Psychometric properties and CAT performance of the PROMIS pediatric sleep disturbance, sleep-related impairment, and fatigue item banks in Dutch children and adolescents.

Abstract

This study's aim is to evaluate the psychometric properties of the pediatric Patient-Reported Outcomes Measurement Information System (PROMIS) sleep disturbance (SD), sleep-related impairment (SRI), and fatigue item banks in Dutch children and adolescents since they are yet to be validated in this population. Children and adolescents aged 8-18 years old (n = 1,325), representative of the Dutch general population, were invited. The following psychometric properties of both item banks were assessed: structural validity by fitting a graded response model (GRM) for which assumptions were checked and assessing item fit, reliability, and efficiency for the full-length item bank, short forms, and computerized adaptive test (CAT). Mean T scores and cutoff scores for mild, moderate, and severe scores were determined. A total of 527 children (response rate 39.7%) were included. SD did not meet the assumption of unidimensionality. The SRI and fatigue item banks showed sufficient structural validity. Three items showed GRM misfit; SRI item w026c and fatigue items 3224R1r and 4191R1r. Both item banks measured reliable (> .90) at the mean of the population and 2SD in the clinical direction. CATs outperformed short forms in efficiency. Mean T scores for the Dutch population were: SRI 47.5 (SD = 10.0) and fatigue 39.8 (SD = 12.4). Cutoffs included SRI minimal ≥ 34.7, moderate ≥ 55.6, severe ≥ 63.2 and fatigue minimal ≥ 23.9, moderate ≥ 49.9, severe ≥ 61.3. PROMIS pediatric SRI and fatigue item banks, short forms, and CAT showed sufficient structural validity and reliability in the Dutch population. Dutch reference values are provided. More research is needed for the PROMIS SD item bank. (PsycInfo Database Record (c) 2022 APA, all rights reserved).