Background: Questionnaires translated into languages other than English are often not validated to the same extent as the English versions. This study examined the concurrent and discriminant validity of selected domains related to physical function from Spanish language versions of the Child Health Questionnaire (CHQ), Pediatric Outcomes Data Collection Instrument (PODCI), and Pediatric Evaluation and Disability Inventory (PEDI).
Methods: Concurrent validity was examined in 93 children with cerebral palsy by correlating questionnaire domain scores with Gross Motor Function Measure and Gillette Functional Assessment Questionnaire walking scale scores. Discriminant validity with respect to Gross Motor Function Classification System (GMFCS) level was examined using analysis of variance and nonparametric discriminant analysis.
Results: Concurrent validity was demonstrated for 3 domains from the PEDI (Mobility functional skills, tau = 0.62; Mobility caregiver assistance, tau = 0.46-0.55; and Self-care functional skills, tau = 0.30-0.36), 3 domains from the PODCI (Sports and physical function, tau = 0.48-0.51; Transfer and basic mobility, tau = 0.48-0.51; and Upper extremity physical function, tau = 0.28), and 1 domain from the CHQ (Physical function, tau = 0.31-0.36). Discriminant validity was demonstrated for the same domains based on significant decreases in domain scores with increasing GMFCS level. Discriminant validity was highest for the PODCI, which correctly classified 98% (91/93) of subjects into the correct GMFCS level when all 3 domains were considered.
Conclusions: For the first time, concurrent validity and discriminant validity have been demonstrated for the physical function domains of Spanish language versions of the PODCI, PEDI, and CHQ questionnaires. PODCI and PEDI had the highest concurrent validity, and PODCI had the best discriminant ability.
Clinical relevance: It is important to examine the validity of instruments when they have been translated from English into other languages. This importance will only increase as the population of non-English-speaking patients expands.