Measuring activities of daily living in children : validation of the German version of the pediatric evaluation of disability inventory (PEDI-G)

Abstract

Activities of daily living (ADL) are crucial for children with and without activity limitations because they enable them to participate in everyday life. For the evaluation of children’s ADL performance, health professionals such as occupational therapists use standardized ADL assessments, e.g. the Pediatric Evaluation of Disability Inventory (PEDI).

The general aim of this thesis was to evaluate evidence for the validity and the reliability of the German version of the PEDI (PEDI-G) for use in children with and without activity limitations in Austria, Germany, and Switzerland.

In Study I, a qualitative approach was chosen to gain knowledge about the process of translating and adapting an assessment in a review committee. Members from all three countries were recruited to represent a range of competencies and knowledge. The findings describe the adaptation of the PEDI for use in a new context as a question of striking a triple balance between (i) staying close to the original assessment while adapting the translation to the new context, (ii) making decisions and reaching consensus, (iii) catering for group dynamics while sticking to the planned schedule of the review committee.

The precision (inter-rater and test-retest reliability) of the PEDI-G was examined in Study II which included children with (n=64) and without activity limitations (n=53) from Austria, Germany, and Switzerland. The results of Study II demonstrate that the PEDI-G produces reliable results across raters and time points. As the test-retest reliability of the Modification Scale is overall insufficient, this scale may not be useful for measuring change.

The evidence of validity of the internal structure of the PEDI-G was evaluated in children with (n=198) and without (n=64) known activity limitations in Study III. Exploratory factor analysis was used on the subscales of the Functional Skills Scale and the items of the Caregiver Assistance Scale of the six PEDI-G domains. The results of the analysis suggest that a one-factor model should be used for all six PEDI-G domains.

The evidence relating to the validity of the PEDI-G was further examined with a Rasch model evaluating goodness-of fit, differential item functioning by country, and differential test functioning at item level of the Functional Skills Scale and the Caregiver Assistance Scale. Although a larger proportion of differences than expected exists at item level, this does not impact the generated individual PEDIG measure of the PEDI-G domains.

In conclusion, the results of these studies contribute to the evidence for the reliability and the validity of the PEDI-G domains in evaluating ADL of children with and without activity limitations living in Austria, Germany, and Switzerland. Still, it remains necessary to further examine the psychometric properties of the PEDI-G within larger samples and including more details about the parents, caregivers and contexts.