Developmental Neurorehabilitation

ISSN: 1751-8423 (Print) 1751-8431 (Online) Journal homepage: http://www.tandfonline.com/loi/ipdr20

Reliability of the gross motor function

classification system and the manual ability
classification system in children with cerebral
palsy in Tanzania

Daniele Piscitelli, Stefano Vercelli, Roberto Meroni, Giulia Zagnoni &

Leonardo Pellicciari

To cite this article: Daniele Piscitelli, Stefano Vercelli, Roberto Meroni, Giulia Zagnoni &
Leonardo Pellicciari (2017): Reliability of the gross motor function classification system and the
manual ability classification system in children with cerebral palsy in Tanzania, Developmental
Neurorehabilitation, DOI: 10.1080/17518423.2017.1342710

To link to this article: http://dx.doi.org/10.1080/17518423.2017.1342710

Published online: 10 Jul 2017.

Submit your article to this journal

View related articles

View Crossmark data

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

http://www.tandfonline.com/action/journalInformation?journalCode=ipdr20

Download by: [Cornell University Library] Date: 10 July 2017, At: 14:43
DEVELOPMENTAL NEUROREHABILITATION
https://doi.org/10.1080/17518423.2017.1342710

ORIGINAL ARTICLE

Reliability of the gross motor function classification system and the manual ability
classification system in children with cerebral palsy in Tanzania
Daniele Piscitellia,b, Stefano Vercellic, Roberto Meronid, Giulia Zagnonie,b, and Leonardo Pellicciarif
a
PhD Program in Neuroscience, School of Medicine and Surgery, University of Milano-Bicocca, Monza, Italy; bNyumba Ali Organization, Bologna,
Italy; cLaboratory of Ergonomics and Musculoskeletal Disorders Assessment, Division of Physical Medicine and Rehabilitation, Salvatore Maugeri
Foundation - IRCCS, Scientific Institute of Veruno, Novara, Italy; dSchool of Medicine and Surgery, Program in Physical Therapy, University of Milano-
Bicocca, Milan, Italy; eSchool of Medicine, Program in Physical Therapy, University of Bologna, Bologna, Italy; fPhD Program in Advanced Sciences
and Technologies in Rehabilitation Medicine and Sports, Department of Clinical Sciences and Translational Medicine, Tor Vergata University of
Rome, Rome, Italy

ABSTRACT ARTICLE HISTORY

Objectives: Gross Motor Function Classification System (GMFCS) and Manual Ability Classification System Received 23 February 2017
(MACS) are broadly studied in high-income countries, but data concerning their functioning are lacking Revised 3 June 2017
in developing countries. Therefore, we analyzed their reliability and sensitivity to change in children with Accepted 11 June 2017
cerebral palsy in Tanzania. Methods: GMFCS and MACS are two ordinal grading systems used to assess KEYWORDS
motor functions while observing children’s performances. Forty-nine children were classified by two Cerebral palsy; developing
independent physiotherapy students at baseline, after one month and after one year. Reliability and countries; gross motor
sensitivity to change were analyzed using intraclass correlation coefficient (ICC), effect size (ES), standard function; outcome
response mean (SRM), standard error of measurement (SEM), and minimal detectable change (MDC). assessment; psychometrics
Results: Inter- (ICC = 0.97/0.95 for GMFCS/MACS) and intra-rater reliability (ICC = 0.98/0.96 GMFCS/MACS)
were excellent. Sensitivity to change was small (ES = −0.14/0.11, SRM = −0.24;/0.24 GMFCS/MACS). SEM
was 0.2 points, resulting in MDC = 0.5/0.7 for GMFCS/MACS, respectively. Conclusions: GMFCS and MACS
demonstrated excellent reliability, but not sensitivity to change.

