Journal of Pediatric Rehabilitation Medicine: An Interdisciplinary Approach 4 (2011) 312 3

DOI 10.3233/PRM-2011-0148
IOS Press
Original Research
The GMFM, PEDI, and CP-QOL and
perspectives on functioning from children
with CP, parents, and medical professionals
Jilda N. Vargus-Adams
a,b,
, Lauren K. Martin
a
, Stacey H. Maignan
c
, Amy C. Klein
c
and
Shelia Salisbury
b
a
Division of Pediatric Rehabilitation, Department of Pediatrics, Cincinnati Childrens Hospital Medical Center
and Department of Physical Medicine and Rehabilitation, University of Cincinnati School of Medicine, Cincinnati,
OH, USA
b
Division of Biostatistics and Epidemiology, Department of Pediatrics, Cincinnati Childrens Hospital Medical
Center, Cincinnati, OH, USA
c
Department of Occupational Therapy and Physical Therapy, Cincinnati Childrens Hospital Medical Center,
Cincinnati, OH, USA
Accepted 2 December 2010
Abstract. Objective: Many outcome measures assess function of children with cerebral palsy (CP), but establishing meaningful
clinical change remains challenging. This study explored correlations between subjective status ratings in several functional
domains, made by children, parents, and medical professionals. The ratings were compared with three outcome measures in
preparation for longitudinal work to establish minimal clinically important change.
Method: Children were assessed with the Gross Motor Function Measure (GMFM), Pediatric Evaluation of Disability Inventory
(PEDI), and Cerebral Palsy Quality of Life Questionnaire for Children (CP-QOL). Respondents provided Likert scale and Linear
Analogue Scale ratings of gross motor function, self care, social function, quality of life, and overall function. Correlations were
calculated for outcome measure scores and ratings.
Results: 122 children with CP across all GMFCS and MACS levels, 79 male, aged 8.1 2.9 years generated status ratings by
27 child reports, 122 parent reports, and 110 medical professional reports. Most ratings were moderately to highly correlated
between parents and medical professionals. Outcome measure scores were frequently signicantly correlated with pertinent
ratings from medical professionals and parents but usually not with child ratings.
Conclusions: Parents and medical professionals have similar perceptions of gross motor, self-care, quality of life, and overall
status for children with CP and these perceptions correlate with standard outcome measures, but often do not agree with childrens
ratings. Longitudinal use of subjective status ratings from parents and professionals should contribute to establishing minimal
clinically important differences for CP outcome measures.
Keywords: Cerebral palsy, parent, outcome measures, function

