Disability and Rehabilitation

ISSN: 0963-8288 (Print) 1464-5165 (Online) Journal homepage: http://www.tandfonline.com/loi/idre20

Implementation of a respiratory rehabilitation

protocol: weaning from the ventilator and
tracheostomy in difficult-to-wean patients with
spinal cord injury

Ibrahim Gundogdu, Erhan Arif Ozturk, Ebru Umay, Ozgur Zeliha Karaahmet,
Ece Unlu & Aytul Cakci

To cite this article: Ibrahim Gundogdu, Erhan Arif Ozturk, Ebru Umay, Ozgur Zeliha Karaahmet,
Ece Unlu & Aytul Cakci (2016): Implementation of a respiratory rehabilitation protocol: weaning
from the ventilator and tracheostomy in difficult-to-wean patients with spinal cord injury,
Disability and Rehabilitation, DOI: 10.1080/09638288.2016.1189607

To link to this article: http://dx.doi.org/10.1080/09638288.2016.1189607

Published online: 23 Jun 2016.

Submit your article to this journal

Article views: 30

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=idre20

Download by: [Library Services City University London] Date: 04 July 2016, At: 07:50
 DISABILITY AND REHABILITATION, 2016
http://dx.doi.org/10.1080/09638288.2016.1189607

RESEARCH ARTICLE

Implementation of a respiratory rehabilitation protocol: weaning from the

ventilator and tracheostomy in difficult-to-wean patients with spinal cord injury
Ibrahim Gundogdu, Erhan Arif Ozturk, Ebru Umay, Ozgur Zeliha Karaahmet, Ece Unlu and Aytul Cakci
Physical Therapy and Rehabilitation Clinic, Ministry of Health Ankara Diskapi Yildirim Beyazit Education and Research Hospital, Ankara, Turkey

ABSTRACT ARTICLE HISTORY

Purpose: Following repeated weaning failures in acute care services, spinal cord injury (SCI) patients who Received 13 October 2015
require prolonged mechanical ventilation and tracheostomy are discharged to their homes or skilled nurs- Revised 30 April 2016
ing facilities, with a portable mechanical ventilator (MV) and/or tracheostomy tube (TT) with excess risk of Accepted 10 May 2016
Downloaded by [Library Services City University London] at 07:50 04 July 2016

complications, high cost and low quality of life. We hypothesized that many difficult-to-wean patients with Published online 14 June
2016
cervical SCI can be successfully managed in a rehabilitation clinic. The aim of our study was to develop a
respiratory rehabilitation, MV weaning and TT decannulation protocol and to evaluate the effectiveness of KEYWORDS
this protocol in tetraplegic patients. Mechanical ventilation;
Methods: A multidisciplinary and multifaceted protocol, including respiratory assessment and management respiratory rehabilitation;
themes, was developed and performed based on the findings from other studies in the literature. spinal cord injury;
Tetraplegic patients with the diagnosis of difficult-to-wean, who were admitted to the rehabilitation clinic tracheostomy decannulation;
after having been discharged from the intensive care unit to their home with home-type MV and/or TT, tetraplegia; weaning
were included in this prospective observational study.
Results: The respiratory rehabilitation protocol was applied to 35 tetraplegic patients (10 home-type MV
and tracheostomy-dependent, and 25 tracheostomized patients) with C1-C7 ASIA impairment scale grade
A, B, and C injuries. Seven out of 10 patients successfully weaned from MV and 30 of 35 patients were dec-
annulated. Four patients were referred for diaphragm pace stimulation and tracheal stenosis surgery. The
mean durations of MV weaning and decannulation were 37 and 31 days, respectively.
Conclusions: A multifaceted, multidisciplinary respiratory management program can change the process of
care used for difficult-to-wean patients with SCI.

ä IMPLICATIONS FOR REHABILITATION

 Findings from this study indicate the significance of a multidimensional evaluation of any reversible
factors for prolonged MV- and/or TT-dependent SCI patients. Thus, rehabilitation specialists should
take this into consideration and should provide the appropriate amount of time to these patients.
 The proposed protocol of respiratory rehabilitation for MV- and/or TT-dependent SCI patients shows
promising results in terms of changing the care used for these patients.
 Successful implementation of a respiratory rehabilitation and weaning protocol is dependent on care-
ful planning and detailed communication between the rehabilitation specialist and intensivist during
the respiratory rehabilitation process.
 Because many of the so-called difficult- or impossible-to-wean patients were successfully weaned from
MV and TT in the PMR clinic, the need for such an outlet for countries without specialized centers is
supported.

Introduction
Spinal cord injury (SCI) severely compromises both sensory and
motor function and causes lifelong disability.[1] Respiratory compli-
cations are a major cause of morbidity and mortality during the
acute and chronic phases after SCI.[1–5] All the patients with an
injury level above T12 are at risk for respiratory complications (i.e.
atelectasis, pneumonia, respiratory insufficiency) as a result of
inspiratory/expiratory muscle paralysis; the higher the level of the
injury, the greater the likelihood that a respiratory impairment will
occur.[1] Cervical SCI patients with injuries above the level of the
phrenic motor neurons (C3, 4, and 5), causing both diaphragm
(i.e. leading to inspiratory failure) and expiratory muscle paralysis
(i.e. leading to inadequate clearance of the secretions related to
ineffective cough), and the patients with a lower level of cervical
SCI, causing accessory inspiratory and expiratory muscle paralysis,
may require tracheostomy cannulation and mechanical ventilation
support at the early stage,[4] which is generally initiated in the
intensive care unit (ICU) of short-term acute care (STAC) hospi-
tals.[2] Although the partial recovery of respiratory functions may
occur over the year following injury,[3] 4% of SCI patients require
long-term mechanical ventilation and tracheostomy tube (TT) or
diaphragm-pacing stimulation.[6] However, several risk factors
regarding respiratory complications, quality of life (QoL), financial