Introduction Palisano et al.8 reported good and high reliability of GMFCS

in Canadian children less than two years and aged between
Cerebral palsy (CP) is a movement and posture disorder
two and twelve, respectively. High test–retest and inter-obser-
secondary to brain injury or malformation occurring in the
ver reliability was confirmed in populations of children from
early childhood.1 By the year 2025, the number of people with
Turkey,12 Brazil,13 China14 and Greece.15 In all these studies,
disabilities in low- and middle-income countries (LMICs) is
the inter-observer reliability was independent of the rater
expected to double,2 and the most recent estimate of CP
background (e.g., caregivers, physiotherapists, physicians, or
prevalence in African populations ranges from 23 to 10/
teachers). High stability over time of the GMFCS was also
1000.4 Several gaps currently exist in the management of CP
reported by studies carried out in Australia,16 Canada,17 and
in Africa, including the lack of validated assessment tools.5 To
more recently in Sweden.18 In particular, Alriksson-Schmidt
capture a broad view of the child’s functional level, gross
et al.18 observed that changes were often transient, and down-
motor function and manual ability should be assessed
ward change (higher performance) was more likely in GMFCS
separately.6,7 Two instruments are commonly used worldwide
levels II and III than in the other levels. The reliability of the
for these purposes: the Gross Motor Function Classification
MACS was firstly explored in the original validation study9
System (GMFCS)8 and the Manual Ability Classification
carried out across Australia and Sweden and then confirmed
System (MACS).9 These measures enable health professionals
in different countries such as Iran,19 Turkey,20 and Korea.21
to quantify motor function and manual ability, establish ther-
Findings showed similar results, reporting high intra- and
apeutic goals, and improve the communication with relatives
inter-rater reliability. High stability has been reported for
or caregivers.1,10 Moreover, reliable classification systems
ratings over 12 months16 and up to five years22 in
could help potential resource allocation issues and impact
Australian and Sweden children, respectively. However, func-
on children’s functional well-being. Both instruments have
tional scales developed and validated in high-income coun-
been well studied in different settings of high-income coun-
tries might lack relevance to children in LMICs since severity
tries. Reliability (i.e., giving the same result at different times
of disability does not only depend on impairment but also the
of observation in stable patients) and responsiveness (i.e.,
contextual factors.23 These contextual factors, such as physical
ability to detect changes over time) are two important psycho-
environment, culture, attitudes of the community toward dis-
metric properties that such measures should demonstrate.11
ability and persons with disabilities, may vary country to

CONTACT Daniele Piscitelli, PhD d.piscitelli1@campus.unimib.it Program in Neuroscience, School of Medicine and Surgery, University of Milan Bicocca,
Piazza Madonnina 1, I-20841 - Carate Brianza (MB), Italy.
© 2017 Taylor & Francis
2 D. PISCITELLI ET AL.
country and are thought to have an impact on the functioning
of children with CP.24 Therefore, to be used in LMICs, the
GMFCS and MACS should be locally validated. The influence
of socioeconomic status on the development process of chil-
dren has been demonstrated in a cross-sectional study.25 In
that study, 49 Brazialian children were organized in two
groups considering their high or low socioeconomic status
and classified according to the GMFCS; children with severe
cerebral palsy with low socioeconomic status presented worse
performance in self-care skills (p = 0.021) and mobility
(p = 0.005). These children were also more dependent regard-
ing mobility (p = 0.015) than those with high socioeconomic
status. This further supports the need to analyze the psycho-
metric properties of the MACS and GMFCS before they are
used to classify children living in LMICs. To our knowledge,
no study has analyzed the reliability and sensitivity to change
of these instruments in a LMIC.
The aim of this study was to contribute to the knowledge
base relating to the classification of functions experienced by
children with mobility and manual ability impairments
attending a non-government organization providing services
to children with disabilities in Tanzania.
The main objective was to assess the reliability of the
GMFCS and MACS in a population of Tanzanian children
with CP, while secondary aims were to explore the sensitivity
to change and the correlation between the two instruments.
Methods
Participants
All the children attending two Nyumba Ali day-care centers in the
Iringa district, Tanzania, in the period between April 2014 and
April 2015 were consecutively enrolled for this study. Nyumba Ali
is a non-profit organization whose aim is to improve activity,
participation, and the quality of life of children with disability. It
provides community-based rehabilitation (CBR) services mana-
ged by a multidisciplinary team involving trained caregivers, and
occasionally medical doctors, nurses, physiotherapists.
Children were enrolled if they met the following criteria:
diagnosis of CP as defined by Rosenbaum et al.1 and age under
18 years. Children were excluded if they presented neuromotor
(e.g., spina bifida), or neuromuscular (e.g., muscular dystrophy,
myopathy) disease, or a history of surgery. All the enrolled chil-
dren come from the same low socioeconomic status and under-
went a comprehensive daily physical therapy program following
CBR guidelines26 that included passive movements of the upper
and lower limbs, stretching, functional (e.g., sitting, walking and
postural control training), and gross motor exercises under the
guidance of health professionals and trained caregivers. Prior to
the study, written informed consent was obtained from the par-
ents or guardians of all participants. The study was conducted in
accordance with the Declaration of Helsinki.
Measures
The GMFCS8 is a standardized 5-level ordinal grading
method of classifying motor function in children with CP.
The grade is assigned based on reported or observed
locomotion performances: level I = walks without limitations;
level II = walks with limitations; level III = walks using a
hand-held mobility device; level IV = self-mobility with lim-
itations; level V = requires transport in a wheelchair. In the
original version, separate descriptions are provided for chil-
dren across four age groups: < 2 years, 3–4 years, 5–6 years,
and 7–12 years. In this study, we used the modified version24
that includes a 5th age group for children aged 12–18 years,
with a single rating to describe the child’s ambulatory perfor-
mance across different terrains and distances. The scale is easy
and fairly quick to administer, without need for special train-
ing for raters and takes 5–15 minutes to complete.12
The MACS9 is a 5-level system to assess manipulation tasks in
children aged 4–18 years with CP. Raters assign the MACS score
indicating the level which best represents the child’s usual man-
ual ability while observing the child in daily activities. Each level
is determined based on appropriate activities using objects typi-
cal of the child’s environments and age: level I = handles most
objects easily and successfully; level II = handles most objects but
with somewhat reduced ability and/or speed of achievement;
level III = handles objects with difficulty and requires help in
preparing and/or modifying activities; level IV = handles a
limited selection of easily managed objects in adapted situations;
level V = does not handle objects and has severely limited ability
to perform even the simplest of actions. The MACS administra-
tion requires no special training for assessors.
In this study, the Italian versions of both GMCFS (available
at: https://canchild.ca/en/resources/42-gross-motor-function-
classification-system-expanded-revised-gmfcs-e-r) and MACS
(available at: www.macs.nu) were used.
Procedures
The raters in the study were two physiotherapy students
previously trained in administration and scoring of GMFCS
and MACS that consisted in reading the features of the two
instruments. Children were classified following standardized
procedures at baseline (T0), after one month (T1), and at a
one-year follow-up (T2). Raters were blinded from each other.
The assessment was performed by observing children during
recreational and daily life activities.
Statistical analysis
All the variables were reported by means of descriptive
statistics. Furthermore, the presence of ceiling\floor effects
considered if more than 15% of the subjects obtained the
highest or lowest score27 were reported. Normal distribu-
tion of the sample was evaluated with the Kolmogorov–
Smirnov test. An acceptable range for normality was estab-
lished for significance values >0.05 and for a skewness lying
between -1 and +1; also visual inspection of quantile-quan-
tile plots was used.28
The first aim of this study was to analyze inter- and intra-
observer reliability of the GMFCS and MACS, that is, the extent
to which the two scales are consistent and free from error.29
Intra-observer reliability was calculated by comparing the T0
and T1 datasets, whereas inter-observer reliability was computed
comparing the scores of the two observers on pooled data (T0
 DEVELOPMENTAL NEUROREHABILITATION 3