Corresponding author: Jilda N. Vargus-Adams MD MSc, Di-

vision of Pediatric Rehabilitation, Cincinnati Childrens Hospital
Medical Center, 3333 Burnet Avenue, MLC 4009, Cincinnati, OH
45229, USA. Tel.: +1 513/636 7480; Fax: +1 513/636 7360; E-mail:
jilda.vargus-adams@cchmc.org.
1874-5393/11/$27.50 2011 IOS Press and the authors. All rights reserved
4 J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP
1. Introduction
The impacts of cerebral palsy (CP) are complicated.
Because CP may affect many aspects of a childs life,
we measure outcomes across many dimensions [12,36,
39]. Clinicians and researchers use various outcome
measures (questionnaires, assessments, and tests) in an
effort to understand the current status of children with
CPand also to evaluate change in status over time or fol-
lowing interventions. The Gross Motor Function Mea-
sure (GMFM) [24], the Pediatric Evaluation of Disabil-
ity Inventory (PEDI) [11], and the Cerebral Palsy Qual-
ity of Life Questionnaire for Children (CP-QOL) [38]
are three measures preferred for use in CP [33]. Each
of these measures has established validity, but incom-
plete information on minimal clinically important dif-
ferences (MCIDs) [14,19,36].
The MCID is a useful concept in delineating change
that really matters from that which may be merely sta-
tistically signicant. The MCID is understood to be
the smallest change in a score which patients perceive
as benecial (or detrimental) and which would justi-
fy a change in the management of a patients illness,
in the absence of signicant cost or risk [15]. Thus,
the MCID allows researchers and clinicians to make
sense of a change score and to understand better the
functional implications of a change score for any giv-
en individual. With established MCIDs for common
outcome measures, researchers and clinicians would
be guided in deciding what degree of change in these
scores would merit changes in therapeutic practices or
inuence their medical decision making.
MCIDs have been determined for wide ranging out-
come measures in various elds [3,18,21]. A preferred
method of calculating MCIDs involves comparing out-
come measure scores over time with patient or profes-
sional assessments serving as anchors [1,15,16,40]. As
a simplied example, in order to establish the MCIDfor
a particular questionnaire or evaluation, patients would
complete the questionnaire or evaluation at two time
points when they would also rate their overall status
or functioning (or whatever construct the questionnaire
or evaluation was intended to assess). The amount of
change would be calculated for both the questionnaire
scores and for the patient self-ratings. The patients
whose self-ratings reected a minimal, but present,
amount of change would be expected to have change
scores on the outcome measure that represent a mini-
mal, but present, amount of change which is the de-
nition of an MCID. In related methodologies, other an-
choring assessments might be employed to determine
which subgroups of patients experienced minimal im-
portant change. These ratings might be generated with
different reporters such as parents or medical profes-
sionals, rather than the patient directly, or the ratings
of change over time might be assessed by direct ques-
tioning (how much are you better or worse?), rather
than asking at two time points and calculating a change
score.
Understanding the perceptions of various informants
is key to interpreting these reports or ratings (these
terms will be used interchangeably to reect a quanti-
ed response regarding an individuals status) or out-
come measure scores. Much work on parent-child
agreement in outcome measures focuses on health-
related quality of life (QOL). Studying children with
CP, Varni et al. found that fewer than half could self-
report, but that in the cases where both parent and child
versions of the Pediatric Quality of Life Inventory Ver-
sion 4.0 were completed, correlations were only fair
to moderate and were particularly poor for emotion-
al functioning [34]. In work with the same outcome
measure, Majnemer et al., had a similar proportion of
children who could self-report, but found more moder-
ate correlations including those in emotional function-
ing [17]. The wider literature on health-related QOL
acknowledges that parent-child agreement is best for
observable and chronic illness, but poorer for emotional
concerns [4,8,30]. In children with CP[41] as well as in
other clinical samples [32], self-reported health-related
QOL is often higher than parent proxy reports. These
discrepancies may be due to different understanding
or weighting of the concepts being assessed [6,30] or
different reasoning and response styles [5].
Beyond QOL discussions, little is known about the
relationship between status reports or outcome measure
scores from children and those from parents, although
levels of agreement in settings such as limb decien-
cy, visual impairment, or physical activity are often no
better than moderate [22,28,29]. Similarly, the rela-
tionship between professional and child reports has not
been explored in depth. While parents and teachers
demonstrate some agreement in evaluating behavior [2,
31], specic variations were identied between parents
and professionals in evaluating QOL in CP [42]. The
relationships between child and other perspectives are
of particular interest in CP wherein proxy reports are
common due to the large prevalence of signicant con-
comitant intellectual disability. Whilst proxy reporting
is not preferable, it may be unavoidable if the child
cannot self-report.
The rst step toward establishing MCIDs for com-
mon CPoutcome measures is the demonstration that the
J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP 5
outcome scores and subjective ratings (that come from
patients, parents, or professionals) at a single point in
time are correlated. This process is a primary aim of
this study.
Secondarily, this study explores the relationships
amongst ratings from various reporters with an aim
of evaluating the possible similarities. Moreover, this
study aims to inform discussion about shared under-
standing of clinical or functional status between chil-
dren with CP, their parents and medical professionals as
greater agreement may enhance the identication and
management of salient problems, the selection of inter-
ventions, adherence, and, ultimately, the improvement
of outcomes.
2. Methods
2.1. Study site and population
Children with a diagnosis of CP between the ages of
4 and 12 years were recruited over 18 months from a
tertiary care childrens hospital and on-line postings on
the hospital website. The ages 412 years were select-
ed to reect target populations for the outcome mea-
sures employed. Exclusion criteria included serious
concurrent illness not typically associated with CP. In
order to have a very inclusive study population, efforts
were made to enroll a range of ages, levels and distribu-
tion of physical impairments, and degrees of intellectu-
al disability. This range of participants was intended to
provide a broad sampling of children with CP, although
this diversity of population was also anticipated to re-
sult in a signicant number of participants who would
not be able to self-report due to either intellectual dis-
ability or age. Consent was obtained fromparticipants
parents or guardians and from all children capable of
giving assent. The local Institutional Review Board
approved this study. All clinical evaluations were per-
formed at the hospitals outpatient therapy center by an
experienced pediatric physical therapist or occupation-
al therapist with training and competence in the use of
all the measures. Questionnaires were completed at the
hospital or at home.
2.2. Outcome measures and status ratings
Participants were assessed using: 1) the Gross Mo-
tor Function Measure (GMFM), 2) the Pediatric Eval-
uation of Disability Inventory (PEDI), and 3) the Cere-
