Outcome Measures of Activity for Children With Cerebral Palsy: A Systematic Review : Pediatric Physical Therapy

Secondary Logo

Journal Logo


Outcome Measures of Activity for Children With Cerebral Palsy

A Systematic Review

Debuse, Dorothée PhD, MCSP; Brace, Hal MSc, MRes, MCSP

Author Information
doi: 10.1097/PEP.0b013e318227bbc6


Systematic evaluation of outcome is an integral part of evidence-based practice.1 Outcome measures (OMs) need to be reliable (consistent over time and when used by different raters) and valid (appropriate to assess what the clinician or researcher wants to assess in a manner that makes intrinsic sense).2 They also need to demonstrate specificity, that is, the ability to distinguish between the presence or absence of a condition in people,3 and responsiveness to change, that is, detect minimal, but clinically relevant, changes.2 The judicious choice of OM(s) is essential, particularly in patients with complex disabilities, such as cerebral palsy (CP).4

Rosenbaum et al recently defined cerebral palsy as “a group of permanent disorders of the development of movement and posture, causing activity limitation, that are attributed to non-progressive disturbances which occurred in the developing foetal or infant brain.”5 This definition reflects the shift from a predominant focus on impairments associated with CP (eg, increased muscle tone) to one that is more consistent with the biopsychosocial model of health embodied in the International Classification of Functioning, Disability and Health (ICF).6 Consistent with this shift of focus has been the development of measures of function/activity, rather than of impairment. A range of measures of function/activity is available now, developed specifically for children with CP,79 or more generally for children with long-term disabilities.10,11

However, evidence across time and specialist areas within physiotherapy and rehabilitation points to investigation of improving the systematic use of OMs to evaluate the effects of interventions.1216 A recent study of the effects of hippotherapy on people with CP found that while both users and physiotherapists agreed that improvements in function were the most important effects of hippotherapy, the effects were not routinely or systematically assessed.1719 This was at least in part, because therapists were unsure about the available OMs' validity, reliability, and clinical utility.19

Three previous systematic reviews on OMs of function/activity for children with CP have been undertaken. Whereas they offered valuable perspectives on the subject, 2 were narrative and did not follow contemporary search strategies; the process of critically appraising the literature was also not specified.4,20 The third and much more recent systematic review21 provides a good audit trail of the search strategy, but poor transparency in terms of the quality assessment of included papers. Its other major flaw is the inclusion of studies that used samples that included a minority of children with CP. Of the 29 studies reviewed, only a minority had samples consisting exclusively of children with CP, with 3 of the included studies having samples of less than 10% of children with CP. This means that of a total of 4565 children participating in the included studies, only 2235 (49%) had CP.21 In view of the fact that the systematic review of Harvey et al21 specifically focused on psychometric properties of measures of activity limitation for children with CP, the minority of children with CP included means that its findings need to be interpreted with caution.

The lack of previous rigorous systematic reviews in this area and the publication of new studies in the last few years call for a new systematic review on the topic to rectify previous methodological errors and omissions. Our aim was to identify valid, reliable, and clinically practical measures of function/activity in children with CP.


Search Strategy

Initially a search was undertaken through the UK National Health Service National Library for Health22 to ensure that there were no ongoing or published systematic reviews in this area. Computerized bibliographic databases (MEDLINE, CINAHL, AMED, PEDro, the Cochrane Library, and ScienceDirect) were then searched via EBSCO Host to identify papers. Keywords were used alone or in combination, using MeSH terms where available. A list of the exact terms used with CINAHL is given in Table 1. Search terms were adapted for different databases.

List of Search Terms Used With CINAHL

Outcome measures were identified and their common abbreviations were then used as terms for further searches of electronic databases. Finally, the references of identified studies were hand-searched for relevant articles to further reduce the possibility of missed studies. The database and subsequent searches as outlined earlier were performed between June 2008 and June 2010. The searches were repeated in February 2011 to ensure that relevant articles published since 2010 were included, but none were identified.