and T1). Because the two assessments points were separated by 3
to 4 weeks, a non-parametric Wilcoxon test was used to confirm
that the classification of the sample had not changed. Then, an
intraclass correlation coefficient (ICC2,1) with 95% confidence
intervals (CI) was used to examine the reliability.30 ICC values
>0.75 are the minimal requirement for a reliable measure,31 but
values >0.90 are usually considered as essential for indicating
excellent reliability in a clinical measurement at the individual
level.29 The sample size required to test reliability was deter-
mined expecting to obtain ICC values of .90, with a 95%CI close
to 0.2;32 therefore, a minimum number of 21 children were
required. The standard error of measurement (SEM) and mini-
mum detectable change (MDC) were also calculated to investi-
gate the degree of uncertainty connected with the measurement
error. Unlike the ICC, the SEM and MDC have the advantage of
being largely independent of the population from which they
were calculated, and of being expressed in the same unit as the
tests.33 SEM was computed with the following formula: SD*√(1–
ICC), where SD is the baseline standard deviation of the mea-
surements and the ICC value is that of intra-rater reliability.
MDC was calculated by multiplying the SEM by the z-score
associated with the 95% confidence level and the square root
of 2.
The secondary purpose of this study was to evaluate the
sensitivity to change of the two scales. Data of T0 and T2 were
used to calculate the effect size (ES) and the standardized
response mean (SRM). ES was calculated by dividing the
average difference in change scores (pre-test scores minus
post-test scores) by the SD of the baseline. SRM is the average
difference (pre-test scores minus post-test scores) divided by
the standard deviation of the differences between the paired
measurements. ES and SRM were interpreted as suggested by
Cohen:34 large effect for values >0.80, moderate effect for
values between 0.80 and 0.50, and small effect for values
between 0.50 and 0.20.
Finally, the correlation between GMFCS and MACS was
calculated by means of Spearman’s rank correlation coefficient
(ρ) on the T0 and T2 pooled dataset. Results were rated as follows:
ρ > 0.70 = strong correlation, 0.50 < ρ < 0.70 = moderate correla-
tion, and ρ < 0.50 = weak correlation.35 The intention here was to
analyze the relationship between gross motor function and man-
ual ability, not to establish concurrent validity. As the two scales
assess different—but related—functions, a moderate correlation
was considered a positive finding.
For all statistical analyses, the α value was set at p < 0.05.
SPSS statistical software program, version 21.0 for Windows
(SPSS Inc., Chicago, IL, USA) was used.
Results
Study participants
Demographics and clinical data of the study population are
reported in Table 1. A total of 49 children (21 females, 28 males)
were enrolled. The sample size used to test reliability was adequate
for both the GMFCS and MACS. The sample for the analyses of
two instruments consists of a different children number, namely
49 subjects for the GMFCS and 36 subjects for the MACS. This
mismatch was due to the different range of years covered by the
two tools, that is, MACS is not applicable for children < 4 years.
Tables 2 and 3 show the distribution of GMFCS and MACS
according to age group and score distribution at each time point,
respectively. An important ceiling effect was reported for the
GMFCS (24 children out of 49—about 50%—obtained the max-
imum score), and MACS (14 children of 36—about 38%—
obtained the maximum score). No significant floor effect was
found for either scale (about 10% and 11% for GMFCS and
MACS, respectively).
Reliability
As data were not normally distributed, a non-parametric test
was used. The Wilcoxon test did not show significant
(p > 0.05) differences between T0 and T1 for either scale,
confirming that the conditions of children were stable enough
to proceed with the reliability analysis.
Excellent reliability was observed for both scales, both for
the intra- and inter-rater reliability. ICCs with their CIs are
summarized in Table 4. GMFCS and MACS scores showed to
be almost free from error, as SEM and MDC values were
below one scale point (Table 4).
Sensitivity to change
Low values for ES and SRM (respectively –0.14 and –0.24 for
GMFCS; 0.11 and 0.18 for MACS) were found. This indicates
a small magnitude of the change score, that is, a small sensi-
tivity to change.

