Functional Classification Systems for Children with Cerebral Palsy:An ICF-Based Approach

Health Condition

(Disorder/disease)

Activities

(Activity limitation)

Body Structures and Functions

Participation

(Participation restriction)

Environmental Factors

Personal Factors

Table 1: The method of development and validation of functional classification systems and psychmetric properties

Year of publication

Authors

Subject of study

 

Population of study

 

Study method

 

Study Results

 

2019

 

Giovanni Baranello et al.

 

Construction and validation of a visual function classification system (VFCS)

The reliability study included 29 specialists, 39 parents and an overall sample of 160 children with CP (mean age 6 years and 6 months in the range of 1-19 years)

 

Includes four steps:

(1) Draft five levels of literature review analysis and clinical experience

(2) Validation of structures and revision of levels for conceptual significance using the nominal group process

(3) Amended by the Delphi International Survey

(4) Assessing the reliability between evaluators among specialists as well as caregivers, and the reliability of test-retest

The absolute agreement between the evaluators was 86%.

Test-retest reliability was high.

Reliability between parents and professionals in 39 average children.

As a final conclusion, they stated that VFCS is properly constructed and is a reliable system for classifying the visual abilities of children with CP in the clinic and research environment.

 

2016

 

 

 

 

 

Eliasson et al.

 

Development of

Manual ability classification for children under four years

(Mini-MACS)

 

 

A sample of 61 children with CP aged 12 to 51 months was scored by a parent and two therapists for a total of 64 assessments.

 Level 1: Drafting the classification system

 Level 2: Nominal group process

 Level 3: Delphi Survey

Level 4: Reliability

The ICC coefficient was 0.90 between parents and therapists and 0.97 between the two therapists.

Moreover, this system seems to be applicable for children with CP 1-4 years.

2014

 

 

 

 

 

 

 

 

Diane  Sellers

 et al.

 

 

Establishing a system for classifying the eating and drinking ability (EDACS) of children with CP and evaluate its reliability

 

Reliability study consisting of 25 speech and language therapists and 48 parents and 129 children and people with CP (4-22 years and mean age 14 years)

 Level 1: Drafting the system

 Level 2: Nominal group process

 Level 3: Delphi Survey

Level 4: Reliability

Agreement and reliability between two groups of excellent therapists

and between the parent group and the therapists, the agreement was moderate to substantial and the reliability was good to excellent.

Finally, they concluded that the EDACS system is a valid and reliable system for classifying the ability to eat and drink in children with CP.

Ø Determining the validity and reliability of the Farsi version of the EDACS system by Riyahi et al. in 2019.

 

2011

Mary Cooley et al.

Establishing and validating a Communication Performance Classification System (CFCS)

 

Reliability between evaluators

 By 61 specialists and 68 parents / relatives and 69 children with CP

 2 to 18 years

 

An 11-member development team developed comprehensive descriptions of CFCS levels and reviewed four nominal groups of 27 participants. In a Delphi survey, 112 respondents commented on the clarity and usefulness of CFCS.

Reliability between evaluators was completed by 61 specialists and 68 parents / relatives who classified 69 children with CP.

Test-retest reliability was completed by 48 specialists with a minimum interval of 2 weeks between classifications.

Reliability between the two specialists was 0.66 and between the parent and the specialist was 0.49.

Reliability among evaluators among specialists for classifying children over 4 years of age had reached 0.77.

The reliability of the test-retest was 0.82.

CFCS interpretation showed content validity and test-retest reliability very good.

Reliability between good evaluators among experts

 and reliablity between average evaluators among parents/experts.

Ø  Determining the validity and reliabilty of the Farsi version of CFCS system by Soleimani et al. in 2015.

2008

Rosenbaum

 et al.

 

Construction, validity and reliability of a large motor function classification system in children with CP

 

110 children with CP 2-18 years

 

 Level 1: Drafting the system

 Level 2: Nominal group process

 Level 3: Delphi Survey

 Level 4: Reliability

 

In this study, the reliability between the evaluators as well as the reliability of the test repetition was high.

The positive predictive validity of GMFCS for predicting gait from 1 to 2 years of age up to 12 years of age was 0.74. 

Ø Determining the validity and reliability of the Farsi version of the GMFCS system by Dehghan et al. in 2010

Ø Evaluation of validity and reliability of the test - Parental version retest and reliability between therapists' prescription and GMFCS parental prescription by Riahi et al. in 2012

2006

Eliasson

 et al.

 

Manual Ability Classification System (MACS) for children with CP, how valid and reliable

 

Reliability between therapists for 168 children aged 4-18 years and 25 parents and pediatricians

 

Its validity was based on the experience of a skilled group, a review of articles, and a comprehensive analysis of children in a range of practices, resulting in a consensus on structures as well as on the content of the five levels. Then parents and therapists were interviewed about the content, simplicity and ease, explanation and preference of each level.

Reliability was also assessed between therapists for 168 children aged 4-18 years and 25 parents and pediatric therapists.

 

The results showed that MACS has good validity and reliability.

