Secondary Logo

Journal Logo


Linking the Pediatric Evaluation of Disability Inventory-Computer Adaptive Test (PEDI-CAT) to the International Classification of Function

Thompson, Sharon V. PT, DHS, PCS; Cech, Donna J. PT, DHS, PCS; Cahill, Susan M. PhD, OTR/L, FAOTA; Krzak, Joseph J. PT, PhD, PCS

Author Information
doi: 10.1097/PEP.0000000000000483
  • Free


Individuals in rehabilitation programs strive to attain independence in functional abilities so that they can participate in meaningful life activities and fulfill life roles. The World Health Organization's biopsychosocial model of health and disability, the International Classification of Functioning, Disability, and Health (ICF), defines a person's participation as the ability to be involved in life situations.1 Within this model, functional abilities are captured among the components of activity and participation, which reflect the individual's ability to execute a task and the person's involvement in life situations. The ICF model has defined 9 chapters of activity and participation that comprehensively describes the Activity and Participation components: (1) Learning and Applying Knowledge, (2) General Tasks and Demands, (3) Communication, (4) Mobility, (5) Self-care, (6) Domestic Life, (7) Interpersonal Interactions and Relationships, (8) Major Life Areas, and (9) Community, Social, and Civic Life. As rehabilitation professionals strive to support a person's functional ability and participation, it is important to consider all aspects of activity and participation, which have been defined by the ICF. A version of the ICF, the International Classification of Functioning, Disability, and Health–Children and Youth (ICF-CY) was subsequently developed to include meaningful activities for children and youth.2 The ICF-CY has now been integrated into the ICF.3

Consistent with the ICF, federal laws such as the Individuals with Disabilities Education Act (IDEA), Parts B and C, support the importance of participation among children with disabilities within their natural environment including home, school, and in the community.4 Physical therapists working in early intervention and educational settings provide services in natural environments and support the child's ability to participate to the greatest extent possible in family, school, and community situations.

Measurement of a child's functional ability is an important element in developing a physical therapy plan of care and assisting children in actively participating in meaningful life activities to the greatest extent possible. It is recommended that therapists use outcome measures with valid psychometric properties to document a child's functional status and progress. The tool should be used to measure the breadth of age-appropriate, meaningful functional activities for children. The ability of rehabilitation outcome measures to comprehensively measure all aspects of activity and participation has been assessed by matching the outcome measure content to the numeric coding system established within the classification system of the ICF. A model using linking rules has been used to compare both adult and pediatric functional outcome measure content to the ICF.5–9 The linking rule model defines specific methods to identify meaningful concepts in outcome assessment tools and match them to ICF numeric codes, using the ICF online browser.10 Several pediatric outcome measures have been identified that incorporate content related to functional ability and participation including the Pediatric Evaluation of Disability Index (PEDI), the Pediatric Outcomes Data Collection Instrument, the pediatric version of the Functional Independence Measure (Wee-FIM), and the Children's Assessment of Participation and Enjoyment.5,6

The Pediatric Evaluation of Disability Inventory-Computer Adaptive Test (PEDI-CAT) is a standardized, norm-referenced outcome measure developed in 2011, as a revision and expansion of the PEDI, to examine the functional performance of children and youth from birth through 20 years of age.11 It includes 276 functional activities across 4 domains: daily activity, mobility, social/cognitive function, and responsibility. The PEDI-CAT uses a computerized adaptive testing method, which individualizes the assessment to each child, decreasing test burden and improving test precision.11 This process reduces administration time when compared with the original PEDI, and maintains the relevance of the outcome assessment measure. The PEDI-CAT is reliable, valid, and responsive in rehabilitation settings.12,13 The PEDI-CAT conceptual model was designed to be consistent with the ICF and ICF-CY framework.11 Although the original PEDI was representative of the ICF Activities and Participation components,5,14 there are currently no published reports describing the extent to which the PEDI-CAT addresses all aspects of activity and participation in the ICF.