Table 1. Demographics and clinical data of study population at baseline.

Characteristics GMFCS (N = 49) MACS (N = 36)
Age (years) 6.5 ± 4.1 8.0 ± 3.8
Sex
Male 28 (57%) 23 (64%)
Female 21 (43%) 13 (36%)
Type of cerebral palsy
Hypotonia 5 (10%) 4 (11%)
Spasticity 37 (76%) 27 (75%)
Dystonia/athetosis 6 (12%) 4 (11%)
Ataxia 1 (2%) 1 (3%)
Distribution of Motor Impairment
Quadriparetic 20 (41%) 17 (47%)
Diparetic 24 (49%) 17 (47%)
Hemiparetic 5 (10 %) 2 (6%)
Values are mean ± standard deviation (SD) or frequency and percentage. GMFCS: Gross Motor Function Classification System; MACS: Manual
Ability Classification System.
4 D. PISCITELLI ET AL.

Table 2. Distribution of gross motor function classification system and manual ability classification system levels according to the age groups.
GMFCS scores MACS scores
Age groups I II III IV V I II III IV V
2–4 years 1 – 1 4 7 – – – – –
4–6 years 1 – 1 2 6 – 3 1 – 6
6–12 years 2 2 2 5 8 3 6 2 2 6
12–18 years 1 – – 3 3 1 – 2 2 2
Total 5 2 4 14 24 4 9 5 4 14
GMFCS: Gross Motor Function Classification System; MACS: Manual Ability Classification System.

Table 3. Distribution of gross motor function classification system and manual ability classification system scores.
GMFCS MACS
T0 T1 T2 T0 T1 T2
N 49 44 22 36 30 15
Median 4 4.5 5 3.5 3 5
25th percentile 4 4 4 2 2 3
75th percentile 5 5 5 5 5 5
GMFCS: Gross Motor Function Classification System; MACS: Manual Ability Classification System; T0: baseline assessment; T1: assessment after one month;
T2: assessment after one year. Note that not everyone got rated with both classification systems because of the restricted age of interest of MACS.

Table 4. Inter- and intra-rater reliability of gross motor function classification system and manual ability classification system.
Inter-rater reliability Intra-rater reliability SEM MDC
Scale ICC (95% CI) Z-score p ICC (95% CI) Z-score p
GMFCS 0.97 (0.96–0.98) −1.67 0.096 0.98 (0.97–0.99) −0.45 0.65 0.2 0.5
MACS 0.96 (0.93–0.97) −1.73 0.083 0.96 (0.94–0.98) 0.00 1.0 0.2 0.7
CI: Confidence Intervals; GMFCS: Gross Motor Function Classification System; ICC: Intraclass Correlation Coefficient; MACS: Manual Ability Classification System; MDC:
Minimum Detectable Change; SEM: Standard Error of Measurement.