CONTACT Ibrahim Gundogdu ibrahimftr@gmail.com Rehabilitation Specialist, Department of Physical Therapy and Rehabilitation, Ministry of Health Ankara
Diskapi Yildirim Beyazit Education and Research Hospital, Irfan Bastug Cd. Diskapi. Ankara, Turkey
ß 2016 Informa UK Limited, trading as Taylor & Francis Group
 2 I. GUNDOGDU ET AL.
cost and family caregiving burden are frequently associated with
mechanical ventilator (MV) and TT;[7,8] thus, it is important to
identify patients who are candidates for weaning from MV and
decannulation.
In some countries, patients with slowly resolving acute respira-
tory dysfunction remain MV- and TT-dependent and are trans-
ported to long-term acute care (LTAC) hospitals, transitional care/
weaning units of STAC or rehabilitation hospitals for weaning trials.
Other countries, including our country, have neither specialized
weaning centers nor adequate ICU beds for the weaning of SCI
patients requiring prolonged mechanical ventilation (PMV); many
of these patients do not require the intensity of treatment or mon-
itoring available in an ICU, and a rehabilitative care setting would
be appropriate for PMV patients.[2] Although there is a growing
body of research on MV and TT weaning of SCI patients, the
implementation of a comprehensive respiratory rehabilitation
protocol, including multifaceted respiratory evaluation and man-
Downloaded by [Library Services City University London] at 07:50 04 July 2016
agement methods during the liberation from MV and TT support,
has not been reported. Moreover, there is an ongoing decline for
home-type MV- and/or TT-dependent SCI patient admission to
rehabilitation clinics in our country, which may worsen the situ-
ation and result in a vicious circle of worsening anxiety and per-
sisting physical disability.
To address the above issues, we have developed a respiratory
rehabilitation, home-type MV weaning and TT decannulation proto-
col (RRWT), including respiratory assessment and management
themes for cervical SCI patients based on findings obtained from
other studies in the literature, and evaluated the effectiveness of
these protocols in a rehabilitation clinic of a tertiary hospital.
Methods
Subjects
We studied 35 tetraplegic patients; 10 with MV and TT, and 25 with
only TT, who were consecutively admitted to the physical medicine
and rehabilitation clinic of a tertiary hospital over a three-year
period (from 2012 to 2015). Inclusion criteria were patients aged
between 15 and 70 years with cervical SCI levels ranging from C1
to C7 ASIA impairment scale (AIS) grade A, B, and C injuries, and
with MV and/or TT for a prolonged period was defined as the need
for 21 consecutive days of mechanical ventilation.[2] Exclusion cri-
teria included patients requiring invasive monitoring.
The reasons for admission into the physical medicine and
rehabilitation (PMR) clinic were as follows: (i) to attempt weaning
from PMV; and/or (ii) to attempt TT decannulation and monitor their
respiratory status due to the high risk of respiratory complications.
MV-dependent patients underwent respiratory rehabilitation,
MV weaning and a TT decannulation program, and tracheostom-
ized patients underwent respiratory rehabilitation and a TT decan-
nulation program for periods from 3 to 12 weeks. During their
hospitalization, patients also performed all the activities of a com-
prehensive rehabilitation program. All patients and their caregivers
received information about RRWT upon admission, and their writ-
ten consents were obtained at the beginning of the study. The
approval of the ethical board of the hospital was obtained, and
the study was conducted according to the principles of the
Declaration of Helsinki.
Setting
The PMR clinic is a forty-bed service in a tertiary hospital. There
were four out of 40 beds equipped with monitoring facilities, such
as pulse oximetry and heart rate control, as well as continuous
electrocardiography, if needed. There are no resources for
noninvasive MV, invasive MV or end-tidal CO2 monitoring, and
patients were expected to be weaned from home-type MV in the
PMR clinic. The nurse-to-patient ratio was 1:2 for the day shift and
1:4 for the night shift for this unit of four beds. To maximize the
potential of the unit, we devised a management strategy that uti-
lized all available resources, including ICU, pulmonary and oto-
laryngology clinics. A pulmonologist and intensivist who staffed
the main neurological ICU at our hospital consulted MV-dependent
patients daily, and patients were monitored by the ENT specialist
as needed. In addition, daily assessment by a physiatrist, physical
and respiratory therapists, clinic psychologist, social service special-
ist and rehabilitation nurse is a component of the routine care.
The patients’ readiness for weaning was determined by an inter-
disciplinary team: a rehabilitation physician, intensivist, and readi-
ness for decannulation, which were determined by the
rehabilitation specialist, in consultation with the ENT physician.
Designing an evidence-based RRWT for home-type MV weaning
of SCI patients
We performed a PubMed literature search using the MeSH terms
and keywords: ‘‘spinal cord injuries’’, ‘‘respiratory therapy’’, respira-
tory rehabilitation’’, ‘‘mechanical ventilation’’, ‘‘ventilator weaning’’
and ‘‘tracheostomy decannulation’’ for relevant abstracts prior to
designing the RRWT. Full-length papers containing relevant find-
ings were screened by two independent reviewers, and studies
involving randomized and non-randomized controlled trials and
case series were used as extremely few studies involving SCI
patients are prospective or randomized trials. We were unable to
locate clinically useful and evidence-based respiratory rehabilita-
tion, weaning from home-type ventilator or decannulation proto-
cols for cervical SCI patients. Thus, our clinical team developed
and implemented one such protocol. Analysis of the literature
revealed that respiratory system assessment, respiratory rehabilita-
tion including secretion removal techniques, respiratory muscle
training, MV weaning and tracheostomy decannulation were
essential components of any regimen designed to rehabilitate and
wean MV- and/or TT-dependent cervical SCI patients. The protocol
used was as follows:
A. Respiratory assessments
1. Nutritional assessment
2. Respiratory muscle strength evaluation
3. Cough strength evaluation
4. Diaphragm assessment
5. Dysphagia and aspiration assessment
6. Periodic assessments
7. Red flags
B. Respiratory management
1. Clearance of airway secretions
2. Ventilator muscle training
3. MV weaning
4. Swallowing therapy
5. Tracheostomy decannulation
6. Discharge planning
Respiratory assessments
Nutritional assessment
Because malnutrition can affect the strength and endurance of the
diaphragm and accessory muscles, patients were assessed using
the spinal nutrition screening tool [9] and energy expenditure was
measured by a dietician to ensure sufficient caloric intake to
improve respiratory muscle function.