The purpose of this study was to examine the degree to which the PEDI-CAT items comprehensively reflect the activity and participation components of the ICF, using the ICF linking rules established by previous researchers.7,8 It was hypothesized that all 9 chapters of the Activities and Participation component of the ICF would be represented in the PEDI-CAT. The results could provide further insight into the PEDI-CAT's content validity and add to the ongoing work on activity and participation in pediatric rehabilitation.



Three experienced physical therapists, with board-certified clinical specialties in Pediatric Physical Therapy by the American Board of Physical Therapist Specialties, participated in this study, 2 as the primary raters and the third to assist with establishing consensus.

Training Module

All raters participated in a 4-hour training module prior to the start of the study. This training included the following activities: completing an online ICF overview available through the American Physical Therapy Association Web site15; reviewing assigned literature on ICF linking rules7,8; reviewing literature related to content analysis of pediatric outcome measures5,14; and practice applying linking rules with examples. Raters completed a trial application of the linking rules to a pediatric outcome measure, the Pediatric Outcomes Data Collection Instrument. Results from the trial application were similar to previously reported findings5 demonstrating competency of the raters.

Linking Process and the PEDI-CAT

Two raters independently completed the ICF linking process by matching the 276 PEDI-CAT items11 to ICF codes using the online ICF browser10 and available ICF updates.16 Raters independently identified meaningful concepts among the PEDI-CAT items. Meaningful concepts were defined as words, phrases, or sentences within a PEDI-CAT item that provided subject matter relevant to the ICF codes. Raters then used the revised ICF linking rules as defined by Cieza et al8 to match individual meaningful concepts to individual ICF codes. Following the independent matching process, a third reviewer was included in the discussion to resolve disagreements between the 2 primary raters and reach consensus in (1) identifying meaningful concepts among items, and (2) matching the most precise ICF code(s) to each meaningful concept. Meaningful concepts that did not match to ICF codes were classified as either “not definable” (ND) or “not covered” (NC). If a meaningful concept of an item was explained by an example or pictorial illustration, both the concept and the example/illustration were considered for matching.

Data Analysis

To evaluate interrater reliability of the linking process, percent agreement, as well as Cohen's κ, the standard error (SE), and the 95% confidence interval (95% CI) were calculated between the 2 independent raters.17 Multiple interpretation schemes of Cohen's κ exist in the literature. As a result, the most common interpretation schemes and criteria were used to interpret interrater reliability including (1) Landis' classification: less than 0: less than chance agreement, 0.01 to 0.20: slight agreement, 0.21 to 0.40: fair agreement, 0.41 to 0.60: moderate agreement, 0.61 to 0.80: substantial agreement, and 0.81 to 0.99: almost perfect agreement,18 and (2) Altman's classification: less than 0.20: poor, 0.21 to 0.40: fair, 0.41 to 0.60: moderate, 0.61 to 0.80: good, and 0.81 to 1.00: very good.19 Descriptive statistics, including frequencies and percentages, of the ICF codes used among the PEDI-CAT domains were calculated to test the hypothesis that all 9 chapters of the Activities and Participation component of the ICF were represented in the PEDI-CAT.


The 276 items in the PEDI-CAT had 329 meaningful concepts (Table 1). Two-hundred twenty-six items had 1 meaningful concept, 47 items had 2 meaningful concepts, and 3 items had 3 meaningful concepts. The 329 meaningful concepts were linked to 363 ICF codes.

TABLE 1 - Meaningful Concepts Identified Among PEDI-CAT Domains
PEDI-CAT Domains Items, n Meaningful Concepts, n
Daily Activities 68 73
Mobility 97 135
Social-Cognitive 60 61
Responsibility 51 60
Total 276 329
Abbreviation: PEDI-CAT, Pediatric Evaluation of Disability Inventory-Computer Adaptive Test.