Relationship between GMFCS and MACS
A strong correlation was found between GMFCS and MACS
scores (ρ = 0.808; p < 0.001).
Discussion
The GMFCS is considered a valid instrument for classifying
the gross motor abilities and limitations of children with CP,36
while the MACS gives a valid single grade for the child’s level
of collaborative use of both hands when handling objects in
daily life.
In the present study, we assessed the reliability and the
sensitivity to change of the two instruments in a sample of
children with CP living in Tanzania, which differs from high-
income countries in terms of health policies, socioeconomic and
cultural values.37 For example, in the western world, it is cus-
tomary to sit on a chair, while in Africa sitting on the floor or
using the Asian low-level squatting type toilet is widely prac-
ticed. The ability to get into and out of a squatting position (and
maintain it) requires higher muscle activities, causing more
difficulty for the children. Many other environmental and per-
sonal factors may negatively influence the activity and mobility
of children in Tanzania. One might assume that all materials to
perform manual activities would be in place, so a child would
just have to perform them. This could be common for high-
income countries and hospital settings, but not for community
settings of low-income countries.38 Walking outside can be also
a very big challenge: hand-held mobility devices, wheelchairs, or
assistive technology are only available for some children, and
their mobility is hampered by physical barriers such as narrow
and overcrowded footpaths, presence of unpaved roads with
uneven surface and potholes, or absence of handrails on stairs.
Furthermore, even those who overcome architectural barriers
will face the disability-unfriendly systems which restrict their
participation in many activities, including participation in lei-
sure activities and sports.
To our knowledge, this is the first study to investigate the
psychometric properties of GMFCS and MACS in a LMIC.
Excellent inter- and intra-observer reliability coefficients
were found for both scales, in line with previous studies
carried out in higher socioeconomic levels. El et al.12 studied
the reliability of the GMFCS in 136 Turkish children with CP
and found ICC values of 0.97 for inter-observer and .94 for
intra-observer reliability. Comparable results were reported
also for the Brazilian13 (ICC 0.95; CI 0.86–0.98) and Greek15
(kappa value 0.80; CI 0.67–0.94) GMFCS versions. Although
socioeconomic levels of participants in the mentioned studies
were not specified, according to World Bank classification39,
these countries fall into high-income (Greece) or upper-mid-
dle-income economies (Brazil and Turkey). The GMFCS was
found to be less reliable in the lowest8 age group (children <
2 years) compared to other age groups (children 2–12 year of
age). However, in the present study, no children < 2 years old
were enrolled, and sub-group analysis on reliability was not
possible due to underpowered statistics.
Excellent inter-rater reliability of the MACS was reported
both for the original9 (ICC 0.97; CI 0.96–0.98), and for the
Turkish20 (ICC 0.96; CI 0.95–0.97) versions. The above-men-
tioned studies included physiotherapy degree candidates as
raters, indicating that administration of the two question-
naires probably does not require particular skills to produce