Respiratory muscle strength evaluation
To measure the maximum inspiratory pressure (MIP) and max-
imum expiratory pressure (MEP) (Micro Medical Micro RPM, Micro
Medical Ltd, Rochester, UK), reflecting the strength of the respira-
tory muscles and weaning potential, the subject performed a max-
imum inspiratory effort after the maximum expiration, and a
maximum expiratory effort after a maximum inspiration, respect-
ively. The lowest negative value for the MIP and the highest posi-
tive value for the MEP in three or more attempts were chosen for
data analyses.[10,11]
Cough strength evaluation
To assess expiratory muscle weakness (reduced cough capacity)
and the decannulation potential in patients with SCI, peak cough
flow (PCF) was measured using peak flow meter (REF: HS756 -012
Phillips Respironics, Edison, NJ), by having the patient cough as
Downloaded by [Library Services City University London] at 07:50 04 July 2016
forcefully as possible through the peak flowmeter. The highest
value in at least three trials was used for data analyses.[10]
Diaphragm assessment
While the MIP reflects the strength of the inspiratory muscles,
weaning problems are usually ascribed to the involvement of the
phrenic nerves and the diaphragmatic function should be eval-
uated to predict weaning from MV.[12] Thus, four methods were
used to assess diaphragmatic function; PA chest radiograph (i.e.
diaphragmatic paralysis is suspected when asymmetric diaphrag-
matic elevation is observed on the PA chest radiograph), fluoros-
copy (i.e. the paradoxical movement of the paralyzed
hemidiaphragm), phrenic nerve conduction studies (i.e. provide
useful measures in assessing the recovery of lower–motor neuron
diaphragm function) and diaphragm needle EMG (i.e. demon-
strates voluntary active motor unit action potentials that provides
additional electrophysiological support for phrenic nerve function).
Phrenic nerve conduction studies and diaphragm needle EMG
were completed prior to attempted weaning, as demonstrated by
Kimura and Bolton.[13,14] These methods were also used in the
preoperative evaluation of high tetraplegic candidates for the dia-
phragm pace system.
Dysphagia and aspiration assessment
Because MV and TT are thought to be risk factors for dysphagia,
which can increase the risk for pulmonary complications in individ-
uals with cervical SCI, the bedside dysphagia screening test (BDST)
and flexible fiber optic endoscopic evaluation of swallowing (FEES),
which are routinely used in our clinic, were used to test dysphagia
[15,16] after MV-dependent patients had been weaned from venti-
lation. Tracheostomized patients were evaluated on the first day of
admission. A BDST score of 0–2 was accepted as normal swallowing
and a score between 3 and 6 was considered to be dysphagia. In
cases of negative BDST, further bedside evaluation using the FEES
test was performed by the same otolaryngology practitioner using
a non-ducted fiber optic nasopharyngoscope of 3.4 mm diameter, a
light source, camera, monitor, and DVD recorder. At the end of the
examination, dysphagia level was scored from 1 to 6 according to
the endoscopic assessment of dysphagia protocol.[16] Accordingly,
1 point was considered as ‘‘normal swallowing’’ while 2–6 points
were considered ‘‘dysphagia’’. Flexible fiber optic endoscopic evalu-
ation was also used to diagnose complications related to tracheos-
tomy (i.e. tracheal stenosis, tracheomalasia, vocal cord paralysis)
and to remove mucous plugs during RRWT.
RESPIRATORY REHABILITATION FOR SCI PATIENTS 3
Periodic assessments
The initial laboratory assessment included; PA chest radiograph,
ECG, routine laboratory studies (complete blood count, chemistry
panel, C-reactive protein, arterial blood gas). All patients were
monitored with indicators including rising temperature, change in
respiratory rate, shortness of breath, increasing pulse rate, increas-
ing anxiety, increased volume of secretions, frequency of suction-
ing, tenacity of secretions, and declining PCF rate for the
development of atelectasis and pneumonia,[17] and each under-
went careful auscultation to detect abnormalities associated with
the lung on a daily basis. O2 saturation (SO2) was simultaneously
measured using a pulse oximeter (Beurer PO 30, Ulm, Germany),
with a fingertip sensor at the right index finger throughout the
hospitalization period (i.e. patients on MV, during MV weaning,
respiratory rehabilitation and tracheostomy decannulation proc-
esses). Arterial blood gas (ABG) analysis was also performed on full
ventilatory support prior to spontaneous breathing trials, and prior
to tracheostomy decannulation. Chest X-rays, routine laboratory
studies and ABG were repeated when necessary for further evalu-
ation of respiratory symptoms. The MIP, MEP and PCF were meas-
ured and recorded before and after the protocols.
Red flags
The patients’ responses to the effects of RRWT were meticulously
monitored. To continue respiratory rehabilitation, weaning or the
decannulation protocol, patients were required to remain within
normal ranges for SO2 (90–100%), heart rate (>70 bpm and <120
bpm), systolic blood pressure (>90 mmHg and <140 mmHg), and
RR (>12 bpm and <30 bpm) and show no subjective evidence of
respiratory distress or anxiety.
Respiratory management
Clearance of airway secretions
All the patients were encouraged to maintain maximum mobility
and upright positioning. An abdominal binder was used in the
seated or in an upright position to enhance cough in cervical SCI
patients.[7,18] High frequency chest wall oscillation (Vest Airway
Clearance System, Hill-Rom, St. Paul, Minnesota) therapy (HFCWO)
[19] was administered to patients at 20 Hz for 20 min once or twice
a day for 3 to 12 weeks according to the need for secretion
removal. After completion of HFCWO therapy, all patients received
chest physiotherapy including turning, postural drainage, vibration,
percussion and shaking for 30 to 60 min once or twice a day by
the same respiratory therapist throughout the rehabilitation period.
Manual assisted cough, cough assist machine (CoughAssist CA 3000,
Murrysville, PA) and tracheal suctioning were used to remove secre-
tions and to enhance cough.[20–22] Excessive coughing, wheezing
or decreased SO2 measurements were deemed potential indica-
tions for manual assisted cough, cough assist machine or suction-
ing. In addition, bronchodilators (aerosolized ipratropium bromide)
[17,23] and mucolytics (nebulized sodium bicarbonate) [17] were
prescribed by a pulmonologist as needed.
Inspiratory muscle training
After completing secretion clearance, inspiratory muscle training
(IMT) was initiated when the attending physicians determined that
the patient was medically stable.[21,24] All patients about to
undergo IMT had their trach tube cuffs deflated and MV-depend-
ent patients were removed from the ventilator. A red
 4 I. GUNDOGDU ET AL.
decannulation cap was placed on the TT and nose clips were
applied. IMT sessions were performed with the patient in bed with
an approximately 30 to 45 head-up tilt after chest physiotherapy.
A threshold IMT device (REF HS730EU, Parsippany, NJ) was used
with training sessions consisted of 10 repetitions, three times a
day, five days per week until TT decannulation. Supplemental oxy-
gen was added and MV-dependent patients were returned to MV
for rest between training sets as needed. Following each training
session, the patient indicated a rating of perceived inspiratory
exertion on a linear scale of 0 to 4, with 0 representing no inspira-
tory effort and 4 indicating maximal inspiratory effort. If an exer-
tion rating was between 0–1, and if the patients were able to
perform three training sessions without any exertion, then the
pressure was increased. If the exertion rating was 3 or 4, then the
pressure was reduced. Patients were observed during training for
the red flags mentioned above.
Downloaded by [Library Services City University London] at 07:50 04 July 2016
MV weaning
All MV-dependent patients received pressure-support MV with
back-up rates of six to fifteen breathes. Patients were considered
eligible for weaning under the following conditions: the patient
was fully awake, afebrile, medically and respiratorily stable for at
least 24 h, and psychologically willing and ready to participate, and
had PaO2  75 mmHg, PCO2 ¼ 35–45 mmHg, pH ¼7.35–7.45,
PaO2/FiO2 > 150, PEEP 5–8, manageable secretions, and a clear
chest X-ray. The start of a weaning attempt was defined as ventila-
tor-free breathing for a minimum of one minute. Patients were
weaned from pressure support ventilation through a process of tri-
als of spontaneous breathing,[21,24–26] which were progressively
greater in duration and frequency as tolerated. Spontaneous
breathing trials (SBT) were terminated when the patients exhibited
signs of respiratory distress such as tachypnea, SO2 decreasing 5%
from baseline or below 90%, hypotension, tachycardia, or subject-
ive signs of respiratory distress or anxiety. The patients were
returned to MV for rest between the trials. Each patient’s progress
was monitored by ICU and rehabilitation specialists. In addition,
anxiolytics [27] were prescribed for weaning anxiety by a psych-
iatrist as needed. Patients were considered to have been success-
fully weaned from MV when full-time removal from the ventilator
for 2 weeks or more had been achieved.
Swallowing therapy
All dysphagic patients were included in the traditional swallowing
therapy during the TT decannulation process and received daily
care for oral hygiene, thermal (cold) and tactile stimulation, swal-
lowing maneuvers, head and trunk positioning, dietary modifica-
tion, and oral motor exercises, including the lip, tongue and jaw
movements according to the different patient characteristics for
60 min a day, 5 days a week for 3 to 6 weeks.
Tracheostomy decannulation
Tracheotomy removal was considered if patients were clinically
stable, demonstrated adequate swallowing and were able to
expectorate. The criteria for tracheotomy decannulation were sta-
ble ABG (PaO2 > 60 mmHg, normal PaCO2), absence of distress,
hemodynamic stability, absence of fever or active infection, mean
SO2 greater than 90%, normal fiber optic endoscopic examination
of the trachea, normal swallowing tests, ability to tolerate TT clos-
ure for 24 h and ability to generate PCF greater than 160 L/
min.[7,8,10,17,28] Following assessments, patients demonstrated
gradually increasing TT closure trials as tolerated. TT closure trials
were terminated when the patients exhibited signs of respiratory
distress or subjective signs of respiratory distress or anxiety. The
TTs were removed by an ENT specialist. A spare tracheostomy of
the same size and one size smaller was at the bedside in case of
the need for recannulation.
Discharge planning and vaccination
Before discharge, all patients and caregivers were given details of
a telephone hotline run by the PMR clinic that they could contact
and were educated about the disease state, respiratory complica-
tions and follow-up, respiratory exercises and prevention of
respiratory infections. All patients were vaccinated against
pneumococcus at the time of discharge and advised regarding
seasonal influenza vaccination.[17]
Data analysis
Statistical analysis was performed using the Statistical Package for
Social Sciences (SPSS, Inc.; Chicago, IL) version 16.0 for Windows.
Normality of the continuous variables was assessed using a
Shapiro–Wilk test. Descriptive statistics were expressed as the
mean ± standard deviation. Statistically significant differences in
repeated measurements within the groups were evaluated using a
Wilcoxon signed rank test. A Mann–Whitney U test was used for
the comparison of non-normally distributed continuous variables.
Differences were considered significant if p < 0.05.
Results
The RRWT protocol was applied to 35 tetraplegic patients (10
tetraplegic patients with MV and TT, 25 patients with only TT);
seven patients had high tetraplegia (injury level cephalad to C4)
and 28 patients had low tetraplegia (injury level C4 or caudal).
Twenty-two patients had complete SCIs and 13 patients had
incomplete SCIs, as defined by the International Standards for
Neurological Classification of Spinal Cord Injury. All patients
(n ¼ 35) had a cuffed TT, and MV-dependent patients were venti-
lated with Puritan Bennett 560 (Boulder, CO) home-type MV.
The mean age was 29.2 years, 20% (n ¼ 7) were female, 80%
(n ¼ 28) were male. Motor vehicle accident was the main cause of
injury (21 [60%]). The mean durations of ICU, MV and TT were 57,
64 and 90 days, respectively. In total, 4 pneumohemothorax, 4
atelectasis and 5 pneumonia cases were detected in the ICU fol-
low-up recordings of the 35 patients. Four out of 10 MV-dependent
patients (C4, C5 and C6 injuries) had a history of pneumohemo-
thorax and 2 of these patients also had ventilator-associated pneu-
monia during the ICU period. The frequency of endotracheal
suctioning for patients ranged between 4 to 15 times daily at the
time of admission. Malnutrition was detected in 18 (51.4%) out of
35 patients. The demographic and disease characteristics of the
patients are shown in Table 1, and the summary of the PMR
admissions are shown in the flow diagram (Figure 1).
Diaphragmatic evaluation of MV-dependent patients revealed
that seven patients had normal chest X-ray, phrenic nerve conduc-
tion study and diaphragm needle EMG results upon admission.
Phrenic nerve conduction studies revealed an absent response
from both diaphragms in one patient (patient 1) and reduced
amplitude from the left diaphragms of two patients (patient 5 and
7). Diaphragm needle EMG showed fibrillations, positive sharp
waves and single polyphasic MUAP, suggesting severe partial
axonal degeneration of the phrenic nerves, which indicated bilat-
eral (patient 1) or unilateral diaphragm paralysis (patients 5 and 7)
in these patients. Fluoroscopy revealed no diaphragm movement
 RESPIRATORY REHABILITATION FOR SCI PATIENTS 5