The agreement between the 2 raters was 87% in matching meaningful concepts to the ICF codes. The greatest agreement between the 2 primary raters occurred in the Daily Activities (91% agreement) and Mobility (94% agreement) domains. There was less agreement between the 2 primary raters in the Social-Cognitive (85% agreement) and Responsibility (78% agreement) domains of the PEDI-CAT. Interrater reliability of the overall matching process using Cohen's κ demonstrated very good to almost perfect agreement between the 2 primary raters (κ = 0.84; SE = 0.13; 95% CI = 0.57-1.10; P < .001).18,19 Among the 3 raters, consensus was reached for each item in the PEDI-CAT. Disagreement between the 2 primary raters occurred 36 (13%) times, requiring further discussion with the third rater to reach consensus.

All 9 chapters in the Activity and Participation component of the ICF were represented in the PEDI-CAT (358/363 codes, 99%). Table 2 provides the frequencies and percentages of how often the ICF chapters were represented among the PEDI-CAT domains following the consensus process. Two items were considered ND, and 3 items/meaningful concepts were linked to Body Function codes. The highest frequency of representation in the Activities and Participation components of the ICF was in the D4: Mobility chapter (157 codes; 43%) followed by the D5: Self-care chapter (73 codes; 20%). The lowest frequency of representation in Activities and Participation was in the D8: Major Life Areas chapter (5 codes; <1%).

TABLE 2 - Frequencies and Percentages of ICF Code Representation Among PEDI-CAT Domains
ICF Components and Chapters Daily Activities PEDI-CAT Domains Total, n (%)
Mobility Social-Cognitive Responsibility
Body Structure and Function (B)
Functions of the Digestive, Metabolic and Endocrine Systems (B5) 1 0 0 0 1 (<1)
Neuromusculoskeletal and Movement-Related Functions (B7) 0 2 0 0 2 (<1)
Activity and Participation (D)
Learning and Applying Knowledge (D1) 5 0 9 3 17 (5)
General Tasks and Demands (D2) 0 1 6 9 16 (4)
Communication (D3) 0 0 31 0 31 (9)
Mobility (D4) 18 137 0 2 157 (43)
Self-care (D5) 45 1 3 24 73 (20)
Domestic Life (D6) 11 0 0 14 25 (7)
Interpersonal Interactions and Relationships (D7) 0 0 14 0 14 (4)
Major Life Areas (D8) 0 0 2 3 5 (<1)
Community, Social and Civic Life (D9) 0 6 10 4 20 (6)
Not definable (ND) 0 0 0 2 2 (<1)
Total 80 147 75 61 363 (100)
Abbreviation: ICF, International Classification of Functioning, Disability, and Health; PEDI-CAT, Pediatric Evaluation of Disability Inventory-Computer Adaptive Test.


The purpose of this study was to examine the degree to which the PEDI-CAT items comprehensively reflect the activity and participation components of the ICF using the ICF linking rules.7,8 The linking process had very good to almost perfect interrater reliability between the 2 raters. Three hundred and sixty-three codes spanning all 9 chapters of the Activity and Participation component of the ICF were matched to three hundred twenty-nine meaningful concepts identified in the PEDI-CAT. Upon reaching consensus, 99% of ICF codes assigned to PEDI-CAT items/meaningful concepts were from the Activities and Participation component. Together, these findings support the PEDI-CAT's content validity as a measure of activity and participation.

The Importance of Activity and Participation

Participation is a complex and multidimensional, functional activity level that is important to children with disabilities and their families.20 As a result, health care workers have a responsibility to use outcome assessment tools and interventions focused on functional activity levels, as well as social and community participation. The PEDI-CAT, an updated version of the PEDI, has an expanded age range (birth to 21 years of age) and addresses a much larger segment of the pediatric population than the original test. The PEDI-CAT also includes a “Responsibility” subscale, which assesses how well children and youth can independently perform tasks important to independent living. The expanded age range and responsibility subscale make the PEDI-CAT especially useful for therapists working in educational environments, as it encompasses the goal of the IDEA to “prepare students for employment and independent living.”4 The computerized adaptive testing (CAT) platform of the PEDI-CAT provides an efficient and precise measurement of a student's function. The current study indicates that 99% of the items included in the PEDI-CAT are representative of the Activities and Participation chapters of the ICF. A tool with these characteristics is suitable in all pediatric clinical settings that promote activity and participation, such as the educational and home-based environments, rehabilitation, and outpatient centers.