The internal correlation coefficient between therapists was 0.97 and between parents and therapists was 0.96, which is a sign of excellent agreement.

 

Ø Determining the validity and reliability of the Farsi version of the MACS system by Riahi et al. in 2013

Source

Type of Reliability

Subject Age Range, y

Reliability Statistics

Palisano et al [34]

 

 

Inter-rater reliability

 

ICC =0.96

Inter-rater reliability

older than 2 years of age

K=0.75

Inter-rater reliability

under 2 years of age

K=0.55

Randall et al [35]

Inter-rater reliability

4-11

0.98

Wood et al [36]

Inter-rater reliability

2-12

G=0.93

Test–retest reliability

G=0.79

Palisano et al [37]

Inter-rater reliability

under 2 years of age

K=0.55

Inter-rater reliability

2-12

K=0.75

Russell et al [38]

Test-retest reliability

over 12 months

ICC =0.99

Bodkin et al [39]

Interrater reliability

younger than two years old

K=0.55

Interrater reliability

two to 12 years old

K=0.75

El et al [40]

 

12-18

ICC=0.97

Test-retest reliability

ICC=0.94

Papavasiliou et al [41]

 

mean age 5.4 years

K=0.80

Piscitelli et al [42]

Inter- reliability

 

K=0.97

Intra-rater reliability

 

K=0.98

 

 GMFCS by parents

Riyahi et al [43]

 

2-12

ICC=0.92

 

K=0.61

GMFCS by therapists And parents

 

2-12

ICC=0.88

 

K=0.68

Dehghan et al [44]

Interrater reliability (IRR)

 

P<0.05

Source

Type of Reliability

MACS Version

 

Subject Age

Range, y

Reliability Statistics

Eliasson et al [21]

Interrater reliability

Original version

4-18

ICC = 0.97

Van Meeteren et al [45]

Interrater reliability

Original version

18-24

ICC = 0.83

Plasschaert et al [46]

Interrater reliability

Original version

1-5

Κ = 0.55 (younger than 2 y)

κ = 0.67 (2-5 y of age)

Jang et al [47]

Interrater reliability

Korean version

4-14

ICC = 0.92-0.96

Interrater reliability

Korean version

4-14

ICC = 0.96-0.98

Akpinar et al

Interrater reliability

Turkish version

4-18

ICC = 0.89-0.98

Test retest reliability

Turkish version

4-18

ICC = 0.91-0.98

Riyahi et al [20]

Interrater reliability

Farsi version

4-18

ICC = 0.96

Test retest reliability

Farsi version

4-18

ICC = 0.97

Morris et al [48]

Interrater reliability

Original version

6-12

ICC = 0.7-0.9

Mutlu et al [49]

Interrater reliability

Turkish version

4-18

ICC = 0.96

Randall et al [35]

Interrater reliability

Original version

4-11

κ = 1

Source

Type of Reliability

Subject Age Range, y

Reliability Statistics

 

Hidecker et al [50]

Inter-rater reliability

2-18

K = 0.82

 

Randall et al [35]

Inter-rater reliability

4-11

0.98

 

Soleymani et al [51]

between speech and language pathologists and occupational therapists

 

inter-rater reliability

2-18

0.81

 

between parents and occupational therapists

 

inter-rater reliability

2-18

0.74

 

between parents and speech and language pathologists

 

inter-rater reliability

2-18

0.88

 

for occupational therapists

 

test-retest reliability

2-18

0.96

for speech and language pathologists

 

test-retest reliability

2-18

0.98

 

for parents

 

test-retest reliability

2-18

0.94

 

Vander Zwart et al [52]

between parents and SLTs

 

Interrater reliability

range 2y 9mo–12y 10mo

r=0.54

 

between SLTs

 

Interrater reliability

range 2y 9mo–12y 10mo

r=0.78

 

SLT

 

intrarater reliability

range 2y 9mo–12y 10mo

r=0.85

 

Source

Reliability of EDACS levels

Subject Age Range, y

Reliability Statistics

Sellers et al [53]

Reliability of EDACS levels I to V

4-22

K = 0.72

ICC= 0.93

 

Reliability of EDACS levels of assistance

k=0.80

ICC= 0.92

Tschirren et al [54]

Interrater reliability of EDACS

mean age 9y 7mo

κ=0.94

 

Interrater Reliability of EDACS level of assistance

κ=0.89

                   Reliability measures associated with use of EDACS by speech and language therapists and parents

Sellers et al [53]

Reliability of EDACS levels I to V

4-22

k=0.45

 

ICC=0.86

 

Reliability of EDACS levels of assistance

k=0.64

 

ICC=0.77

Tschirren et al [54]

Interrater reliability of EDACS

mean age 9y 7mo

κ=0.82

 

Interrater Reliability of EDACS level of assistance

κ=0.89

Table 6: General summary headings for the GMFCS, MACS, Mini-MACS, CFCS, EDACS and VFCS

Level 5

Level 4

Level 3

Level 2

Level 1

Scale

Transported in a Manual Wheelchair

Self-Mobility with Limitations; May Use Powered Mobility

Walks Using a Hand-Held Mobility Device

Walks with Limitations

Walks without Limitations

GMFCS (aged 0–18)

Does not handle objects and has severely limited ability to perform even simple actions.