bral Palsy Quality of Life Questionnaire for Children
(CP-QOL). These outcome measures were chosen as
well-regarded and widely employed measures in CP
research [33]. The measures address the primary do-
mains of gross motor function, self-care, social func-
tion, and QOL. Additionally, the assessment asked each
participant, their parent, and a medical professional to
rate the participants status in the four domains and
overall.
The GMFM is a robust, validated, and widely-
employed instrument [26] designed to measure gross
motor function in children. It is valid in detecting
change in gross motor function in children with cere-
bral palsy and is intended for use in clinical trials in-
volving such children [27]. Used in both clinical and
research settings, the GMFM-66 [25] involves a trained
pediatric therapists observation of a childs actual mo-
tor activity, scoring from 0 to 100, with higher scores
reecting better motor function. Rigorous evaluation
for MCIDs have not been pursued for the GMFM-66,
although preliminary data were published for the prior,
non-interval scaled GMFM-88 version [26].
The PEDI [11] assesses functional capabilities and
performance in children with disabilities. It evaluates
self-care, mobility, and social function and is valid and
reliable. To complete the PEDI, a trained clinician or
investigator conducts a parent interview. The PEDI
was designed for children less than 8 years of age,
but is appropriate for older children with functional
limitations. Subscale scores are calculated with higher
scores reecting better function.
The CP-QOL [38] is a condition-specic health-
related QOL instrument for children with CP, with es-
tablished reliability and validity [37]. The question-
naire is available in two versions, parent-report (used
for all participants, 7 subscales and 66 items) and child
self-report (used with participants aged 912 years with
the cognitive capacity to respond, 5 subscales and 53
items) and these versions show moderate correlation in
psychometric testing [37]. Scores range from 0100
with higher scores reecting higher QOL.
Subjective status ratings were solicited to describe
the participants current status for overall functioning
and for the domains of gross motor, social function-
ing, self-care, and QOL. The domains were selected
to reect the primary domains of the three outcome
measures being employed. These ratings were devel-
oped for this study, employ quantied response options
and closely resemble status ratings used in other work
to quantify clinical change. Two types of scales were
used, a 100 mm linear analogue scale (LAS) and an
11-point Likert scale, providing two ratings for each
6 J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP
question and permitting comparison between the scale
types. Each participant with CP (if cognitively able
to compare values for status ratings, as decided by the
child and parent and veried by study staff), one of
their parents or guardians, and a medical professional
provided status ratings. The medical provider, identi-
ed by the subject and family, was a treating therapist,
doctor, or nurse and had frequent and recent contact
with the participant. Each rater (participant, parent,
medical professional) was informed of the aims of the
study and responded to two types of questions describ-
ing the participants current health status. The rst rat-
ing involved responding to a stem sentence of In the
last two weeks, how are you/your child/your patient
doing OVERALL? by placing a mark on a 100 mm
LAS anchored with extremely bad and a sad-face at
the lower endpoint and excellent and a happy-face at
the upper endpoint. The second rating used the same
stem sentences with 11 response options of numbers
from 0 to 10, labeled as extremely bad at 0 and ex-
cellent at 10 on Likert scale. Each of these ques-
tions was asked in terms of OVERALL functioning (as
stated above) and then again for the domains that are
reected by the GMFM, PEDI, and CP-QOL (replac-
ing the word OVERALL with WITH GROSS MOTOR
AND MOBILITY SKILLS SUCH AS WALKING?,
WITH SELF-CARE?, WITH SOCIAL FUNCTION-
ING?, and WITH QUALITY OF LIFE?). The status
rating questions were piloted with several parents and
professionals prior to the study and found to be easy to
understand and complete.
The Gross Motor Function Classication System
(GMFCS) [20] is a convenient and reliable means of
stratifying children with CP by severity [23,43] and is
stable over time. An experienced pediatric occupation-
al or physical therapist classied each participant with
the GMFCS, separating the study population into 5 stra-
ta of severity. Similarly, each participant was classied
in terms of the Manual Abilities Classication System
(MACS), an upper extremity analogue of the GMFCS,
creating 5 strata of severity [9].
2.3. Statistical analysis
Measures were analyzed with standard scoring algo-
rithms. Scaled scores for the Functional Skills scales
of the PEDI were used. Means were calculated from
the subscales of the CP-QOL (5 subscales for the child
version and 7 for the parent version) to create a sin-
gle value. Descriptive statistics were generated for de-
mographic information. Associations among variables
were evaluated using schematic plots and correlations.
All correlations were generated using Spearmans rank
correlation. Fishers transformation was used to cal-
culate 95% condence intervals for each correlation.
A condence interval that included zero indicated that
the correlation was not signicantly different fromzero
(i.e. there is no association between the two variables).
The Spearman correlations for age and CP severity
variables (GMFCS and MACS) were calculated for the
scores for GMFM, PEDI, and CP-QOL. Status rating
correlations were also examined, both amongst the var-
ious raters and between LAS and Likert ratings. Last-
ly, correlations were calculated between status ratings
and scores on the GMFM, PEDI, and CP-QOL. All
analyses were performed using SAS 9.2 (Cary, NC).
3. Results
Demographic information is in Table 1.
Status ratings for the 122 children with CP were pro-
vided by 27 child reports, 122 parent reports, and 110
medical professional reports. Because 3 participants
each had 2 medical professionals report for them and
20 professionals reported for more than 1 participant,
a total of 85 professionals were involved including 60
outpatient therapists, 20 school therapists, 3 doctors
and 2 nurses. Some status ratings were not provid-
ed, especially from medical professionals, resulting in
missing data. When two professional reports were sub-
mitted, the domains were separated so that only one
professional reported for each domain (for example,
the physical therapist completed Gross Motor and the
occupational therapist completed Self Care). Only 27
children were able to self-report, due to age or intellec-
tual disability, so data from the child respondent group
is limited which affects the robustness of the analysis.
LAS status ratings were highly correlated with Lik-
ert scale status ratings for each respondent group with
Spearman correlation coefcients () 0.740.87 for
children, 0.920.96for parents, and 0.920.98for med-
ical professionals, all with p value < 0.001. These
correlations demonstrate a reliability of responses from
participants between these two methods of assessing
status reports. When status ratings were compared be-
tween respondent groups, signicant correlations were
observed between parents and medical professionals in
gross motor ( = 0.67, 0.73 for LAS and Likert, re-
spectively), self care ( = 0.83, 0.82), QOL ( = 0.26,
0.24), and overall ( = 0.38, 0.36) status ratings (Ta-
ble 2). Ratings of social function as well as most ratings
J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP 7
Table 1
Demographics
Total population Population who could self report
Total population 122 children 27 children