Inclusion/Exclusion Criteria

In line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations,23 inclusion criteria were set for both OMs and papers. To be included in the review, OMs must

  1. have been used in psychometric studies with children with CP;
  2. have evaluative, not only discriminative properties;
  3. be feasible for use by a competent physiotherapist without extensive specialist training, in a nonspecialist environment and with basic equipment; and
  4. assess aspects of activity/function as defined in the ICF.6

To be included in the review, articles must

  1. demonstrate content/face validity;
  2. have a sample made up of at least 75% children with CP; and
  3. be primary, original research published in peer-reviewed journals in English (because of lack of resources available for translation).

Measures of individualized performance (eg, Goal Attainment Scale, Canadian Occupational Performance Measure) were excluded because of the absence of objective criteria for scoring and the difficulty in ascertaining their reliability.24

Data Extraction and Assessment of Methodological Quality/Risk of Bias

The data extraction template used in this study is shown in Figure 1. Neither the Cochrane Library25 nor the PRISMA group23,26 offers suggestions as to how studies on OMs should be appraised for risk of bias. In this study, all papers, which met the inclusion criteria, were subjected to a systematic critical appraisal, which was developed and piloted by HB. It was adapted from existing appraisal frameworks for OMs developed by Jerosch-Herold2 and Terwee et al27 (for more detailed information, please see the Appendix, Supplement Digital Content 1, at https://links.lww.com/PPT/A21). After piloting the appraisal tool, it was used by both authors to examine the psychometric properties of OMs purported to be examined and any risk of bias, including study design and statistical tests used. The critical appraisal tool is very specific and detailed and provides very clear criteria for assessment (see the Appendix at https://links.lww.com/PPT/A21). The papers were independently appraised by both authors. Because of the tool's detailed specification of appraisal criteria, there was overwhelming agreement on individual items between authors. In the case of differences between authors, these were discussed and a consensus decision was reached. Unlike other systematic reviews,28,29 we did not attempt to give papers a numerical quality score, as an equal ranking of individual criteria could not be assumed or justified, and the Cochrane Collaboration30 explicitly discourages the use of numerical weighing scales. The results of the appraisal of the studies were tabulated using a table based on RevMan 5.25

Fig. 1:
Data-extraction template.

Data were synthesized in 3 ways. First, included studies were summarized in terms of the psychometric properties they tested, participants, and sampling. Then, to provide an overview of the OMs, OMs' key characteristics in terms of target group, method of administration, ICF domains or categories covered, description of OM, equipment needed, and time to complete, were summarized. Finally, each OM's psychometric properties investigated in each of the included papers were detailed in terms of their statistical test results and the sample on which they were tested (children's age and Gross Motor Function Classification System [GMFCS] levels).


Selection of Papers

Of 1533 studies initially identified through the database searches and 56 studies identified through hand-searching of references and searching databases for specific OMs, 371 studies were identified as potentially relevant based on their citations using the search strategies outlined earlier. When their abstracts were screened, 302 were excluded as they did not meet the inclusion criteria. Of the remaining 69 full-text papers screened, 62 were excluded as they turned out either not to meet the inclusion criteria or not to be rigorous or far enough in their development for the OMs to be recommended for clinical practice at this point.9,31,32 The PRISMA diagram26 in Figure 2 gives an overview of the selection process. The remaining 7 studies on 6 OMs met the inclusion criteria and underwent the data extraction and critical appraisal process. The characteristics of the included papers in terms of psychometric properties tested and participants are reported in Table 2. Because of differences in study methodology, a meta-analysis of results was not possible.

Fig. 2:
PRISMA diagram of study selection.
Characteristics of Studies in Terms of Psychometric Properties Tested, Participants, and Sampling

Overview of Included Papers and Outcome Measures

The tabulated critical appraisal of the 7 included studies and a visual representation of their strengths and weaknesses are presented in Table 3. In the 7 studies identified for inclusion in the review, 6 evaluative OMs were specifically examined. These OMs were the 88- and the 66-item versions of the Gross Motor Function Measure (GMFM-88 and GMFM-66),7 the Pediatric Evaluation of Disability Inventory (PEDI),10 the Paediatric Outcomes Data Collection Instrument (PODCI),39 the Functional Independence Measure for Children (WeeFIM),40 and the Gillette Functional Assessment Questionnaire (FAQ).41

Results of Critical Appraisal of Core Studiesa

All the included studies used samples consisting exclusively of children with CP. There was a wide range of sample sizes in the studies (n ranged from 41 to 562). The proportion of female to male participants across studies was relatively even. The age ranges of samples were detailed in all the studies. Inclusion/exclusion criteria were provided, and clear, standardized protocols were described or referenced in all studies. All studies also provided information on their participants' severity of CP.

