Inter-rater reliability in assigning ICF codes to children with disabilities

Purpose: The purpose of this study was to determine if independent raters would arrive at similar decisions about which codes from the Activities and Participation component of the International Classification of Functioning, Disability, and Health (ICF) to assign to children with disabilities based on the results of standard paediatric functional assessment measures. Method: Children (N= 60), 9 months to 17.75 years old, with a range of disabilities were assessed using either the Vineland Adaptive Behavior Scales ( n= 20), the Pediatric Evaluation of Disability Inventory (PEDI) ( n= 20), or the School Function Assessment (SFA) ( n= 20). Two raters independently determined which of 40 codes from the Activities and Participation component of the ICF applied to each child based on items and standard scores from the assessment measure. The Cohen's kappa statistic was used as the measure of inter-rater reliability. Results: The kappa statistic was greater than or equal to0.70, the criterion for good agreement in this study, for 23/40 codes using the Vineland, 17/40 codes using the PEDI, and 17/40 codes using the SFA. The mean kappa statistic reached greater than or equal to0.70 for the Self care domain, using all three tools. The mean kappa statistic also reached greater than or equal to0.70 for Learning and Applying Knowledge and Mobility using the Vineland. Conclusions: Independent raters reached high rates of agreement when assessment test items were structured developmentally and corresponded to a single ICF code. For domains other than Self care, alternative assessment strategies may be necessary to improve assigning ICF codes to children with disabilities.