Table 1. Descriptive characteristics of patients. in C1 AIS-A patients (patients 1, 2 and 3) and paradoxical move-
MV-dependent Tracheostomized ment in 2 patients (patients 5 and 7) (Table 2).
Parameter patients patients Total A C1 AIS-A patient (patient 1) with severe bilateral diaphragm
Patients (n) 10 25 35 paralysis could tolerate SBT only for one minute on the first day of
Mean age (years) 27.1 ± 16.4 30.0 ± 10.3 29.2 ± 12.1 admission. After 30 days of RRWT, the SBT time increased to
Gender (F/M) 1/9 6/19 7/28
100 min/day. On the third month, needle EMG revealed the same
Mean ICU duration (days) 67.8 ± 22.8 53.2 ± 44.0 57.4 ± 39.4
Mean MV duration (days) 136.9 ± 73.7 35.2 ± 35.7 64.3 ± 67.1 results but with polyphasic MUAPs and the patient tolerated SBT
Mean TT duration (days) 118.3 ± 68.8 78.8 ± 55.8 90.1 ± 61.5 for 12 to 16 h during the daytime before being discharged from
Etiology the hospital.
MVA 4 17 21 The other two C1 AIS-A patients (aged 20 and 25 years) with
Fall from height 2 4 6 intact diaphragms (patients 2 and 3) had a history of 2 months of
Sports injury 2 3 5
an ICU, and 9 and 10 months of MV duration. Respiratory stabiliza-
Violence 1 1 2
Diving injury 1 1 tion was established, and the need for endotracheal suctioning
AIS was diminished, such that they could tolerate SBT for 5 to 10 min
Motor complete after 6 weeks of the program. These patients were deemed eligible
C1 3 3 for the diaphragm pacing stimulation (DPS) system and they were
C4 2 5 7 successfully implanted (intramuscular) in another hospital.
Downloaded by [Library Services City University London] at 07:50 04 July 2016