Clinical Utility of Outcome Measures

The GMFM-88 and -66 are condition-specific instruments designed to assess gross motor function in children with CP.7 The PEDI and the WeeFIM were designed for chronically ill children with disabilities. The PODCI was designed to assess orthopedic problems, and the FAQ to assess walking ability across different terrains in children with chronic neuromuscular conditions. The PEDI, WeeFIM, and PODCI are generic questionnaires that measure the effect of a condition on a person's functioning, health, and/or self-care in a range of environments. Administration time varies across OMs from 15 to 45 minutes and over, and assessor training in some form is required for all of them. Two require specialized scoring software.

All OMs in the included papers are feasible for use in a nonspecialist environment if basic resources are available, and all included studies had numerical results with sufficient significance to provide clinically relevant conclusions.

Psychometric Properties of Outcome Measures

Completion rates can provide information on the relevance of questionnaires to the people who complete them.42 McCarthy et al35 were the only authors to examine completion rates. They reported the highest completion rates for the GMFM-88 and the PEDI, which confirms the relevance of these OMs to the functional ability of children with CP. In contrast, 3 of PODCI's 7 subscales had high-missing scores (11%–35% of the sample), which suggests that many of the PODCI items are irrelevant for children with CP and, therefore, inappropriate for this client group.

Reliability of measures was examined in only 2 of the 7 included studies. Both the GMFM-88 and -66 were shown to have excellent intratester reliability (intraclass correlation coefficient = 0.9944 and 0.9932, respectively).37 Cronbach α coefficients of 0.9 or greater demonstrate excellent internal consistency for both the GMFM-88 and the PEDI (for all scales).35

Responsiveness to change is a fundamental aspect of validity.43,44 Responsiveness to change, when tested over 18 months and calculated as effect size and standardized response mean, was excellent for all the PEDI scales (≥0.8 for age range 2–7 years), and excellent (0.91) for GMFM-88 total scores for children younger than 48 months, but only good (0.6) for children older than 48 months.38 The area under the curve (receiver operating characteristic [ROC]) is another method of calculating the effect of an intervention.45 Bagley et al33 found the GMFM-66 to be the most sensitive OM she tested in terms of its ability to discriminate between GMFCS I and II and II and III, with the area under the curve of 0.903 and 0.935, respectively. Oeffinger et al36 found that the GMFM-66 was the most responsive OM in terms of minimally clinically important difference, second only to oxygen cost (instrumented gait analysis), and followed by the GMFM-88 dimensions D and E. They found the WeeFIM to be less responsive across all GMFCS levels than the GMFM-88 dimensions D + E. This concurs with findings of Bagley et al33 as outlined earlier.

Russell et al7 found the GMFM-66 sensitive over 12 months in children younger than 5 years (P < .0001); they also found responsiveness to change to be greater in children younger than 5 years with higher functional ability (GMFCS levels I and II).

Ceiling effects occur when a substantial number of study participants achieve the highest scores available. This occurs with OMs that only evaluate participants in the low to moderate ranges of function and means that an OM is not sensitive to detect change at the higher end of function.46 Such a lack of responsiveness to change clearly affects an OM's validity.43,44 The same principle, but at the opposite end of the spectrum, occurs with floor effects.