Handles a limited selection of easily managed objects in adapted situations

Handles objects with difficulty; needs help to prepare and/or modify activities.

Handles most objects but with somewhat reduced quality and/or speed of achievement.

Handles objects easily and successfully.

MACS

(aged 4–18)

Does not handle objects and has severely limited ability to perform even simple actions.

Handles a limited selection of easily managed objects in simple actions.

Handles objects with difficulty.

Handles most objects, but with somewhat reduced quality and/or speed of achievement.

Handles objects easily and successfully

Mini-MACS (aged 1–4)

Seldom effectively sends and receives, even with familiar partners

Inconsistently sends  and/or receives even  with familiar partners

Sends and receives with familiar partners  effectively, but  not  with  unfamiliar partners

Sends and receives with familiar and unfamiliar partners but  may need extra time

 

Sends and receives with familiar and unfamiliar partners effectively and efficiency

CFCS

 (aged 2–18)

Unable to eat and drink safely – tube feeding  may be  considered to  provide nutrition

Eats and drinks with  significant limitations to safety

Eats and drinks with  some limitations to  safety; there may be limitations to efficiency

Eats and drinks safely but with some  limitations to efficiency

Eats and drinks safely and efficiently

EDACS

 

(aged 3–18)

Does not use visual function even in very adapted  environments

Uses visual function in very adapted environments but  performs just part of vision- related activities

Uses visual function but needs some adaptations

Uses visual function successfully but  needs self- initiated compensatory strategies

Uses visual function easily  and successfully in vision- related activities

VFCS
(aged 1–19)

Riyahi et al. in 2022  [29]

 

MACS- GMFCS-CFCS-EDACS

There is a moderate and significant relationship between the MACS and the EDACS, the MACS and the CFCS, and the CFCS and the EDACS in children with CP.

Mutlu A et al. in 2018 [55]

GMFCS-MACS-CFCS

GMFCS levels showed a strong correlation with MACS levels (rs=0.78). MACS level was strongly correlated with CFCS levels (rs=0.73), particularly in quadriplegic children (rs=0.78). GMFCS levels were moderately correlated with CFCS levels (rs=0.71).

Margarta et al. in 2017 [27]

GMFCS-MACS-CFCS-EDACS

A high correlation between the four classifications was found (r> 0.70, p < 0.01).

Killian et al. in 2014 [26]

GMFCS-MACS-CFCS

There were moderate positive correlations between three FCS: GMFCS and MACS (T were no significant correlations between CFCS and the other FCS).

Compagnone E et al. in 2014 [56]

GMFCS-MACS-CFCS

A strong correlation was found between the three classifications: Level V of the GMFCS-E&R corresponds to Level V of the MACS (rs=0.67, p=0.001); the same relationship was found for the CFCS and the MACS (rs=0.73, p<0.001) and for the GMFCS-E&R and the CFCS (rs=0.61, p=0.001).

Oskoui et al. in 2013 [57]

GMFCS–MACS

The overall agreement between GMFCS and MACS Levels was moderate (kappa 0.457, standard error 0.034) with a strong positive correlation (Spearman rho of 0.820, standard error 0.023).

Himmelman et al. in 2013 [58]

CFCS-GMFCS-MACS

CFCS correlated with the GMFCS, MACS and cognitive function (p < 0.01).

Riyahi et al. in 2013 [20]

GMFCS-MACS

Correlation between GMFCS and MACS was also calculated, this demonstrated a moderate relationship (P=0.727).

Heidecker et al. in 2012 [50]

GMFCS–MACS

GMFCS levels were strongly correlated with MACS levels (rs=0.69, p<0.001).

Heidecker et al. in 2012 [50]

MACS–CFCS

MACS levels were moderately correlated with CFCS levels (rs=0.54, p<0.001).

Heidecker et al. in 2012 [50]

GMFCS–CFCS

GMFCS levels were moderately correlated with CFCS levels (rs=0.47, p<0.001),

Akpinar P. et al. in 2010 [30]

GMFCS–MACS

Total agreement between the GMFCS and the MACS occurred in only 45% of the children.

Gunel MK et al. in 2009 [59]

GMFCS–MACS- WeeFIM

A good correlation between the GMFCS and MACS was found in all children (r=0.735, p<0.01). There was also a correlation between the GMFCS and WeeFIM subscales according to subtypes and all parameters were correlated at the level of p<0.01, the same as the MACS.

Carnahan KD et al. in 2007 [60]

GMFCS–MACS

The overall agreement between GMFCS and MACS was poor (kappa value 0.35, 95% confidence interval 0.27–0.41).

Eliasson AC et al. in 2006 [21]

GMFCS–MACS

Correlation between GMFCS and MACS was also calculated, this demonstrated a moderate relationship.