reliable results. Thus, their use seems appropriate also in
contexts with limited health professional resources. This was
confirmed by studies where raters were families or health care
operators administrating the MACS (ICC range 0.73 to
0.85).40,41 However, it should be emphasized that reliability
coefficients of the two scales may have been inflated by the
ceiling effects observed in this population and must be inter-
preted with caution.
A low ability to detect change over time after a 1-year interval
was found for both scales. Results are in line with those of previous
studies conducted in high-income countries. Ohrvall et al.22 found
a high MACS stability at 1-year interval (ICC 0.97; CI 0.97–0.97),
and Imms et al.16 found high stable ratings (ICCs >0.9) for MACS
and GMFCS administered by caregivers over 12 months. The low
sensitivity to change may be also attributable to other factors such
as the presence of a ceiling effect, or to the fact that the variables
are classified into five levels only42 and the broadness of each
category appears to contain a large amount of the measured
variable.22 Therefore, to achieve a higher score, the children’s
functional abilities need to improve greatly. However, GMFCS
and MACS are intended be used as classification tools to quantify
the motor function and the manual ability characteristics of
children16,22 and should not replace other instruments—such as
the Gross Motor Function Measure and the Assisting Hand
Assessment—specifically developed as outcome measures.43,44
Furthermore, their use in clinical practice may improve the com-
munication between parents and professionals, help families to
understand the child’s functional capabilities, and plan for future
needs and treatments.17,45 These instruments can be useful also in
research in order to classify children into homogenous clusters
(e.g., matching control and experimental groups).45
A moderate correlation between GFMCS and MACS total
score was found. Earlier studies reported either high46,47
(rs = 0.69–0.73, p < 0.001) or poor7,20 (kappa value = 0.35,
CI 0.27–0.41, with a significantly lower level of MACS than
GMFCS) agreement between the two scales. These conflicting
results may be due to differences in the type of populations
studied. In this study, almost half of the sample had severe
quadriparetic CP, leading to major manual ability and motor
impairments. Therefore, the level of correlation between the
two scales was likely inflated by a ceiling effect.
The clinical characteristics of the sample in this study
warrant some considerations. Diparetic and spastic CP were
the most common clinical subtypes, in line with epidemiolo-
gic data from Western7 and African48 countries. However,
children’s functioning was much lower than the average
score of matched age children in high-income countries,
confirming the hypothesis that socioeconomic and cultural
values would have a negative impact on disability. At first
evaluation, almost all of the children were classified at level IV
or V on the GMFCS, while in the large study of Palisano
et al.36, the children were fairly evenly distributed among the
five levels, with more than half classified at level I to III.
Similarly, the distribution of MACS level was skewed to
worst manual ability in this study, whereas more than 50%
of the children belonged to level I or II in the study of
Eliasson et al.9 conducted in Sweden and Australia.
Previous study outlined that by middle childhood children
with CP do not make substantial changes in the gross motor
DEVELOPMENTAL NEUROREHABILITATION 5
abilities.36 Based on this remark, we expected to observe little
or none improvement after 1 year. Instead, a number of
children (22% and 13% for GMFCS and MACS, respectively)
showed an unexpected lowering of functioning. This may
indicate that socioeconomic factors in LMICs—such as poor
nutrition, lack of facilities, low availability of medication or
amount of therapy provided, inadequate caregiver assistance,
or scant use of assistive technology—could play a significant
role in mobility and manual ability of children with CP. The
present findings also caution against the use of predictive
curves of gross motor functioning previously validated in a
large sample of Canadian patients with CP, when children
from LMICs are considered.36
This study had some limitations. The sample size was
adequate to analyze reliability, but did not allow to calculate
ICCs for the different age groups or functional levels. The
sample size used for the sensitivity to change analysis was
small, although previous studies involving a larger sample
obtained similar results. Also, the selection of children was
based on a convenience sampling and was available for a
limited period only in a setting of humanitarian organizations.
Although the two day-care centers in the Iringa district follow
the World Health Organization guidelines on CBR strategy
and they can be considered representative of those existing in
Tanzania, results may be not generalizable to other contexts.
Hence, further validation studies in different geographical
areas, clinical settings, or with different pathologies are
recommended. Finally, our results need to be confirmed for
the MACS and GMFCS when administered by Tanzanian
caregivers.
Conclusions
This study investigated the psychometric properties of the
GMFCS and MACS in children with CP attending two day-
care centers providing CBR in a LMIC. Both scales showed
excellent reliability, but a low ability to detect change over
time. Therefore, these instruments should be accounted for
classifying motor function and manual ability characteristics
of the children. Further research is needed to confirm our
results and to implement the use of these motor assessment
tools in children with CP in LMICs.
Acknowledgments
The authors want to thank Paola Ghezzi, Filippo Miotello, Lara Pepponi,
and Federico Tammaro (University of Insubria, Varese, Italy) for collect-
ing data, and Rosemary Allpress for the linguistic revision.
Declaration of interest
The authors report no conflicts of interest. The authors alone are
responsible for the content and writing of the paper.
References
1. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M,
Damiano D, et al. A report: the definition and classification of
cerebral palsy April 2006. Developmental Medicine and Child
Neurology Supplement 2007;109:8–14.
6 D. PISCITELLI ET AL.
2. Peat M. Community Based Rehabilitation. Philadelphia: WB
Saunders Co, 1997.
3. El-Tallawy HN, Farghaly WM, Shehata GA, Metwally NA,
Rageh TA, Abo-Elfetoh N. Epidemiology of cerebral palsy in
El-Kharga District-New Valley (Egypt). Brain & Development
2011;33:406–411.
4. Couper J. Prevalence of childhood disability in rural KwaZulu-
Natal. South African Medical Journal 2002;92:549–552.
5. Donald KA, Kakooza AM, Wammanda RD, Mallewa M, Samia P,
Babakir H, et al. Pediatric cerebral palsy in Africa: where are we?
Journal of Child Neurology 2015;30:963–971.
6. Ohrvall AM, Eliasson AC. Parents’ and therapists’ perceptions of