McCarthy et al35 found very low ceiling effects and no floor effects with the PEDI, but moderately high ceiling and floor effects (18%–33%) with the GMFM-88, and similar floor, but even higher ceiling effects (up to 43%) with the PODCI. Vos-Vromans et al38 also found no ceiling effects for the PEDI, but considerable ceiling effects for the GMFM-88, particularly in children 2 years and older. Bagley et al33 demonstrated moderate to high ceiling effects of 12% to 56% in 2 of the 3 WeeFIM domains across GMFCS levels I to III, indicating that the WeeFIM is unable to detect change in these areas. This is further supported by these authors' finding that effect size indexes for the WeeFIM were only moderate or low by GMFCS comparisons between levels I and II, and II and III (0.0-0.5), except in the mobility domain between levels II and III (1.2). Bagley et al33 identified that the FAQ, too, has a high ceiling effect (45%) for children in GMFCS level I. As the FAQ assesses walking ability, which can be expected to be present only in children with GMFCS I to III, this indicates a lack of responsiveness to change in the key target population.

Because of a high number of missing scores, considerable ceiling effects, and/or otherwise inadequate responsiveness of the PODCI, Gillette FAQ, and WeeFIM as detailed earlier, we can only consider the GMFM-88 and -66 and the PEDI as valid measures of function for use with children with CP. Table 4 gives an overview of these OMs and their properties. It outlines the administration method, ICF domains covered,6 a short description of the OM, the equipment needed, and the time taken for administration of the tool. Table 5 provides the data synthesis of all the relevant psychometric property results with the age ranges and samples, for which they were validated in the papers reviewed in this study for the GMFM-88 and -66 and the PEDI.

Key Characteristics of the 3 Most Valid and Reliable OMs Reported in the Included Studies
Synthesis of Reported Psychometric Properties


Differences Between Different Studies' Findings

In spite of some broad agreement between authors, there are differences between different studies' findings. Russell et al37 suggest that after the age of 5 years, changes in a child's motor abilities are more related to developing and refining motor functions in specific environments, rather than the development of basic gross motor skills. They argue that GMFM is, therefore, not able to detect these changes and suggest that the PEDI may be a better tool to assess children 5 years and older.37 This would explain the finding of several studies3,38 that children younger than 5 years demonstrated greater change in gross motor function than older children. In fact, Russell et al37 and Rosenbaum et al47 found that there was no improvement at all in GMFM scores in children older than 5 years. However, other authors33,36 found the GMFM-66 to be sensitive in children between 5 and 18 years.

There is some disagreement between authors as to the responsiveness of the GMFM-88 versus -66. Wang and Yang3 found both GMFM versions responsive to children with GMFCS levels IV and V over a period of only 3.5 months, and other authors identified the GFMF-66 as the most sensitive OM they tested.33,36 However, this conflicts with the findings of Lundkvist Josenby et al,34 who report that whereas the GMFM-66 was as responsive as the GMFM-88 for identifying change at 3 and 5 years postintervention, the GMFM-66 was less sensitive at 12 and 18 months postintervention. The reason for this discrepancy might be found in the specific context of the studies and in the fact that Lundkvist Josenby et al34 worked with the smallest sample (n = 41). Wang and Yang3 found both GMFM-88 and -66 capable of detecting clinically meaningful change. However, they reported that the specificity of the GMFM-66 was 21% greater than that of the GMFM-88, meaning that the GMFM-88 is more likely to detect false positives than the GMFM-66. The latter may have further contributed to the finding of Lundkvist Josenby et al34 that the GMFM-88 was more sensitive than the GMFM-66.

Clinical Considerations

A disadvantage the GMFM-66 and -88 share is that, as unidimensional OMs, they only examine gross motor function. It has been argued that they need to be complemented by other OMs to provide a comprehensive picture of a child's functioning/activity and participation.34,37 Furthermore, as measures of capacity, they do not give insight into a child's actual motor behaviors in his or her normal environment. Instead, they measure a child's single “performance” in a clinically controlled test setting.