The Linking Rules Process

The linking rules process was an effective method to critically evaluate the PEDI-CAT content in relation to the ICF. The process included a training module for all raters and followed a systematic, stepwise process directed by the updated linking rules.15 Beyond the ICF browser, raters used additional resources including available ICF updates and the ICF-CY to accurately match meaningful concepts within the unique pediatric items of the PEDI-CAT to the 9 activity and participation chapters of the ICF. These factors contributed to the favorable agreement (κ) between the 2 primary raters. The highest degree of agreement between raters occurred when the PEDI-CAT item/meaningful concept and ICF chapter constructs were immediately apparent (ie, the same word(s) were used in both the PEDI-CAT item/meaningful concept and the ICF chapter). The rater agreement in the current study (87%) was consistent with previous comparative investigations of pediatric outcome measure content and the ICF. Using the linking rules process, Chien et al21 reported between rater agreement to the chapter-level ranging from 80.7% to 98.5%.

The code “ND” (not definable by the ICF) was used twice when linking items in the PEDI-CAT Responsibility domain. While raters felt that those 2 items would have been associated with the ICF major life area classification chapter D8, there were no classification choices explicitly pertaining to “personal paperwork” or “personal documentation.”

Some items of the PEDI-CAT contained more than one meaningful concept, and some meaningful concepts required linking with more than one ICF chapter code to most accurately represent the constructs within the item. The items within the PEDI-CAT domains of Daily Activities and Mobility linked with 4 chapters of Activities and Participation within the ICF, and Social/Cognitive and Responsibility domains each linked with 7 of the 9 chapters. These findings support the complex, multidimensional nature of many of the skills, tasks, and life roles developed throughout the age range encompassed by the PEDI-CAT.

Difficulties Matching PEDI-CAT Items to ICF Chapters

Some meaningful concepts within the PEDI-CAT items were difficult to capture, either because individual concepts were not covered by ICF chapters, the aim of the item was not obvious, or because different levels of chapter organization were identified within the ICF. These difficulties in linking the PEDI-CAT items with the ICF were minimal and did not significantly impact the outcome of this study. For example, when evaluating the PEDI-CAT item on the Social/Cognitive subscale, “Teaches another person a new game or activity by giving examples and explanations,” the ICF does not contain a specific chapter for the concept “teaching.” Similarly, items containing concepts such as “safety” and “opening or closing a door” are not exclusively captured by any one ICF code. In such instances, raters used similar strategies as those used by Fayed et al22 and identified the target domain of the item by asking “What is it about?” When the aim of the item was not immediately apparent, we found it helpful to look beyond the literal wording in breaking down the complexity of the item. Due to the complexity of item “Teaches another person a new game or activity by giving examples and explanations,” 2 ICF codes were provided by the raters. New games and activities were considered “general tasks and demands” (D2), and teaching by giving examples and explanations requires “interpersonal interactions and relationships” (D7). Therefore, meaningful concepts were linked to more than one code.

When matching the PEDI-CAT items within different ICF chapters of Activities and Participation, raters also identified different levels of organization of the ICF constructs. For example, in chapter D5, Self-care, D550 eating, and D560 drinking are structured comprehensively to include all aspects of eating and drinking, including manipulation of utensils. However, the definition for D540 dressing seemed to be less comprehensive in its inclusion of manipulative skills necessary for buttoning, causing us to choose a second code denoting hand function along with the dressing code for item DA061 “Puts on and buttons a front-buttoning shirt.”