Age mean (SD) 8 years 1 month (2y10m) 10 years 9 months (1y6m)
Sex 79 Male (65%) 14 Male (52%)
Race Asian 6 (5) 0
n (%) Black 13 (11) 1 (4)
White 99 (81) 24 (89)
Other 4 (3) 2 (7)
GMFCS I 37 (30) 15 (56)
n (%) II 20 (16) 3 (11)
III 18 (15) 5 (19)
IV 22 (18) 3 (11)
V 25 (20) 1 (4)
MACS I 46 (38) 16 (59)
n (%) II 30 (25) 9 (33)
III 12 (10) 1 (4)
IV 11 (9) 1 (4)
V 23 (19) 0
CP Distribution Quadriparesis 52 (43) 7 (26)
n (%) Diparesis 29 (24) 7 (26)
Hemiparesis 36 (30) 12 (44)
Triparesis 5 (4) 1 (4)
SD standard deviation, GMFCS Gross Motor Function Classication System,
MACS Manual Abilities Classication System, CP cerebral palsy.
Table 2
Correlations between Linear Analogue Scale (LAS) and Likert status ratings of different respondent groups. Correlation
coefcients (95% condence intervals) and pvalues
Domain Child-parent Child-medical professional Parent-medical professional
n = 27 n = 2124 n = 96106
LAS Ratings (95%CI) p value (95%CI) p value (95%CI) p value
Gross Motor 0.38 (0.01, 0.66) 0.049 0.36 (0.07, 0.67) 0.088 0.67 (0.55, 0.77) < 0.0001
Self Care 0.35 (0.04, 0.64) 0.073 0.10 (0.51, 0.35) 0.664 0.83 (0.76, 0.88) < 0.0001
Social Function 0.23 (0.17, 0.56) 0.243 0.13 (0.30, 0.51) 0.553 0.12 (0.08, 0.30) 0.254
Quality of Life 0.35 (0.04, 0.64) 0.075 0.02 (0.40, 0.43) 0.929 0.26 (0.07, 0.44) 0.009
Overall 0.08 (0.31, 0.44) 0.705 0.06 (0.35, 0.45) 0.786 0.38 (0.19, 0.53) < 0.0001
Likert Status Ratings
Gross Motor 0.47 (0.09, 0.71) 0.015 0.42 (0.00, 0.71) 0.044 0.73 (0.62, 0.80) < 0.0001
Self Care 0.39 (0.00, 0.67) 0.046 0.10 (0.51, 0.35) 0.671 0.82 (0.74, 0.87) <0.0001
Social Function 0.37 (0.13, 0.66) 0.054 0.14 (0.29, 0.52) 0.514 0.16 (0.03, 0.34) 0.114
Quality of Life 0.24 (0.15, 0.57) 0.217 0.17 (0.16, 0.54) 0.435 0.24 (0.05, 0.42) 0.014
Overall 0.20 (0.19, 0.54) 0.311 0.03 (0.43, 0.38) 0.896 0.36 (0.18, 0.51) 0.0002
LAS Linear Analogue Scale.
between children and other raters were not correlated
and the correlations rarely reached statistical signi-
cance. Child ratings of gross motor status demonstrat-
ed signicant correlations with parent ratings ( =0.47
Likert) but not with most medical professional ratings.
Outcome measure scores were available for 122
children on the GMFM (mean 52.9, standard devi-
ation (SD) 24.2), 122 children on the PEDI (Self
Care mean 54.2, SD 18.7, Social Function mean
57.2, SD 14.5, Mobility mean 56.1, SD 24.8),
and 27 children and 120 parents on the CP-QOL
(Child mean 72.0, SD 9.7, Parent mean 64.8, SD
8.4). Outcome measure scores (with the exception
of the child form of the CP-QOL) showed strong
relationships with severity of impairment (GMFCS
and MACS) (Table 3). Age was not related to the
outcome measure scores. Pertinent status ratings were
evaluated with the outcome measures in Table 4. Sig-
nicant correlations were evident between the outcome
measures and status ratings for the parents in all cas-
es except the Child CP-QOL and overall status. For
the medical professionals, correlations existed between
status ratings and the GMFM and PEDI, but not with
the CP-QOL. The strongest relationships for the par-
8 J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP
Table 3
Outcome Measure Mean Scores (standard deviations) and Spearman Correlation Coefcients (95% condence intervals) and pvalues with
demographic information
Mean score (SD) Age p GMFCS p MACS p
GMFM n = 122 52.9 (24.2) 0.10 (0.27, 0.08) 0.263 0.94 (0.96, 0.92) < 0.0001 0.81 (0.86, 0.73) < 0.0001
PEDI Self Care 54.2 (18.7) 0.05 (0.13, 0.23) 0.578 0.84 (0.88, 0.78) < 0.0001 0.82 (0.87, 0.75) < 0.0001
n = 122
PEDI Mobility 56.1 (24.8) 0.02 (0.20, 0.16) 0.841 0.92 (0.95, 0.89) < 0.0001 0.80 (0.85, 0.72) < 0.0001
n = 122
PEDI Social Function 57.2 (14.5) 0.14 (0.04, 0.31) 0.137 0.68 (0.76, 0.57) < 0.0001 0.70 (0.78, 0.59) < 0.0001
n = 122
CP-QOL Child 72.0 (9.7) 0.30 (0.61, 0.09) 0.123 0.03 (0.40, 0.36) 0.894 0.05 (0.34, 0.42) 0.813
n = 27
CP-QOL Parent 64.8 (8.4) 0.03 (0.20, 0.15) 0.774 0.37 (0.51, 0.20) < 0.0001 0.29 (0.44, 0.11) 0.001
n = 120
GMFCS Gross Motor Function Classication System, MACS Manual Abilities Classication System, GMFM Gross Motor Function Measure,
PEDI Pediatric Evaluation of Disability Inventory, CP-QOL Cerebral Palsy Quality of Life Questionnaire for Children.
ents and the medical professionals were between the
GMFM and PEDI Mobility measures and gross motor
status as well as PEDI Self Care and self care status.
For the children, the few relationships demonstrating
signicant correlations included PEDI Social Function
and social function status. None of the correlation co-
efcients for the children were large.
4. Discussion
This study demonstrates differences between the
self-report of children with CP and the reports of par-
ents and professionals that may be related to differ-
ences in the perspectives of children versus the adult
reporters. Furthermore, this study establishes variable
degrees of correlation of those reports with established
outcome measures. These ndings suggest that there is
no single best measure of CP outcomes; perceptions of
children, arguably the most valid assessors of their own
status, are not the same as the perceptions of their par-
ents or providers nor are they strongly related to highly
regarded CP measurement tools. Conversely, parents
and providers demonstrated reasonable agreement with
each other in some arenas as well as with CP outcome
measures.
Large, signicant correlations between status ratings
were present only between medical professionals and
parents, and only in the gross motor and self-care do-
mains. Moderate correlations were found in overall
ratings (parents and professionals) and in gross motor
(child and both raters) and self care (child and par-
ent). All other correlations were small and/or non-
signicant. As gross motor status is the most obvi-
ous domain, because all parties could directly observe
the childs mobility, higher correlations were expect-
ed. Other domains require greater information to as-
sess, but parents and professionals agreed on self-care.
Therapists were likely to be involved in self-care tasks
with the children, and parents also see these issues dai-
ly, so parents and many of the professionals had a good
understanding of self-care status. The smaller social
function and QOL correlations may reect greater vari-
ability in the respondents individual denitions of the
domains, their knowledge of the childs experience, or
variation in how respondents dene these concepts [6].
Limited correlation in QOL is in line with prior studies
that have shown that children with chronic disease rate
their own QOL higher than others do [13].
Parental and child reports were not well correlated
in general; correlations between professionals and chil-
dren were even smaller. Arguably, children may have
perspectives that vary considerably from the assess-
ments of adults, even the adults who know the children
well. Children, parents and professionals may differ
substantially in how they interpret the various domains
as well, due to differing frames of reference or experi-
ences. We did not ask respondents to dene or clarify
their understanding of each of the domains, so the pre-
cise constructs may have varied between groups or be-
tween individual responses. For example, QOL may be
regarded by some as being highly inuenced by health
status, such that the state of having a physical disability
would directly reduce a persons QOL, whereas others
might view QOL as a more internal state of happiness
or well-being, that could be relatively unaffected by
disability. The CP-QOL was developed to reect the
several domains of QOL that were deemed applicable
by children with CP and their families. Functioning
and physical health were among these domains. If the
constructs addressed by the CP-QOL and those per-
ceived in the status questions were not similar, corre-
J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP 9
T
a
b
l
e
4
S
p
e
a
r
m
a
n
C
o
r
r
e
l
a
t
i
o
n
C
o
e
f