The PEDI (a parent-reported OM) examines motor and self-care function, as well as participation (in its social function domain). The PEDI, therefore, reflects much more closely the ICF domains of activity and participation. Its clinical relevance is further supported by evidence that motor skills are not necessarily representative of overall functional improvements following therapeutic interventions.48

Holsbeeke et al49 warn that it is important to distinguish between the constructs of capacity, capability, and performance in OMs. Capacity is an individual's ability to carry out certain tasks when observed in a standardized environment. In contrast, capability describes what that individual can do in his or her own environment, whereas performance is what that individual actually does in his or her environment on a daily basis.49 Clearly, the results of measures of function for children with CP need to be interpreted with these distinctions in mind. For example, the GMFMs assess gait in the test environment on a smooth floor. This is not necessarily representative of a child's gait in a nonclinical environment for 2 reasons. A child may feel anxious in a clinical environment and may, therefore, “perform” less well than in a nonclinical environment (or not at all). Also possible, a child may find walking easier on a level surface than outside over rough terrain and do better than usual.

The PEDI is an OM that assesses the effect of a condition on a child in terms of his or her function across different domains, that is, how it affects his or her everyday life. On the basis of the parent's knowledge of the child, it reflects a child's function across different dimensions; that is, it assesses performance.35 This is more functionally relevant to children with CP. As the PEDI is a multidimensional OM, its mobility scale is a lot less detailed than the GMFM scales. In spite of this, its responsiveness to motor changes is well-established.35,38

As for the GMFMs, by the authors' own admission,37 children with intermediate motor ability will have greatest possible change with the GMFM-88. In other words, it has moderate ceiling and floor effects, which have been confirmed by other authors.35 This adversely affects the OM's validity. Irrespective of their motor abilities, children still must complete the whole GMFM-88 to arrive at a valid score. This can be particularly challenging with younger children with CP or those with a learning difficulty, as a lot of the motor tasks have no functional relevance, and convincing the child to perform them is likely to be more difficult. Therefore, in clinical practice to save time, the GMFM-88 is often administered in part. However, the reliability and validity of scoring individual dimensions are much decreased by only scoring part or individual sections of the GMFM-88.3,8

When using the GMFM-66, on the contrary, an estimate of a child's motor ability can be made even if there are missing items, as items can be rated as “not tested” rather than given a zero score, on the basis of the GMFM-66's interval-level scaling.37,50 The latter also improves interpretability, as an equal rise in score corresponds to an equal rise in ability. It also allows therapists to predict what the next item that might be suitable as a therapeutic goal may be as each item is arranged in ascending difficulty.37 A further advantage of the GMFM-66 is that it has 22 fewer items than the GMFM-88 allowing for faster completion of the test. The disadvantage of the GMFM-66 is that it has a floor effect in children with low motor ability and ceiling effects in children older than 5 years.37,47 This means that it is much less useful when scoring children of low-motor ability and those older than 5 years.

sThe PEDI was designed with its scales deliberately skewed toward the lower end of ability in terms of functional skills.35 This means that it is useful in conjunction with the GMFM as it assesses some of the emerging motor skills that cannot be examined using the GMFM. However, despite the PEDI's ability to capture function at the lower end of the scale, it has no ceiling effects.35,38 On the contrary, because of its ability to assess the effect of a child's motor skills on function in his or her own environment, the PEDI emerges as being highly relevant for the age range where the GMFMs show least change (≥5 years). This is supported by Russell et al,37,47 who, as mentioned earlier, suggest that after the age of 5 years, changes in a child's motor abilities are related to developing and refining motor functions in specific environments, rather than the development of basic gross motor skills, as assessed by the GMFM. They suggest, therefore, that the PEDI is likely to be a better tool to assess children aged 5 years and older.37

Recent Developments

In a recent paper, Haley et al51 outline the development of a new instrument to succeed the current PEDI. The authors report that in spite of the addition of some new items, which extends the functional assessment and makes the new tool appropriate for children up to the age of 15 years, something that is to be much welcomed with the current bias of OMs for children with CP toward the younger age range. The introduction of Item Response Theory methods and computer adaptive testing (CAT) means that the new OM is less time consuming to administer than the current PEDI and will result in an interval measure. The current dichotomous scale in the self-care, mobility, and social function sections will be replaced by a 4-point scale to reflect difficulty. The new OM will include illustrations of all mobility and self-care items, and a new “Responsibility” section will replace the current Caregiver Assistance section, to reflect the bigger age range to which the new tool will apply. For more detailed information, readers are referred to Haley's original article.51 The new electronic tool PEDI-CAT is expected to be available over the Internet in spring 2011 (S. Haley, e-mail communication, November 22, 2010).