Additional issues when linking child health instruments to the ICF previously identified by Fayed et al22 were not encountered. PEDI-CAT items were not complicated by questions about behavioral or personal factors, and the items are worded to consistently reflect the child's function, not the parent's or caregiver's. The responses to some individual PEDI-CAT items did not require linking as specified in the ICF linking rules.7 Possible responses are used to describe the child's ability or their amount of responsibility and do not contain meaningful concepts. Raters did link accompanying illustrations, examples, or additional content information that either clarified meaningful concepts or contributed to additional meaningful concepts within the items.

The PEDI-CAT and Environmental Factors

The PEDI-CAT conceptually describes function and disablement consistently with the ICF framework.11 The Environment dimension of the ICF is not measured as a separate aspect of function within the PEDI-CAT. Instead the PEDI-CAT assumes that functional assessment takes place within a child or youth's typical daily environment, and “tries to define the relevant context of performance within each item.”21 PEDI-CAT item administration instructions encourage the parent/caregiver to answer based on the child's typical performance in their typical environment, allowing for a variety of methods to be used (including the use of mobility devices, alternative methods of communication, adaptation, and modifications).11 We did not link any PEDI-CAT items to environmental factor codes and believe that any environmental reference served to supply a context to the item.

The PEDI-CAT manual also states that “the three functional skills domains address the Activity dimension” while “the fourth domain, Responsibility, examines Participation.”11 As stated in the updated linking rules,8 when linking items with meaningful concepts that address aspects of the component Activities and Participation, the differentiation between activities and participation cannot be made through the linking process. We matched the PEDI-CAT items/meaningful concepts to the component Activities and Participation without making attempts to differentiate between the 2 concepts. During discussions regarding consensus, however, we did identify that descriptions of certain items/meaningful concepts extended beyond task completion (ie, activity) and included the description of a standardized environment consistent with the ICF definition of participation. For example, item 58 in the Mobility Section “Walks several hours at family or school outing (such as zoo, etc.)” extends beyond the activity of walking and includes the environmental context of family or school outing. The use of the environmental context would classify this item under participation based on the ICF definition. Analyzing the items of the PEDI-CAT with the ICF classifications with the purpose of differentiating activities and participation would be a suggested area of future study.

The findings of this study should be interpreted in light of its limitations. First, this current study only focused on the linking of PEDI-CAT item content with the ICF, potentially supporting its content validity. Second, use of the updated linking rules created a systematic process for linking the PEDI-CAT items to the ICF, but challenges occurred. Interpretation of meaningful concepts within each item and of the ICF may have been influenced by differences in the perspectives, expertise, and knowledge bases between the raters. Due to the length of the PEDI-CAT and expansiveness of the item bank, we were not able to complete linking of any one PEDI-CAT domain within one sitting (2-3 hours). Although coding for all 4 PEDI-CAT domains was completed by each rater individually prior to the first scheduled discussion, we allowed the raters to conduct an independent review of their linking following the first consensus meeting. Our subsequent discussions may have been influenced by the earliest one when we became more comfortable with the linking process, and developed a better familiarity with the ICF classifications. While these factors may have influenced our final agreement results, we felt that our process was similar to other studies of linking multiple outcome measures with the ICF.9,21 Additional studies analyzing the PEDI-CAT items with the ICF are recommended to confirm the linking results of this study. Third, the current study was completed using the online ICF browser and available ICF updates. Previous studies14 had been completed using the ICF-CY prior to the merging of the ICF-CY with the ICF. The ongoing ICF update process may have contributed to variations in final coding of outcome measure items and comparison of study results should take this into account.


The current study showed that the PEDI-CAT, with its revised item bank, updated CAT platform, and expanded age range from the original PEDI, was strongly representative of the Activities and Participation components of the ICF, with item content of the PEDI-CAT distributed across all 9 chapters of the Activities and Participation components of the ICF. The linking process followed by the 2 primary raters showed a favorable agreement and reliability. These findings support the content validity of the PEDI-CAT as a measure of activity and participation. Health care professionals can be encouraged to select the PEDI-CAT as a measure of activity and participation within clinical research and patient management settings.