C5 2 9 11
The means of the MIP, MEP and PCF values of MV-dependent
C6 1 4 5
C7 3 3 patients before/after the weaning protocol were 18.6 ± 11.2/
Motor incomplete 41.1 ± 18.1, 19.0 ± 9.8/40.0 ± 13.9 and 46.0 ± 49.9/130 ± 71.3,
C2 1 1 respectively. All three variables were significantly improved after
C3 1 2 3 the weaning protocol (p ¼ 0.005, p ¼ 0.005 and p ¼ 0.012, respectively).
C4 2 2
Pre/post treatment PCF values were also significantly improved in tra-
AIS: ASIA impairment scale; MVA: motor vehicle accident; MV: Mechanical ventila- cheostomized patients (p < 0.001).
tion; ICU: intensive care unit; TT: tracheostomy tube.

Reason for admission: N=35

Weaning from MV + TT = TT decannulation:25

10

RESPIRATORY
ASSESSMENT

Paralytic diaphragm= 3 TT decannulation criteria

Intact diaphragm = 7

RESPIRATORY
REHABILITATION

Respiratory rehabilitation Respiratory rehabilitation

MV Weaning protocol TT Decannulation protocol
TT Decannulation protocol

Discharge Discharge

Weaned + decannulated =7 Decannulated = 23

Partially weaned =1 Referred for tracheal stenosis surgery
Referred for DPS=2 =2

Figure 1. Outcome according to reason for admission. TT: tracheostomy tube; MV: mechanical ventilator; DPS: diaphragmatic pace stimulation.
 6 I. GUNDOGDU ET AL.

Table 2. Diaphragmatic evaluation of MV-dependent patients.

Patient AIS Chest X-ray Conduction study Needle EMG Fluoroscopy
1 C1 AIS-A Normal Absent response Bilateral diaphragmatic paralysis No movement
2 C1 AIS-A Normal Normal Normal No movement
3 C1 AIS-A Normal Normal Normal No movement
4 C2 AIS-C Normal Normal Normal Normal
5 C3 AIS-C Left diaphragmatic elevation Reduced amplitude of left phrenic nerve Left diaphragmatic paralysis Paradoxical movement
6 C4 AIS-A Normal Normal Normal Normal
7 C4 AIS-B Left diaphragmatic elevation Reduced amplitude of left phrenic nerve Left diaphragmatic paralysis Paradoxical movement
8 C5 AIS-A Normal Normal Normal Normal
9 C5 AIS-A Normal Normal Normal Normal
10 C6 AIS-A Normal Normal Normal Normal
AIS: ASIA impairment scale; EMG: Electromyography.

Table 3. Respiratory and swallowing evaluation of patients.

MV-dependent patients (n ¼ 10) Tracheostomized patients (n ¼ 25)
Mean MIP (cmH2O) (Pre/post) 18.6 ± 11.2/40.0 ± 16.9 p ¼ 0.005
Mean MEP (cmH2O) (Pre/post) 19.0 ± 9.8/40.0 ± 13.9 p ¼ 0.005
Downloaded by [Library Services City University London] at 07:50 04 July 2016

Mean PCF (L/d) (Pre/post) 46.0 ± 49.9/130 ± 71.3 p ¼ 0.012 88.0 ± 28.5/166.4 ± 24.9 p < 0.001
Mean BDST (Pre/post) 2.9 ± 1.37/1.4 ± 1.5 p ¼ 0.004 2.0 ± 1.19/0.52 ± 0.58 p < 0.001
Mean FEES (Pre/post) 3.0 ± 0.7/1.0 ± 0.0 ¥ p ¼ 0.039 1.56 ± 0.85/1.0 ± 0.0 ‡ p ¼ 0.023
Dysphagia with BDST (Pre/post)(n) 12/2
Dysphagia with FEES (n) 11/0
Total dysphagia (Pre/post) (n) 23/2
Weaning time (d) 37.0 ± 11.6*
TT decannulation time (d) 31.7 ± 16.9†
Weaned from MV (n) 7/10 (70%)
Decannulated (n) 30/35 (85.7%)
MV: Mechanical ventilator; MIP: maximum inspiratory pressure; MEP: maximum expiratory pressure; PCF: peak cough flow;
BDST: bedside dysphagia screening test; FEES: fiberoptic endoscopic evaluation of swallowing; *: Mean of 7 patients
excluding partially weaned or unweanable patients; †: mean of 30 patients; ¥: Mean of 5 patients; ‡: mean of 18 patients.