While a range of OMs has been used to assess function in children with CP, not all of these have been validated for this user group. In the 7 studies identified for inclusion in this review, 6 evaluative OMs were specifically examined. Of these, only 3 OMs emerge as potentially appropriate for testing function in children with CP, the GMFM-88, GMFM-66, and PEDI. However, the GMFM-88 has weaknesses, namely its long completion time and floor and ceiling effects, which caution against its use in clinical practice. A weakness the GMFM-88 and -66 share is their unidimensionality; they only test gross motor capacity, that is, the gross motor ability a child demonstrates in a controlled environment, and not performance in other functional domains, including a child's actual motor and general activity behavior in his or her own daily environment.

In spite of its wider focus, the PEDI demonstrates excellent responsiveness to change across domains, including motor function. It is a measure of performance, that is, a child's actual activity in his or her everyday environment, and is, therefore, much more closely aligned with the ICF's activity and participation domains.6 A new version of the PEDI, the PEDI-CAT, is currently being developed to bring it up-to-date with recent trends in outcome measurement and to extend its range up to the age of 15 years. The latter is to be particularly welcomed with the current bias of validated measures of function for individuals with CP toward the younger age range.


1. Chartered Society of Physiotherapy. Outcome Measures. London: Chartered Society of Physiotherapy; 2008.
2. Jerosch-Herold C. An evidence-based approach to choosing outcome measures: a checklist for the critical appraisal of validity, reliability and responsiveness studies. Br J Occup Ther. 2005;68(8):347–353.
3. Wang H, Yang Y. Evaluating the responsiveness of 2 versions of the Gross Motor Function Measure for children with cerebral palsy. Arch Phys Med Rehabil. 2006;87(1):51–56.
4. Ketelaar M, Vermeer A, Helders P. Functional motor abilities of children with cerebral palsy: a systematic literature review of assessment measures. Clin Rehabil. 1998;12(5):369–380.
5. Rosenbaum P, Paneth N, Leviton A, et al. The definition and classification of cerebral palsy. Devel Med Child Neurol. 2007;49(109):1–44.
6. World Health Organization. International Classification of Functioning, Disability and Health (ICF). Geneva, Switzerland: World Health Organization; 2001.
7. Russell D, Rosenbaum P, Avery L, et al. Gross Motor Function Measure (GMFM-66 & GMFM-88) User's Manual. Clinics in Developmental Medicine. London: MacKeith Press; 2002:159.
8. Russell D, Rosenbaum P, Cadman D, et al. The Gross Motor Function Measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol. 1989;31(3):341–352.
9. Pountney T, Cheek L, Green E, et al. Content and criterion validation of the Chailey Levels of Ability. Physiotherapy. 1999;85(8):410–416.
10. Haley S, Coster W, Ludlow L, et al. Pediatric Evaluation of Disability Inventory. Boston, MA: New England Medical Centre; 1992.
11. Msall M, DiGaudio K, Rogers B, et al. The Functional Independence Measure for Children (WeeFIM). Conceptual basis and pilot use in children with developmental disabilities. Clin Pediatr. 1994;88:421–430.
12. Abrams D, Davidson M, Harrick J, et al. Monitoring the change: current trends in outcome measure usage in physiotherapy. Man Ther. 2006;11(1):46–53.
13. Duckworth M. Outcome measurement selection and typology. Physiotherapy. 1999;85(1):21–27.
14. Haigh R, Tennant A, Biering-Sorensen F, et al. The use of outcome measures in physical medicine and rehabilitation within Europe. J Rehabil Med. 2001;33(6):273–278.
15. Maher C, Williams M. Factors influencing the use of outcome measures in physiotherapy management of lung transplant patients in Australia and New Zealand. Physiother Theory Pract. 2005;21(4):201–217.