c
i
e
n
t
s
b
e
t
w
e
e
n
s
t
a
t
u
s
r
a
t
i
n
g
s
a
n
d
o
b
j
e
c
t
i
v
e
m
e
a
s
u
r
e
s
w
i
t
h
9
5
%
c
o
n

d
e
n
c
e
i
n
t
e
r
v
a
l
s
a
n
d
p

v
a
l
u
e
s
P
a
r
e
n
t
s
(
n
=
1
1
5

1
2
0
)
M
e
d
i
c
a
l
p
r
o
f
e
s
s
i
o
n
a
l
s
(
n
=
1
0
0

1
0
8
)
Y
o
u
t
h
(
n
=
2
7
)
S
t
a
t
u
s
R
a
t
i
n
g
O
b
j
e
c
t
i
v
e
M
e
a
s
u
r
e
L
i
k
e
r
t
r
a
t
i
n
g
L
i
n
e
a
r
A
n
a
l
o
g
u
e
L
i
k
e
r
t
r
a
t
i
n
g
L
i
n
e
a
r
A
n
a
l
o
g
u
e
L
i
k
e
r
t
r
a
t
i
n
g
L
i
n
e
a
r
A
n
a
l
o
g
u
e
S
c
a
l
e
r
a
t
i
n
g
S
c
a
l
e
r
a
t
i
n
g
S
c
a
l
e
r
a
t
i
n
g

(
9
5
%
C
I
)

(
9
5
%
C
I
)

(
9
5
%
C
I
)

(
9
5
%
C
I
)

(
9
5
%
C
I
)

(
9
5
%
C
I
)
p
-
v
a
l
u
e
p
-
v
a
l
u
e
p
-
v
a
l
u
e
p
-
v
a
l
u
e
p
-
v
a
l
u
e
p
-
v
a
l
u
e
G
r
o
s
s
M
o
t
o
r
G
M
F
M
0
.
7
4
(
0
.
6
5
,
0
.
8
1
)
0
.
7
0
(
0
.
6
0
,
0
.
7
8
)
0
.
7
9
(
0
.
7
1
,
0
.
8
5
)
0
.
7
8
(
0
.
6
9
,
0
.
8
4
)
0
.
1
9
(

0
.
2
1
,
0
.
5
3
)
0
.
1
5
(

0
.
2
5
,
0
.
5
0
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
3
3
6
p
=
0
.
4
5
4
G
r
o
s
s
M
o
t
o
r
P
E
D
I
M
o
b
i
l
i
t
y
0
.
0
7
2
(
0
.
6
1
,
0
.
7
9
)
0
.
6
8
(
0
.
5
7
,
0
.
7
7
)
0
.
7
9
(
0
.
7
1
,
0
.
8
5
)
0
.
7
9
(
0
.
7
0
,
0
.
8
5
)
0
.
3
0
(

0
.
1
0
,
0
.
6
1
)
0
.
2
8
(

0
.
1
2
,
0
.
6
0
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
1
2
8
p
=
0
.
1
6
0
S
e
l
f
C
a
r
e
P
E
D
I
S
e
l
f
C
a
r
e
0
.
8
4
(
0
.
7
8
,
0
.
8
9
)
0
.
8
6
(
0
.
8
0
,
0
.
9
0
)
0
.
8
0
(
0
.
7
1
,
0
.
8
6
)
0
.
7
9
(
0
.
7
0
,
0
.
8
5
)
0
.
3
1
(

0
.
0
8
,
0
.
6
2
)
0
.
3
9
(
0
.
0
0
,
0
.
6
7
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
1
0
9
p
=
0
.
0
4
5
S
o
c
i
a
l
F
u
n
c
t
i
o
n
P
E
D
I
S
o
c
i
a
l
F
u
n
c
t
i
o
n
0
.
2
7
(
0
.
0
9
,
0
.
4
3
)
0
.
2
5
(
0
.
0
7
,
0
.
4
2
)
0
.
5
9
(
0
.
4
4
,
0
.
7
0
)
0
.
5
5
(
0
.
3
9
,
0
.
6
7
)
0
.
5
0
(
0
.
1
4
,
0
.
7
3
)
0
.
3
9
(
0
.
0
1
,
0
.
6
7
)
p
=
0
.
0
0
3
p
=
0
.
0
0
6
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
0
0
8
p
=
0
.
0
4
2
Q
u
a
l
i
t
y
o
f
L
i
f
e
C
P
-
Q
O
L
P
a
r
e
n
t
0
.
4
1
(
0
.
2
5
,
0
.
5
5
)
0
.
3
6
(
0
.
1
8
,
0
.
5
1
)
0
.
1
8
(

0
.
0
2
,
0
.
3
6
)
0
.
1
6
(

0
.
0
4
,
0
.
3
4
)
0
.
2
0
(

0
.
2
0
,
0
.
5
3
)
0
.
2
9
(

0
.
1
1
,
0
.
6
0
)
p
<
0
.
0
0
0
1
p
=
0
.
0
0
0
1
p
=
0
.
0
7
5
p
=
0
.
1
0
7
p
=
0
.
3
3
0
p
=
0
.
1
4
8
Q
u
a
l
i
t
y
o
f
L
i
f
e
C
P
-
Q
O
L
C
h
i
l
d
0
.
4
3
(
0
.
0
5
,
0
.
6
9
)
0
.
4
7
(
0
.
1
0
,
0
.
7
2
)
0
.
2
9
(