the content of the manual ability classification system, MACS.
Scandinavian Journal of Occupational Therapy 2010;17:209–216.
7. Carnahan KD, Arner M, Hagglund G. Association between gross
motor function (GMFCS) and manual ability (MACS) in children
with cerebral palsy. A population-based study of 359 children.
BMC Musculoskeletal Disorders 2007;8:50.
8. Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi
B. Development and reliability of a system to classify gross motor
function in children with cerebral palsy. Developmental Medicine
and Child Neurology 1997;39:214–223.
9. Eliasson AC, Krumlinde-Sundholm L, Rosblad B, Beckung E,
Arner M, Ohrvall AM, et al. The manual ability classification
system (MACS) for children with cerebral palsy: scale develop-
ment and evidence of validity and reliability. Developmental
Medicine and Child Neurology 2006;48:549–554.
10. Morris C, Bartlett D. Gross motor function classification system:
impact and utility. Developmental Medicine and Child Neurology
2004;46:60–65.
11. Streiner DL, Norman GR. Health Measurement Scales. A Practical
Guide to their Development and Use, 4th. Oxford: Oxford
University Press, 2008.
12. El O, Baydar M, Berk H, Peker O, Kosay C, Demiral Y.
Interobserver reliability of the Turkish version of the expanded
and revised gross motor function classification system. Disability
and Rehabilitation 2012;34:1030–1033.
13. Hiratuka E, Matsukura TS, Pfeifer LI. Cross-cultural adaptation of
the gross motor function classification system into Brazilian-
Portuguese (GMFCS). Revista Brasileira De Fisioterapia (Sao
Carlos (Sao Paulo, Brazil)) 2010;14:537–544.
14. Shi W, Yang H, Li CY, Zhou MQ, Zhu M, Wang Y, et al.
Expanded and revised gross motor function classification system:
study for Chinese school children with cerebral palsy. Disability
and Rehabilitation 2014;36:403–408.
15. Papavasiliou AS, Rapidi CA, Rizou C, Petropoulou K, Tzavara C.
Reliability of Greek version gross motor function classification
system. Brain & Development 2007;29:79–82.
16. Imms C, Carlin J, Eliasson AC. Stability of caregiver-reported
manual ability and gross motor function classifications of cerebral
palsy. Developmental Medicine and Child Neurology
2010;52:153–159.
17. Palisano RJ, Cameron D, Rosenbaum PL, Walter SD, Russell D.
Stability of the gross motor function classification system.
Developmental Medicine and Child Neurology 2006;48:424–428.
18. Alriksson-Schmidt A, Nordmark E, Czuba T, Westbom L.
Stability of the gross motor function classification system in
children and adolescents with cerebral palsy: a retrospective
cohort registry study. Developmental Medicine and Child
Neurology 2017;59:641–646.
19. Riyahi A, Rassafiani M, AkbarFahimi N, Sahaf R, Yazdani F. Cross-
cultural validation of the Persian version of the Manual Ability
Classification System for children with cerebral palsy. International
Journal of Therapy and Rehabilitation 2013;20:19–24.
20. Akpinar P, Tezel CG, Eliasson AC, Icagasioglu A. Reliability and
cross-cultural validation of the Turkish version of Manual Ability
Classification System (MACS) for children with cerebral palsy.
Disability and Rehabilitation 2010;32:1910–1916.
21. Jang DH, Sung IY, Kang JY, Lee SI, Park JY, Yuk JS, et al.
Reliability and validity of the Korean version of the manual ability
classification system for children with cerebral palsy. Child: Care,
Health and Development 2013;39:90–93.
22. Ohrvall AM, Krumlinde-Sundholm L, Eliasson AC. The stability
of the manual ability classification system over time.
Developmental Medicine and Child Neurology 2014;56:185–189.
23. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW,
Knol DL, et al. The COSMIN study reached international con-
sensus on taxonomy, terminology, and definitions of measure-
ment properties for health-related patient-reported outcomes.
Journal of Clinical Epidemiology 2010;63:737–745.
24. Palisano RJ, Rosenbaum P, Bartlett D, Livingston MH. Content
validity of the expanded and revised gross motor function classi-
fication system. Developmental Medicine and Child Neurology
2008;50:744–750.
25. Assis-Madeira EA, Carvalho SG, Blascovi-Assis SM.
Functional performance of children with cerebral palsy from
high and low socioeconomic status. Revista Paulista De
Pediatria 2013;31:51–57.
26. Khasnabis C, Heinicke Motsch K, Achu K, Al Jubah K, Brodtkorb
S, Chervin P, et al. Community-Based Rehabilitation: CBR
Guidelines. Geneva: World Health Organization, 2010.
27. McHorney CA, Tarlov AR. Individual-patient monitoring in clin-
ical practice: are available health status surveys adequate? Quality
of Life Research 1995;4:293–307.
28. Chan YH. Biostatistics 101: data presentation. Singapore Medical
Journal 2003;44:280–285.
29. Portney LG, Watkins MP. Foundations of Clinical Research:
Applications to Practice (Vol. 2). Upper Saddle River, NJ:
Prentice Hall, 2000.
30. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater
reliability. Psychological Bulletin 1979;86:420–428.
31. Post MW. What to do with “moderate” reliability and validity
coefficients? Archives of Physical Medicine and Rehabilitation
2016;97:1051–1052.
32. Bonett DG. Sample size requirements for estimating intraclass
correlations with desired precision. Statistics in Medicine
2002;21:1331–1335.
33. Weir JP. Quantifying test-retest reliability using the intraclass
correlation coefficient and the SEM. Journal of Strength and
Conditioning Research /National Strength & Conditioning
Association 2005;19:231–240.
34. Cohen J. Statistical Power Analysis for the Behavioral Sciences.
2nd Ed. Manwah, NJ: Lawrence Erlbaum, 1988.
35. Munro B. Statistical Methods for Health Care Research.
Philadelphia, PA: JB Lippincott, 2000.
36. Palisano RJ, Hanna SE, Rosenbaum PL, Russell DJ, Walter SD,
Wood EP, et al. Validation of a model of gross motor function for
children with cerebral palsy. Physical Therapy 2000;80:974–985.
37. Organization. WH. World health statistics 2015 [Internet].
Geneva, Switzerland: World Health Organization. [cited 2016 13
May]. Available from: http://www.who.int/gho/publications/
world_health_statistics/en/
38. Mahmud I, Clarke L, Ploubidis G. Developing the content of a
locomotor disability scale for adults in Bangladesh: a qualitative
study. Archives of Physiotherapy.
39. World Bank Country and Lending Groups [Internet].
Washington, DC: The World Bank. [cited 2017 May 11].
Available from: https://datahelpdesk.worldbank.org/knowledge
base/articles/906519
40. Morris C, Galuppi BE, Rosenbaum PL. Reliability of family report
for the gross motor function classification system. Developmental
Medicine and Child Neurology 2004;46:455–460.
41. Morris C, Kurinczuk JJ, Fitzpatrick R, Rosenbaum PL. Reliability
of the manual ability classification system for children with cere-
bral palsy. Developmental Medicine and Child Neurology
2006;48:950–953.
42. Hobart JC, Cano SJ, Zajicek JP, Thompson AJ. Rating scales as
outcome measures for clinical trials in neurology: problems,
solutions, and recommendations. The Lancet Neurology
2007;6:1094–1105.