16. Van Peppen R, Maissan F, Van Genderen F, et al. Outcome measures in physiotherapy management of patients with stroke: a survey into self-reported use, and barriers to and facilitators for use. Physiother Res Int. 2008;13(4):255–270.
17. Debuse D, Chandler C, Gibb C. An exploration of German and British physiotherapists' views on the effects of hippotherapy and their measurement. Physiother Theory Pract. 2005;21(4):219–242.
18. Debuse D, Gibb C, Chandler C. The effects of hippotherapy on people with cerebral palsy from the users' perspective: a qualitative study. Physiother Theory Pract. 2009;25(4):174–192.
19. Debuse D. An Exploration of the Effects of Hippotherapy on People With Cerebral Palsy. Newcastle upon Tyne, UK: Northumbria University; 2006.
20. Young N, Wright J. Measuring pediatric physical function. J Pediatr Orthop. 1995;15:244–253.
21. Harvey A, Robin J, Morris M, et al. A systematic review of measures of activity limitation for children with cerebral palsy. Devel Med Child Neurol. 2008;50(3):190–198.
22. National Health Service. National Health Service National Library for Health. http://www.library.nhs.uk/Default.aspx. Published 2008. Accessed
23. Liberati A, Altman D, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ ONLINE. 2009. doi: 10.1136/bmj.b2700.
24. Cusick A, McIntyre S, Novak I, et al. A comparison of goal attainment scaling and the Canadian occupational performance measure for paediatric rehabilitation research. Pediatr Rehabil. 2006;9(2):149–157.
25. The Cochrane Collaboration 2008 The Cochrane Library Database of Systematic Reviews. http://www3.interscience.wiley.com/cgi-bin/mrwhome/106568753/HOME?CRETRY=1&SRETRY=0. Accessed May 7, 2008.
26. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ. 2009;339:332–336.
27. Terwee C, Bot S, de Boer M, et al. Quality criteria proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007;60(1):34–42.
28. Greaves S, Imms C, Dodd K, et al. Assessing bimanual performance in young children with hemiplegic cerebral palsy: a systematic review. Devel Med Child Neurol. 2010;52:413–421.
29. Jeglinski I, Surakka J, Carlberg E, et al. Evidence on physiotherapeutic interventions for adults with cerebral palsy is sparse. A systematic review. Clin Rehabil. 2010;24(9):771–788.
30. The Cochrane Collaboration 2009 Cochrane Handbook for Systematic Reviews of Interventions. http://www.cochrane.org/training/cochrane-handbook. Accessed May 10, 2010.
31. McDowell B, Kerr C, Parkes J, et al. Validity of a 1 minute walk test for children with cerebral palsy. Devel Med Child Neurol. 2005;47:744–748.
32. Haley S, Fragala-Pinkham M, Dumasd H, et al. Evaluation of an item bank for a computerized adaptive test of activity in children with cerebral palsy. Phys Ther. 2009;89(6):589–600.
33. Bagley A, Gorton G, Oeffinger D, et al. Outcome assessments in children with cerebral palsy, part II: discriminatory ability of outcome tools. Devel Med Child Neurol. 2007;49:181–186.
34. Lundkvist Josenby A, Jarnlo G, Gimmesson C, et al. Longitudinal construct validity of the GMFM-88 total score and goal total score and the GMFM-66 score in a 5-year follow-up study. Phys Ther. 2009;89(4):342–350.
35. McCarthy M, Silberstein C, Atkins E, et al. Comparing reliability and validity of pediatric instruments for measuring health and well-being of children with spastic cerebral palsy. Devel Med Child Neurol. 2002;44(7):265–269.
36. Oeffinger D, Bagley A, Rogers S, et al. Outcome tools used for ambulatory children with cerebral palsy: responsiveness and minimum clinically important differences. Devel Med Child Neurol. 2008;50:918–925.
37. Russell D, Avery L, Rosenbaum P, et al. Improved scaling of the Gross Motor Function Measure for children with cerebral palsy: evidence for reliability and validity. Phys Ther. 2000;80(9):873–885.