1. World Health Organization. International Classification of Functioning, Disability, and Health. Geneva, Switzerland: World Health Organization; 2001.
2. World Health Organization. International Classification of Functioning, Disability and Health—Children and Youth Version. Geneva, Switzerland: World Health Organization; 2007.
3. World Health Organization. WHOFIC Resolution 2012: Merger of ICF-CY into ICF. Published 2012. Accessed October 2014.
4. IES. PL 108-446, Individuals with Disabilities Education Improvement Act of 2004.
5. Schiariti V, Klassen AF, Cieza A, et al. Comparing contents of outcome measures in cerebral palsy using the International Classification of Functioning (ICF-CY): a systematic review. Eur J Paediatr Neurol. 2013;18(1):1–12.
6. Schiariti V, Fayed N, Cieza A, Klassen A, O'Donnell M. Content comparison of health-related quality of life measures for cerebral palsy based on the International Classification of Functioning. Disabil Rehabil. 2011;33(15/16):1330–1339.
7. Cieza A, Brockow T, Ewert T, et al. Linking health-status measurements to the international classification of functioning, disability and health. J Rehabil Med. 2002;34(5):205–210.
8. Cieza A, Geyh S, Chatterji S, Kostanjsek N, Ustun B, Stucki G. ICF linking rules: an update based on lessons learned. J Rehabil Med. 2005;37(4):212–218.
9. Fayed N, Cieza A, Bickenbach JE. Linking health and health-related information to the ICF: a systematic review of the literature from 2001 to 2008. Disabil Rehabil. 2011;33(21/22):1941–1951.
10. World Health Organization. ICF Brouser. Accessed October 2017.
11. Haley SM, Coster W, Dumas HM, Fragala-Pinkham MA, Moed R. PEDI-CAT: Pediatric Evaluation of Disability Inventory Computer Adaptive Test Manual 1-3-6. Boston, MA: Trustees of Boston University, under license to CREcare, LLC; 2011.
12. Dumas HM, Fragala-Pinkham MA. Concurrent validity and reliability of the pediatric evaluation of disability inventory-computer adaptive test mobility domain. Pediatr Phys Ther. 2012;24(2):171–176.
13. Dumas HM, Fragala-Pinkham MA, Rosen EL, Lombard KA, Farrell C. Pediatric Evaluation of Disability Inventory Computer Adaptive Test (PEDI-CAT) and Alberta Infant Motor Scale (AIMS): Validity and Responsiveness. Phys. Ther. 2015;95(11):1559–1568.
14. Ostensjo S, Bjorbaekmo W, Carlberg EB, Vollestad NK. Assessment of everyday functioning in young children with disabilities: an ICF-based analysis of concepts and content of the Pediatric Evaluation of Disability Inventory (PEDI). Disabil Rehabil. 2006;28(8):489–504.
15. Bemis-Dougherty AP, Harwood KJP. Online: The International Classification of Functioning, Disability, and Health Overview (No CEUs). Alexandria, VA: American Physical Therapy Association; February 2010.
16. World Health Organization. List of Official ICF Updates. Geneva, Switzerland: World Health Organization; 2014.
17. Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas. 1960;20(1):37–46.
18. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–174.
19. Altman DG. Practical Statistics for Medical Research. London, England: Chapman and Hall; 1999.
20. Palisano RJ, Chiarello LA, Orlin M, et al. Determinants of intensity of participation in leisure and recreational activities by children with cerebral palsy. Dev Med Child Neurol. 2010;53(2):142–149.
21. Chien CW, Rodger S, Copley J, Skorka K. Comparative content review of children's participation measures using the International Classification of Functioning, Disability and Health-Children and Youth. Arch Phys Med Rehabil. 2014;95(1):141–152.
22. Fayed N, Cieza A, Bickenbach J. Illustrating child-specific linking issues using the Child Health Questionnaire. Am J Phys Med Rehabil. 2012;91(13, suppl 1):S189–S198.

content validity; Disability Inventory-Computer Adaptive Test; ICF; pediatric evaluation of linking rules

© 2018 Academy of Pediatric Physical Therapy of the American Physical Therapy Association