Seven out of 10 MV-dependent patients were successfully Discussion

weaned from MV (70%) and TT, and tracheostomy closure was
possible in 23 out of the 25 patients who had been admitted for
TT decannulation (96%). In total, 30 out of 35 patients were decan-
nulated (85.7%). The mean duration of weaning from MV and TT
decannulation were 37.0 ± 11.6 and 31.7 ± 16.9 days, respectively.
The TT decannulation time was calculated after weaning from the
ventilator for MV-dependent patients.
Pre-treatment BDST scores revealed that 12 out of 35 (34%)
patients had dysphagia. When negative BDST scores were
obtained, further evaluation with FEES (n ¼ 23) indicated that 11
out of 23 (47.8%) patients had dysphagia before swallowing ther-
apy. In total, twenty-three patients (65.7%) had dysphagia. Pre/
post treatment BDST and FEES scores of both groups were signifi-
cantly improved (p ¼ 0.004, p < 0.001, p ¼ 0.039, and p¼ 0.023).
After the rehabilitation period, no patient had dysphagia, exclud-
ing 2 unweanable patients. In addition, the nutritional status of all
patients was normal. A comparison of the pre- and post-treatment
changes of respiratory and swallowing data is provided in Table 3.
Two patients who were not able to undergo decannulation
were diagnosed with severe tracheal stenosis and referred for sur-
gery. Patient outcomes are shown in the flow diagram in Figure 1.
Two patients showed symptoms, including dyspnea, tachypnea
and sudden decrease in SO2 during the hospitalization period.
Pneumonia, atelectasis, aspiration and pulmonary emboli were
excluded in the differential diagnosis. Multiple cough assists and
tracheal suctioning efforts provided no success. They were each
diagnosed with a mucous plug, which were treated with bronchos-
copy in consultation with the chest disease clinic. Two patients
had atelectasis and ventilator-associated pneumonia (VAP), which
were treated with respiratory rehabilitation and appropriate antibi-
otics. Three out of 10 MV-dependent patients were diagnosed with
weaning anxiety by a psychiatrist and were treated with 50 to
100 mg quetiapine.
In patients requiring MV þ TT or only TT, the main goal of respira-
tory rehabilitation is weaning from MV and decannulation of TT.
This process requires a multidimensional evaluation of the respira-
tory system, and aggressive respiratory care interventions need to
begin immediately after injury. If these interventions fail to
improve the respiratory status in acute care services, then more
aggressive specialized respiratory interventions, such as those uti-
lized in respiratory rehabilitation units may need to be
implemented.[2,29]
The patients included in the study, had failed to wean from MV
or TT in an ICU and were discharged with the diagnosis of diffi-
cult- or impossible-to-wean. The length of ICU stay and the dura-
tions of MV and TT were longer than the periods reported in
previous studies.[21,24–26,29] It has been reported that the sever-
ity of SCI is an independent risk factor for weaning from MV and
TT.[8,29] Most of the patients reported in this study, particularly
those with cervical lesions at C4 or higher (45.7%) and complete
injuries (62.8%), were at a higher risk for adverse outcomes. Due
to the heterogeneity of the groups, it is difficult to establish com-
parisons between our study and those performed in previous ser-
ies. Previous studies have been published describing the
effectiveness of one respiratory rehabilitation method,[21,24] com-
parison of two weaning methods [26] or effectiveness of an inter-
mediate care center [25,29] in SCI patients requiring MV or TT, and
most of these studies were performed in acute stages. We propose
that the rates of successful weaning from MV and TT are substan-
tial, 70% and 85%, respectively, which were comparable or higher
than those reported in previous weaning studies, 35%,[29]
57%,[21] and 83% in SCI patients,[26] and 89–90% in unselected
trauma patients,[24,25] and were comparable to previously
reported decannulation rates of 84%.[7,8] Furthermore, the success
rates were even better if we eliminate C1 AIS-A patients, in whom