38. Vos-Vromans D, Ketelaar M, Gorter J. Responsiveness of evaluative measures for children with cerebral palsy: the Gross Motor Function Measure and the pediatric evaluation of disability inventory. Disabil Rehabil. 2005;27(20):1245–1252.
39. Daltroy LH, Liang MH, Fossel AH, et al. The POSNA pediatric musculoskeletal functional health questionnaire: report on reliability, validity, and sensitivity to change. J Pediatr Orthop. 1998;18(5):561–571.
40. McCabe M, Granger C. Content validity of a pediatric functional independence measure. Appl Nurs Res. 1990;3:120–122.
41. Novacheck T, Stout J, Tervo R. Reliability and validity of the Gillette Functional Assessment Questionnaire as an outcome measure in children with walking disabilities. J Pediatr Orthop. 2000;20:75–81.
42. Oppenheim A. Questionnaire Design, Interviewing and Attitude Measurement. London: Pinter; 1992.
43. Hays R, Hadorn D. Responsiveness to change: an aspect of validity, not a separate entity. Qual Life Res. 1992;1(1):73–75.
44. Sim J, Wright C. Research in Health Care: Concepts, Designs and Methods. Cheltenham: Stanley Thornes; 2000.
45. Pham B, Cranney A, Boers M, et al. Validity of area-under-the-curve analysis to summarize effect in rheumatoid arthritis clinical trials. J Rheumatol. 1999;26(3):712–716.
46. Austin P, Brunner L. Type I error inflation in the presence of a ceiling effect. Am Statist. 2003;57(2):97–104.
47. Rosenbaum P, Walter S, Hanna S, et al. Prognosis for gross motor function in cerebral palsy. JAMA. 2002;228(11):1357–1363.
48. Awaad Y, Tayem H, Munoz S, et al. Functional assessment following intrathecal baclofen therapy in children with spastic cerebral palsy. J Child Neurol. 2003;18(1):26–34.
49. Holsbeeke L, Ketelaar M, Schoemaker M, et al. Capacity, capability and performance: different constructs or three of a kind? Arch Phys Med Rehabil. 2009;90(5):849–855.
50. Avery L, Russell D, Raina P, et al. Rasch analysis of the Gross Motor Function Measure: validating the assumptions of the Rasch model to create an interval-level measure. Arch Phys Med Rehabil. 2003;84(5):697–705.
51. Haley S, Coster W, Kao Y-C, et al. Lessons from use of the pediatric evaluation of disability inventory: where do we go from here? Pediatr Phys Ther. 2010;22(1):69–75.
52. Centre for Evidence-Based Medicine 2010 critical appraisal skills programme. http://www.sph.nhs.uk/what-we-do/public-health-workforce/resources/critical-appraisals-skills-programme. Accessed March 4, 2011.
    53. Law M, Stewart D, Pollock N, et al. Critical Review From—Quantitative Studies. 1998. http://www.srs-mcmaster.ca/Portals/20/pdf/ebp/quanreview.pdf. Accessed March 4, 2011.
      54. Law M, Stewart D, Pollock N, et al. Critical Review Form—Qualitative Studies. 1998. http://www.srs-mcmaster.ca/Portals/20/pdf/ebp/qualreview.pdf. Accessed March 4, 2011.
        55. Scottish Intercollegiate Guidelines Network 2010 Critical appraisal: notes and checklists. http://www.sign.ac.uk/methodology/checklists.html. Accessed March 4, 2011.
          56. Bot S, Terwee C, van der Windt D, et al. Clinimetric evaluation of shoulder disability questionnaires: a systematic review of the literature. Ann Rheum Dis. 2004;63:335–341.
          57. De Boer M, Moll A, De Vet H, et al. Psychometric properties of vision-related quality of life questionnaires: a systematic review. Ophtalmic Physiol Opt. 2004;24:257–273.
            58. Portney L, Watkins M. Foundations of Clinical Research: Applications to Practice. Upper Saddle River, NJ: Prentice Hall; 2000.

              activities of daily living; cerebral palsy; child; motor activity; outcomes assessment; psychometrics; severity of illness index; systematic review

              Supplemental Digital Content

              Copyright © 2011 Academy of Pediatric Physical Therapy of the American Physical Therapy Association