0
.
1
4
,
0
.
6
2
)
0
.
3
8
(

0
.
0
5
,
0
.
6
8
)
0
.
3
9
(
0
.
0
0
,
0
.
6
6
)
0
.
2
5
(

0
.
1
5
,
0
.
5
7
)
p
=
0
.
0
2
5
p
=
0
.
0
1
3
p
=
0
.
1
7
6
p
=
0
.
0
7
6
p
=
0
.
0
4
6
p
=
0
.
2
1
0
O
v
e
r
a
l
l
G
M
F
M
0
.
4
0
(
0
.
2
4
,
0
.
5
4
)
0
.
3
7
(
0
.
2
0
,
0
.
5
2
)
0
.
5
6
(
0
.
4
2
,
0
.
6
8
)
0
.
5
2
(
0
.
3
7

0
.
6
5
)

0
.
0
2
(

0
.
4
0
,
0
.
3
6
)

0
.
0
2
(

0
.
3
9
,
0
.
3
7
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
9
0
7
p
=
0
.
9
3
1
O
v
e
r
a
l
l
P
E
D
I
S
e
l
f
C
a
r
e
0
.
4
5
(
0
.
2
8
,
0
.
5
8
)
0
.
4
3
(
0
.
2
7
,
0
.
5
7
)
0
.
5
0
(
0
.
3
4
,
0
.
6
3
)
0
.
4
7
(
0
.
3
0
,
0
.
6
0
)
0
.
1
4
(

0
.
2
5
,
0
.
4
9
)
0
.
1
1
(

0
.
2
8
,
0
.
4
7
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
4
7
7
p
=
0
.
5
9
0
O
v
e
r
a
l
l
P
E
D
I
S
o
c
i
a
l
F
u
n
c
t
i
o
n
0
.
4
5
(
0
.
2
9
,
0
.
5
8
)
0
.
4
3
(
0
.
2
7
,
0
.
5
7
)
0
.
5
4
(
0
.
3
9
,
0
.
6
6
)
0
.
5
2
(
0
.
3
6
,
0
.
6
4
)
0
.
3
2
(

0
.
0
7

0
.
6
2
)
0
.
1
9
(

0
.
2
1

0
.
5
3
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
0
9
9
p
=
0
.
3
4
3
O
v
e
r
a
l
l
P
E
D
I
M
o
b
i
l
i
t
y
0
.
4
0
(
0
.
2
4
,
0
.
5
4
)
0
.
4
0
(
0
.
2
3
,
0
.
5
4
)
0
.
5
2
(
0
.
3
6
,
0
.
6
4
)
0
.
4
8
(
0
.
3
2
,
0
.
6
2
)
0
.
0
7
(

0
.
3
2
,
0
.
4
4
)
0
.
0
3
(

0
.
3
5
,
0
.
4
0
)
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
<
0
.
0
0
0
1
p
=
0
.
7
3
6
p
=
0
.
8
7
6
O
v
e
r
a
l
l
C
P
-
Q
O
L
P
a
r
e
n
t
0
.
3
4
(
0
.
1
7
,
0
.
4
9
)
0
.
3
4
(
0
.
1
6
,
0
.
4
9
)
0
.
3
0
(
0
.
1
1
,
0
.
4
6
)
0
.
3
2
(
0
.
1
4
,
0
.
4
9
)
0
.
1
4
(

0
.
2
6
,
0
.
4
9
)
0
.
2
0
(

0
.
2
0
,
0
.
5
3
)
p
<
0
.
0
0
1
p
<
0
.
0
0
1
p
=
0
.
0
0
2
p
<
0
.
0
0
1
p
=
0
.
4
8
6
p
=
0
.
3
2
2
O
v
e
r
a
l
l
C
P
-
Q
O
L
C
h
i
l
d
0
.
3
7
(