weaning is impossible due to the severity of SCI, and patients who
needed tracheal stenosis surgery.
Two out of three C1 AIS-A patients (patients 2 and 3) with
intact diaphragms experienced modest improvements in SBT
times, potentially due to the strengthening of accessory respiratory
muscles as there is an interruption of the descending bulbospinal
respiratory pathways to C3–4-5 phrenic nerve motor neurons that
innervate the diaphragm. Among the most important reasons for
these patients attempting RRWT is the need to spontaneously ven-
tilate in case of accidental disconnection from the MV and to
detect the eligibility of patients for implanting a DPS system.
Indications for the DPS system correspond to MV dependance sec-
ondary to respiratory failure, which is mainly caused by high-level
cervical SCI.[30] There is no equivalent alternative to propose to
these patients to eliminate morbidity and mortality related to
long-term MV use. In addition, the timing of the DPS system for
high-level cervical SCI is contentious.[31] The two C1 AIS-A SCI
Downloaded by [Library Services City University London] at 07:50 04 July 2016
patients in the study were MV-dependent for 10 to 11 months
when referred for DPS implantation, which is a reasonable time
delay for implantation if we consider the suggested reasonable
time delay of 6–18 months for implantation.[30] In our opinion,
with careful multidisciplinary neurological and respiratory assess-
ment, if respiratory rehabilitation programs were applied to these
patients after the 2 months of the ICU period, then the decision
for the DPS system could be given even earlier. A recent review
suggests that an earlier implantation may be safe and result in
better outcomes in selected patients.[31]
One C1 AIS-A patient with severe diaphragm paralysis (patient
(1) had a C3-C4 level dislocation injury. Although the injury level
was evaluated as C1 AIS-A, we propose that this patient actually
had a C3 level of injury, as the patient had been partially weaned
from MV. This case may reflect a problem regarding the AIS defin-
ition as previously reported,[32] and may indicate that the progno-
sis decision should be based on a multidimensional evaluation of
these patients rather than ASIA examination.
All of the tetraplegic patients in this study, except for the three
C1 AIS-A patients, achieved weaning from MV using the RRWT.
Because most of the low tetraplegic patients (C4–6) do not require
MV or require several weeks for weaning (21), prolonged ventilator
dependency among these patients (C4–6) may be explained by a
history of pneumohemothorax or ventilator-associated pneumonia
during the ICU follow-up. Neither of the C1 AIS-A tetraplegic
patients achieved a discontinuation from mechanical ventilation,
although it may be said that their dependance on mechanical ven-
tilation was lessened and the strength of available accessory
respiratory muscles was improved.
The weaning duration in our patients (37.0 ± 11.6 days) is simi-
lar to other series of acute SCI patients (44 and 47 days).[24,29]
The TT decannulation duration of our patients (31.7 ± 16.9 days)
was also similar to decannulation studies (16 to 45 days),[7,8] but
we neither calculated the weaning duration from MV to weaning
nor the TT decannulation duration from tracheostomy to decannu-
lation. Instead, we calculated this value from the start of the RRWT
protocol because our patients had required long-term MV
(136.9 ± 73.7 days) or tracheostomy (90.1 ± 61.5 days) in the com-
munity after being discharged from acute care services.
Depending on the level and completeness of the injury, inad-
equacy of the respiratory muscles results in ineffective inspiration
and an ineffective cough strength.[21] The MIP and MEP are pre-
dictors of successful weaning while the PCF is a predictive index
for tracheostomy decannulation.[11] In our clinical experience, they
also provide visual biofeedback to patients during respiratory
rehabilitation. In this study, most of the patients experienced
improvements in MIP, MEP and PCF values, which were needed to
RESPIRATORY REHABILITATION FOR SCI PATIENTS 7
either follow the recovery process of the respiratory muscles or
decide the time of decannulation.[28] In addition, long-term
tracheostomy is a risk factor for dysphagia and aspiration pneumo-
nia in patients with SCI.[15] For prolonged tracheostomized
patients, dysphagia evaluation for safe decannulation is an import-
ant clinical issue. We have used swallowing tests as a decannula-
tion criterion. Dysphagic patients showed a significant
improvement in dysphagia tests and had no dysphagia (91%) after
the rehabilitation process and successful decannulation (85%).
These results indicate the significance of the multifaceted rehabili-
tation program for respiratory-compromised patients with SCI.
During the protocol, four complications occurred; two mucous
plugs, one atelectasis and one case of pneumonia. Although suc-
cessfully managed, these potentially dangerous problems merit
consideration of alternative diagnoses (pulmonary thromboembol-
ism, aspiration) and an expanded evaluation, which should be kept
in mind. We conclude that our protocol’s initial experience with
RRWT has had an acceptable complication rate. Moreover, the
occurrence of these complications may support the logic of the
implementation of a multidisciplinary, multifaceted respiratory
evaluation and management protocol, and further improvements
in this protocol will be sought that could minimize the risks to
these patients. A combination of respiratory rehabilitation interven-
tions may be indicated after cervical SCI for individuals who are at
risk for respiratory complications.
In this group of tetraplegic patients receiving home mechanical
ventilation, there is the need for a comprehensive approach to
ongoing surveillance for any reversible factors. Recently, there has
been increasing awareness of the importance of intermediate care
units, which manage SCI patients who need MV and TT. We pro-
pose that this study provides a disease-specific multifaceted
assessment and management of respiratory dysfunction.
Furthermore, during the gradual weaning process in the ward,
the intensivist provided explicit guidance and directed an individu-
alized plan as a resource regarding information about the ventila-
tor. The essential role that the physiatrist played was providing
holistic respiratory care; bringing all of the respiratory rehabilita-
tion strategies together, such as the respiratory evaluation meth-
ods, secretion removal techniques, respiratory muscle training,
weaning from MV, and decannulation of TT, for these patients.
Finally, the coordination of care between the rehabilitation team,
intensivist, and ENT specialist, on which the successful rehabilita-
tion of many of the patients depended, was established.
Several limitations need to be considered with respect to our
study. First, this study was a single-center design and involved a
limited number of patients, which may have affected our results
regarding the higher weaning and decannulation rates compared
to previous studies. Second, it evaluates the SCI patients in a
rehabilitation clinic in our country with a unique system of care
and was affected by regional differences. Thus, it was difficult to
establish comparisons with different settings. Third, our primary
aim was to evaluate the effectiveness of the multifaceted, aggres-
sive respiratory rehabilitation protocol in a rehabilitation clinic.
Before implementing RRWT, our clinic was unable to accept
patients with TT or MV, and thus, we were unable to compare our
findings with previous data. Nevertheless, we compared our data
with those obtained in previous studies. In addition, because most