43. Krumlinde-Sundholm L, Holmefur M, Kottorp A, Eliasson AC.
The assisting hand assessment: current evidence of validity, relia-
bility, and responsiveness to change. Developmental Medicine and
Child Neurology 2007;49:259–264.
44. Ko J, Kim M. Reliability and responsiveness of the gross motor
function measure-88 in children with cerebral palsy. Physical
Therapy 2013;93:393–400.
45. Wood E, Rosenbaum P. The gross motor function classification
system for cerebral palsy: a study of reliability and stability over
time. Developmental Medicine and Child Neurology
2000;42:292–296.
46. Hidecker MJ, Ho NT, Dodge N, Hurvitz EA, Slaughter J,
Workinger MS, et al. Inter-relationships of functional status in
DEVELOPMENTAL NEUROREHABILITATION 7
cerebral palsy: analyzing gross motor function, manual ability,
and communication function classification systems in children.
Developmental Medicine and Child Neurology 2012;54:737–742.
47. Compagnone E, Maniglio J, Camposeo S, Vespino T, Losito L, De
Rinaldis M, et al. Functional classifications for cerebral palsy:
correlations between the gross motor function classification sys-
tem (GMFCS), the manual ability classification system (MACS)
and the communication function classification system (CFCS).
Research in Developmental Disabilities 2014;35:2651–2657.
48. El-Tallawy HN, Farghaly WM, Shehata GA, Rageh TA, Metwally
NA, Badry R, et al. Cerebral palsy in Al-Quseir City, Egypt:
prevalence, subtypes, and risk factors. Neuropsychiatric Disease
and Treatment 2014;10:1267–1272.