0
.
0
1
,
0
.
6
6
)
0
.
3
6
(

0
.
0
3
,
0
.
6
5
)
0
.
4
6
(
0
.
0
6
,
0
.
7
2
)
0
.
3
6
(

0
.
0
6
,
0
.
6
6
)
0
.
3
5
(

0
.
0
4
,
0
.
6
4
)
0
.
2
3
(

0
.
1
7
,
0
.
5
6
)
p
=
0
.
0
5
5
p
=
0
.
0
6
4
p
=
0
.
0
2
p
=
0
.
0
8
5
p
=
0
.
0
6
9
p
=
0
.
2
4
7
G
M
F
M
G
r
o
s
s
M
o
t
o
r
F
u
n
c
t
i
o
n
M
e
a
s
u
r
e
,
P
E
D
I
P
e
d
i
a
t
r
i
c
E
v
a
l
u
a
t
i
o
n
o
f
D
i
s
a
b
i
l
i
t
y
I
n
v
e
n
t
o
r
y
,
C
P
-
Q
O
L
C
e
r
e
b
r
a
l
P
a
l
s
y
Q
u
a
l
i
t
y
o
f
L
i
f
e
Q
u
e
s
t
i
o
n
n
a
i
r
e
f
o
r
C
h
i
l
d
r
e
n
.
10 J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP
lations would not be expected to be large. Alternative-
ly, one (or more than one) group of respondents may
have inaccurate ideas about the childs functional sta-
tus, a situation that might result from limited exposure
to the child in informative settings or variable perfor-
mance by the child. This explanation could be sup-
ported if one groups ratings are more highly correlated
with objective outcome measures like the GMFMor the
PEDI. It may be harder to resolve for concepts like QOL
or overall status where outcome measures are obtained
through proxy reporting or are not well developed.
Correlations between Likert scale ratings and LAS
ratings were quite high and neither rating type demon-
strated consistently better correlation with the mea-
sures, suggesting that only one or the other rating is
needed in similar settings.
The outcome measure scores reected the wide-
ranging functional performance of participants. Rela-
tionships between gross motor impairment (GMFCS)
and manual skills (MACS) and the outcome measures
were large and expected, consistent with prior re-
ports [10,23]. These classication variables should be
highly correlated with mobility and self care as motor
skills (directly related to the GMFCS and MACS) are
intrinsic to mobility and self care tasks. The strong
relationship of GMFCS with social function has been
reported [35] but the relationship may be confounded
with the interplay of intellectual disability or behav-
ior issues and more severe CP, as these comorbidities
may impair interpersonal interactions. Correlations be-
tween QOL and GMFCS and MACS were smaller, but
present with parent scores on the CP-QOL. This may be
the result of a true decrement of QOL in children with
more severe CP, although work in Europe has not sup-
ported a signicant effect of gross motor impairment
on QOL for young people with CP [7]. Children may
have employed a less nuanced understanding of QOL
using more extreme scores or considering fewer con-
tributing ideas, as has been reported [5]. Alternatively,
parental assessments of QOL, which have been noted
to vary from those of their children, may be inaccurate
or overly inuenced by physical disability. Although
the CP-QOL was developed as true QOL measure, it
is possible that parents interpreted QOL to be more of
health-related QOL domain.
Parent and professional perceptions of functional sta-
tus of children with CP are at least moderately accurate.
Their ratings correlated well with established outcome
measures, suggesting that they reect similar concepts
and providing support for using ratings fromparents or
professionals as tool for establishing MCIDs for these
measures. Relationships are most robust for gross mo-
tor and self care domains, which, again, are readily
assessed by observation. More moderate correlations
were present between QOL and overall ratings provid-
ed by parents and professionals and the outcome mea-
sures. These domains may have varying constructs as
QOL has no clear and widely accepted denition and
overall status is also likely to be conceptualized in var-
ious manners. Thus, smaller correlations between per-
ceptions and outcome measure scores could be the re-
sult if outcome measures themselves assess somewhat
different constructs than those the parents and medi-
cal professionals perceive. Alternatively, because it is
harder to directly witness QOL or overall status due to
the breadth of these concepts as well as the degree to
which they may be dened by an individuals internal
state, the reporters may not have adequate understand-
ing of the childrens status.
Childrens perceptions, while potentially meaning-
ful, were not signicantly correlated with the outcome
measures outside of moderate correlations between
child report and social function and QOL. Some of
the other relationships showed small or moderate cor-
relations that did not reach signicance, perhaps due
to the relatively small sample of children who could
self-report. Overall, however, outcome measure corre-
lations were of smaller magnitude with children than
with parents and professionals. We suspect that the
children may have had some difculty fully consider-
ing the meaning of the various concepts addressed, may
have had somewhat inaccurate interpretation of them,
or may have simply been inuenced by recent or more
salient events that altered their responses. Therefore,
we suspect that the childrens perceptions were not as
objective as the other reporters.
The limitations of the study include the small number
of children who could self-report their status ratings
due to age and/or cognitive impairment, which limits
both the statistical power and the generalizability of the
child reports. Nonetheless, data was collected from all
children who could provide it. Data was not complete,
however, from all medical professionals as some of the
professionals did not return status ratings, limiting the
numbers of participants whose data could be compared
with those ratings. A range of professionals served
as respondents. It is possible professional discipline,
experience, or other factors affected ratings or created
undesired variability in the data, but these factors could
not be analyzed.
Parent and professional perceptions of the functional
status of children with CP are largely well correlated
J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP 11
with each other and with scores on the GMFM and
PEDI. Parental perceptions of QOL were correlated
with the CP-QOL. Child reports of status were not cor-
related with other raters or with scores on the measures,
potentially due to the small number of respondents.
Parent and professional status ratings demonstrate ad-
equate correlation with established CP outcome mea-
sures to pursue their use as a rubric for determining
minimal clinically important differences in a longitu-
dinal study. Child reports similar to those used in this
study may not be as valuable in MCID endeavors, but
may be of interest for other reasons.
Acknowledgement/Support
This work was supported by National Institutes of
Health grant K23-HD049552 and the United Cerebral
Palsy Research and Education Foundation (now Cere-
bral Palsy International Research Foundation) grant
EH-008-06. The authors report no conicts of interest.
References
[1] D.E. Beaton et al., A taxonomy for responsiveness, J Clin
Epidemiol 54 (2001), 12041217.
[2] J. Biederman et al., Comparison of parent and teacher reports
of attention-decit/hyperactivity disorder symptoms from two
placebo-controlled studies of atomoxetine in children, Biolog-
ical Psychiatry 60 (2006), 11061110.
[3] H.I. Brunner et al., Minimal clinically important differences of
the childhood health assessment questionnaire, J Rheumatol
32 (2005), 150161.
[4] J. Cremeens, C. Eiser and M. Blades, Factors inuencing
agreement between child self-report and parent proxy-reports