of the patients in this study received nearly all of the respiratory
rehabilitation interventions, it was difficult to identify which inter-
ventions were more effective than other interventions. Fourth,
although multifaceted respiratory assessment and management
strategies were used in this study, end-tidal CO2 monitoring is not
available in our clinic, which is a sensitive and noninvasive method
for the early detection of hypercarbia during MV weaning and can
 8 I. GUNDOGDU ET AL.
diminish the need for blood gas determinations. Similarly, noninva-
sive ventilation methods, which have been established as a safe
and effective weaning strategy, were not used because they were
not available in our clinic. Finally, although we did not evaluate
our patients for quality of life (QoL) and family burden, we were
impressed that our patients and their families responded favorably
to this protocol. Future studies investigating rehabilitation strat-
egies during the liberation of MV and TT may find that assessing
patient QoL and family burden as an additional outcome param-
eter may provide useful information.
Conclusions
Our study indicates the significance of multidimensional evaluation
of the weaning potential of prolonged MV- and TT-dependent
patients and the necessity of providing the appropriate amount of
time for these patients to recover. A multifaceted, multidisciplinary
Downloaded by [Library Services City University London] at 07:50 04 July 2016
respiratory management program can change the process of care
used for difficult-to-wean patients with SCI. Because many of the
so-called difficult- or impossible-to-wean patients were successfully
weaned from MV and TT in the PMR clinic, the need for such an
outlet for countries without specialized centers is supported.
Disclosure statement
The authors report no conflicts of interest.
References
[1] Zimmer MB, Nantwi K, Goshgarian HG. Effect of spinal cord
injury on the respiratory system: basic research and current
clinical treatment options. J Spinal Cord Med.
2007;30:319–330.
[2] MacIntyre NR, Epstein SK, National Association for Medical
Direction of Respiratory Care, et al. Management of patients
requiring prolonged mechanical ventilation: report of a
NAMDRC consensus conference. Chest. 2005;128:3937–3954.
[3] Brown R, DiMarco AF, Hoit JD, et al. Respiratory dysfunction
and management in spinal cord injury. Respir Care. 2006;
51:853–868.
[4] Berlly M, Shem K. Respiratory management during the first
five days after spinal cord injury. J Spinal Cord Med.
2007;30:309–318.
[5] Berney S, Bragge P, Granger C, et al. The acute respiratory
management of cervical spinal cord injury in the first 6
weeks after injury: a systematic review. Spinal Cord.
2011;49:17–29.
[6] Onders RP, Elmo M, Khansarinia S, et al. Complete world-
wide operative experience in laparoscopic diaphragm pac-
ing: results and differences in spinal cord injured patients
and amyotrophic lateral sclerosis patients. Surg Endosc.
2009;23:1433–1440.
[7] Cameron TS, McKinstry A, Burt SK, et al. Outcomes of
patients with spinal cord injury before and after introduc-
tion of an interdisciplinary tracheostomy team. Crit Care
Resusc. 2009;11:14–19.
[8] Nakashima H, Yukawa Y, Imagama S, et al. and others.
Characterizing the need for tracheostomy placement and
decannulation after cervical spinal cord injury. Eur Spine J.
2013;22:1526–1532.
[9] Wong S, Derry F, Jamous A, et al. Validation of the spinal
nutrition screening tool (SNST) in patients with spinal cord
injuries (SCI): result from a multicentre study. Eur J Clin
Nutr. 2012;66:382–387.
[10] Kang SW, Shin JC, Park CI, et al. Relationship between
inspiratory muscle strength and cough capacity in cervical
spinal cord injured patients. Spinal Cord. 2006;44:242–248.
[11] Vitacca M, Paneroni M, Bianchi L, et al. Maximal inspiratory
and expiratory pressure measurement in tracheotomised
patients. Eur Respir J. 2006;27:343–349.
[12] Chiodo AE, Scelza W, Forchheimer M. Predictors of ventila-
tor weaning in individuals with high cervical spinal cord
injury. J Spinal Cord Med. 2008;31:72–77.
[13] Kimura J. Electrodiagnosis in Diseases of Nerve and Muscle.
2nd ed. Philadelphia (PA): FA Davis; 1983.
[14] Bolton CF, Grand’Maison F, Parkes A, et al. Needle electro-
myography of the diaphragm. Muscle Nerve. 1992;15:
678–681.
[15] Shem K, Castillo K, Wong S, et al. Dysphagia in individuals
with tetraplegia: incidence and risk factors. J Spinal Cord
Med. 2011;34:85–92.
[16] Warnecke T, Ritter MA, Kroger B, et al. Fiberoptic endo-
scopic dysphagia severity scale predicts outcome after
acute stroke. Cerebrovasc Dis. 2009;28:283–289.
[17] Cosortium for Spinal Cord Medicine. Respiratory manage-
ment following spinal cord injury: a clinical practice guide-
line for health-care professionals. J Spinal Cord Med.
2005;28:259–293.
[18] Fink JB. Positioning versus postural drainage. Respir Care.
2002;47:769–777.
[19] Whitman J, Van Beusekom R, Olson S, et al. Preliminary
evaluation of high-frequency chest compression for secre-
tion clearance in mechanically ventilated patients. Respir
Care. 1993;38:1081–1087.
[20] Jaeger RJ, Turba RM, Yarkony GM, et al. Cough in
spinal cord injured patients: comparison of three methods
to produce cough. Arch Phys Med Rehabil. 1993;74:
1358–1361.
[21] Gutierrez CJ, Harrow J, Haines F. Using an evidence-based
protocol to guide rehabilitation and weaning of ventilator-
dependent cervical spinal cord injury patients. J Rehabil Res
Dev. 2003;40:99–110.
[22] Bach JR. Respiratory muscle aids for the prevention of pul-
monary morbidity and mortality. Semin Neurol. 1995;
15:72–83.
[23] Schilero GJ, Grimm DR, Bauman WA, et al. Assessment of
airway caliber and bronchodilator responsiveness in sub-
jects with spinal cord injury. Chest. 2005;127:149–155.
[24] Martin AD, Davenport PD, Franceschi AC, et al. Use of
inspiratory muscle strength training to facilitate ventilator
weaning: a series of 10 consecutive patients. Chest. 2002;
122:192–196.
[25] deBoisblanc MW, Goldman RK, Mayberry JC, et al. Weaning
injured patients with prolonged pulmonary failure from
mechanical ventilation in a non-intensive care unit setting.
J Trauma. 2000;49:224–230.
[26] Peterson W, Charlifue W, Gerhart A, et al. Two methods of
weaning persons with quadriplegia from mechanical venti-
lators. Paraplegia. 1994;32:98–103.
[27] Rosenthal LJ, Kim V, Kim DR. Weaning from prolonged
mechanical ventilation using an antipsychotic agent in a
patient with acute stress disorder. Crit Care Med. 2007;35:
2417–2419.
[28] Bach JR, Saporito LR. Criteria for extubation and tracheos-
tomy tube removal for patients with ventilatory failure.

A different approach to weaning. Chest. 1996;110:
1566–1571.
[29] Romero-Ganuza J, Garcia-Forcada A, Vargas E, et al. An
intermediate respiratory care unit for spinal cord-injured
patients. A retrospective study. Spinal Cord. 2015;53:
552–556.
[30] Gonzalez-Bermejo J, LLontop C, Similowski T, et al.
Respiratory neuromodulation in patients with neurological
pathologies: for whom and how? Ann Phys Rehabil Med.
2015;58:238–244.
Downloaded by [Library Services City University London] at 07:50 04 July 2016
RESPIRATORY REHABILITATION FOR SCI PATIENTS 9
[31] Garara B, Wood A, Marcus HJ, et al. Intramuscular diaphrag-
matic stimulation for patients with traumatic high cervical
injuries and ventilator dependent respiratory failure: a sys-
tematic review of safety and effectiveness. Injury.
2016;47:539–544.
[32] Gundogdu I, Akyuz M, Ozturk EA, et al. Can spinal cord
injury patients show a worsening in ASIA impairment scale
classification despite actually having neurological improve-
ment? The limitation of ASIA impairment scale classification.
Spinal Cord. 2014;52:667–670.