303
In: Cote L, Sprinzeles LL, Elliott R, Kutscher AH, eds. Parkinson’s Disease
27 Ramig LO, Sapir S, Fox C, Countryman S. Changes in vocal loudness
and Quality of Life. New York: Haworth Press, 2000:151 –6
following intensive voice treatment (LSVT) in individuals with
9 Oxtoby M. Parkinson’s Disease Patients and Their Social Needs. London:
Parkinson’s disease: a comparison with untreated patients and normal
Parkinson’s Disease Society, 1982
age-matched controls. Mov Disord 2001;16:79 –83
10 Ramig LO, Pawlas AA, Countryman S. The Lee Silverman Voice Treatment
(LSVTw): A Practical Guide to Treating the Voice and Speech Disorders in 28 Fox CM, Ramig LO. Vocal sound pressure level and self-perception of
Parkinson Disease. Iowa City: National Centre for Voice and Speech, 1995 speech and voice in men and women with idiopathic Parkinson disease.
11 Ramig LO, Bonitati CM, Lemke JH, Horii Y. Voice treatment for patients Am J Speech Lang Pathol 1997;6:85 –94
with Parkinson disease: development of an approach and preliminary 29 Kreiman J, Gerratt BR. Validity of rating scale measures of voice quality.
efficacy data. J Medical Speech-Language Pathology 1994;2:191 –209 J Acoust Soc Am 1998;104:598 –608
12 Ramig LO, Countryman S, O’Brien C, Hoehn M, Thompson L. Intensive 30 Mashima PA, Birkmire-Peters DP, Syms MJ, Holtel MR, Burgess LP,
speech treatment for patients with Parkinson’s disease: short- and Peters LJ. Telehealth: voice therapy using telecommunications technology.
long-term comparison of two techniques. Neurology 1996;47:1496 –504 Am J Speech Lang Pathol 2003;12:432 –9
13 Ramig LO, Sapir S, Countryman S, et al. Intensive voice treatment (LSVT) 31 Hornsby D, Hudson L. Videoconferencing for service delivery and
for patients with Parkinson’s disease: a 2 year follow up. J Neurol Neurosurg professional education. Australian Communication Quarterly 1997;
Psychiatry 2001;71:493 –8 Spring:25 –26
104 Journal of Telemedicine and Telecare Volume 16 Number 2 2010

The Assessment and Treatment of Disordered Speech and Voice in Parkinson S Disease Using A PC-based Telerehabilitation System

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Assessment and Treatment of Disordered Speech and Voice in Parkinson S Disease Using A PC-based Telerehabilitation System

Uploaded by

Copyright:

Available Formats

The Assessment and Treatment of

Disordered Speech and Voice in

Gabriella Amalia Constantinescu

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in July 2010

School of Health and Rehabilitation Sciences

This thesis is composed of my original work, and contains no material previously

Gabriella Amalia Constantinescu

This thesis is dedicated to Peggy and Dawn.

I must acknowledge my colleagues in the Telerehabilitation Research Unit, Dr Anne Hill,

Subsequent to the validation of the online assessment of hypokinetic dysarthria, three

Parkinson‟s disease, telerehabilitation, assessment, treatment, Lee Silverman Voice Treatment,

Australian and New Zealand Standard

1.1 INTRODUCTION ......................................................................................................................... 1

CHAPTER 2 TELEREHABILITATION SYSTEM DESIGN AND CALIBRATION ...... 45

CHAPTER 3 ASSESSING DISORDERED SPEECH AND VOICE IN PARKINSON’S

3.1 INTRODUCTION ....................................................................................................................... 71

CHAPTER 4 TREATING DISORDERED SPEECH AND VOICE IN PARKINSON’S

4.1 INTRODUCTION ..................................................................................................................... 115

5.1 INTRODUCTION ..................................................................................................................... 155

CHAPTER 6 HOME-BASED SPEECH TREATMENT FOR PARKINSON’S DISEASE

6.1 INTRODUCTION ..................................................................................................................... 185

CHAPTER 7 CONCLUSIONS .............................................................................................. 203

REFERENCES .......................................................................................................................... 209

APPENDIX – PUBLISHED PAPERS...................................................................................... 233

Table 5.1 Descriptive Characteristics of Participants.................................................................... 160

Figure 1.2 Representation of the basal ganglia control circuit. ......................................................... 6

Figure 1.4 A schematic representation of the associated telehealth, telemedicine, telerehabilitation

Figure 2.1 Telerehabilitation system at participant site. ................................................................... 48

Figure 2.2 Telerehabilitation system at SLP site. . ........................................................................... 48

Figure 2.3 Web cameras on robotic arm at participant site. . ........................................................... 49

Figure 2.5 Telerehabilitation system‟s acoustic speech processor. ................................................. 54

Figure 3.6 Bland and Altman LA for maximum F0 range. ............................................................... 96

Figure 3.7 Bland and Altman LA for word intelligibility. ............................................................... 99

Figure 4.2 Online LSVT® environment at participant site. ............................................................ 126

Figure 4.4 Face-to-face LSVT® environment. ................................................................................ 127

Figure 6.1 SLP treatment site. . ....................................................................................................... 189

Figure 6.2 Participant treatment site. ............................................................................................. 189

ANZTC Australian and New Zealand Telehealth Committee

However, despite the advancements in general healthcare services, concomitant

1.2 Parkinson’s Disease

PD refers to a group of progressive neurodegenerative disorders that are associated with

1.2.1 Economic Burden of Parkinson’s Disease

1.3.1 Neuropathology and Clinical Presentation of Motor Symptoms

1.3.2 Non-motor Symptoms

1.3.3.1 Respiratory Subsystem

1.3.3.3 Articulatory Subsystem

Velopharyngeal function may also be affected by muscle rigidity, bradykinesia and

1.3.3.5 Prosodic Features

1.3.3.6 Overall Speech Intelligibility

1.3.4 Occurrence of Hypokinetic Dysarthria

1.3.6 Impact of Hypokinetic Dysarthria on Communication

1.4 Parkinson-plus Syndromes

1.4.1 Progressive Supranuclear Palsy

1.4.2 Multiple System Atrophy

MSA represents a number of disorders of similar neuropathology, with extensive neural

1.5.1 Pharmacological Management

1.5.2 Neurosurgical Management

Unlike the permanent effects of pallidotomy and thalamotomy, DBS is a reversible

1.5.2.1 Speech and Voice Changes Associated with Neurosurgical Management

1.5.3.4 The Lee Silverman Voice Treatment ® and Parkinson-plus Syndromes

If telehealth is understood to mean the integration of telecommunication systems into the

Figure 1.4 A schematic representation of the associated telehealth, telemedicine, telerehabilitation