onthe Pediatric Qualityof Life Inventory4.0(PedsQL) generic
core scales, Health Qual Life Outcomes 4 (2006), 58.
[5] E. Davis et al., Parent-proxy and child self-reported health-
related quality of life: using qualitative methods to explain the
discordance, Quality of Life Research 16 (2007), 863871.
[6] E. Davis et al,, Quality of life of adolescents with cerebral
palsy: perspectives of adolescents and parents, Developmental
Medicine and Child Neurology 51 (2009), 193199.
[7] H.O. Dickinson et al., Self-reported quality of life of 8-12-
year-old children with cerebral palsy: a cross-sectional Euro-
pean study, Lancet 369 (2007), 21712178.
[8] C. Eiser and R. Morse, Can parents rate their childs health-
related quality of life? Results of a systematic review, Quality
of Life Research 10 (2001), 347357.
[9] A.C. Eliasson et al., The Manual Ability Classication System
(MACS) for children with cerebral palsy: scale development
and evidence of validity and reliability, Dev Med Child Neurol
48 (2006), 549554.
[10] M.K. Gunel et al., Relationship among the Manual Abili-
ty Classication System (MACS), the Gross Motor Function
Classication System (GMFCS), and the functional status
(WeeFIM) in children with spastic cerebral palsy, European
Journal of Pediatrics 168 (2009), 477485.
[11] S.M. Haley et al., Pediatric Evaluation of Disability Inventory
(PEDI), ed, Psychological Corp, 1998.
[12] A. Harvey et al., A systematic review of measures of activi-
ty limitation for children with cerebral palsy, Developmental
Medicine and Child Neurology 50 (2008), 190198.
[13] L.M. Ingerski et al., Health-related quality of life across pedi-
atric chronic conditions, Journal of Pediatrics 156, 639644.
[14] L.V. Iyer et al., Establishing minimal clinically important dif-
ferences for scores on the pediatric evaluation of disability
inventory for inpatient rehabilitation, Physical Therapy 83
(2003), 888898.
[15] R. Jaeschke, J. Singer, and G.H. Guyatt, Measurement of
health status. Ascertaining the minimal clinically important
difference, Control Clin Trials 10 (1989), 407415.
[16] E.F. Juniper et al., Determining a minimal important change
in a disease-specic Quality of Life Questionnaire, Journal of
Clinical Epidemiology 47 (1994), 8187.
[17] A. Majnemer et al., Reliability in the ratings of quality of life
between parents and their children of school age with cerebral
palsy, Quality of Life Research 17 (2008), 11631171.
[18] K.L. Miller et al., Minimal clinically important difference for
the ocular surface disease index, Archives of Ophthalmology
128 (2010), 94101.
[19] D. Oefnger et al., Outcome tools usedfor ambulatorychildren
with cerebral palsy: responsiveness and minimum clinical-
ly important differences, Developmental Medicine and Child
Neurology 50 (2008), 918925.
[20] R. Palisano et al., Development and reliability of a system to
classify gross motor function in children with cerebral pal-
sy, Developmental Medicine and Child Neurology 39 (1997),
214223.
[21] A.L. Quittner et al., Determination of the minimal clin-
ically important difference scores for the Cystic Fibrosis
Questionnaire-Revised respiratory symptom scale in two pop-
ulations of patients with cystic brosis and chronic Pseu-
domonas aeruginosa airway infection, Chest 135 (2009),
16101618.
[22] M.J. Rah et al., Levels of agreement between parents and
childrens reports of near work, Ophthalmic Epidemiology 9
(2002), 191203.
[23] P. Rosenbaum et al., Prognosis for gross motor function in
cerebral palsy: creation of motor development curves, Jama
288 (2002), 13571363.
[24] D. Russell et al., The Gross Motor Function Measure (GMFM-
66 and GMFM-88) Users Manual Clinics in Developmental
Medicine, ed., Vol. No. 159, MacKeith Press, London, 2002.
[25] D.J. Russell et al., Improved scaling of the gross motor func-
tion measure for children with cerebral palsy: evidence of
reliability and validity, Physical Therapy 80 (2000), 873885.
[26] D.J. Russell et al., Gross Motor Function Measure (GMFM-66
and GMFM-88) Users Manual, in Clinics in Developmental
Medicine, MacKeith Press, London, 2002.
[27] D.J. Russell et al., The gross motor function measure: a means
to evaluate the effects of physical therapy, Developmental
Medicine and Child Neurology 31 (1989), 341352.
[28] L.C. Shefer et al., Comparison of self-reports and par-
ent proxy-reports of function and quality of life of children
with below-the-elbow deciency, Journal of Bone and Joint
Surgery 91 (2009), 28522859.
[29] F. Sithole and P.J. Veugelers, Parent and child reports of chil-
drens activity, Health Reports 19 (2008), 1924.
[30] H. Sweeting and P. West, Health at age 11: reports from
schoolchildren and their parents, Archives of Disease in Child-
hood 78 (1998), 427434.
12 J.N. Vargus-Adams et al. / GMFM, PEDI, and CP-QOL and perspectives on functioning in CP
[31] M.J. Tasse and L. Lecavalier, Comparing parent and teacher
ratings of social competence andproblembehaviors, American
Journal of Mental Retardation 105 (2000), 252259.
[32] P. Upton, J. Lawford and C. Eiser, Parent-child agreement
across child health-related quality of life instruments: a review
of the literature, Quality of Life Research 17 (2008), 895913.
[33] J.N. Vargus-Adams and L.K. Martin, Measuring what matters
in cerebral palsy: a breadth of important domains and outcome
measures, Archives of Physical Medicine and Rehabilitation
90 (2009), 20892095.
[34] J.W. Varni et al., Health-related quality of life of children
and adolescents with cerebral palsy: hearing the voices of the
children, Developmental Medicine and Child Neurology 47
(2005), 592597.
[35] J.M. Voorman et al., Social functioning and communication
in children with cerebral palsy: association with disease char-
acteristics and personal and environmental factors, Develop-
mental Medicine and Child Neurology (2009).
[36] D.C. Vos-Vromans, M. Ketelaar, and J.W. Gorter, Responsive-
ness of evaluative measures for children with cerebral palsy:
the Gross Motor Function Measure and the Pediatric Evalua-
tion of Disability Inventory, Disability and Rehabilitation 27
(2005), 12451252.
[37] E. Waters et al., Psychometric properties of the quality of life
questionnaire for children with CP, Developmental Medicine
and Child Neurology 49 (2007), 4955.
[38] E. Waters et al., A New Condition Specic Quality of Life
Scale for Children with Cerebral Palsy, Patient Reported Out-
comes Newsletter 35 (2005), 1012.
[39] E. Waters et al., Quality of life instruments for children and
adolescents with neurodisabilities: how to choose the appro-
priate instrument, Developmental Medicine and Child Neurol-
ogy 51 (2009), 660669.
[40] G. Wells et al., Minimal clinically important differences: re-
view of methods, Journal of Rheumatology 28 (2001), 406
412.
[41] M. White-Koning et al., Determinants of child-parent agree-
ment in quality-of-life reports: a European study of children
with cerebral palsy, Pediatrics 120 (2007), e804814.
[42] M. White-Koning et al., Parent and professional reports of the
quality of life of children with cerebral palsy and associated
intellectual impairment, Developmental Medicine and Child
Neurology 50 (2008), 618624.
[43] E. Wood and P. Rosenbaum, The gross motor function clas-
sication system for cerebral palsy: a study of reliability and
stability over time, Developmental Medicine and Child Neu-
rology 42 (2000), 292296.
Copyright of Journal of Pediatric Rehabilitation Medicine is the property of IOS Press and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.