Psychometric Properties of the Patient-Reported Outcomes Measurement Information System (PROMIS®) Pediatric Anger Scale in the Dutch General Population

This study aimed to validate the PROMIS® pediatric v2.0 Anger scale in the Dutch general population, provide reference data, and compare reliability and relative efficiency between the full-length scale, its short-form, computerized adaptive test (CAT), and Pediatric Quality of Life Inventory (PedsQLTM) emotional functioning (EF) subscale scores. Children (N = 1,328), representative of the Dutch population, were asked to complete the PROMIS pediatric Anger scale (8–18 years) and PedsQLTM (8–17 years). A graded response model (GRM) was fit to the data. Structural validity was assessed by checking item-fit statistics (S−X2, p <.001 = misfit). For construct validity, a moderate correlation (Pearson’s r > 0.50) was expected between the Anger scale and PedsQLTM EF subscale score. Dutch mean T score based on the U.S. model was calculated to provide reference data and cut-offs. Standard error of measurement (SE(θ)) was used to assess reliability (SE(θ) <.32 =.90 reliability). Relative efficiency was calculated (1 − SE(θ)2/N items) to compare how good the measures performed relative to the amount of items administered. In total, 527 children completed the PROMIS pediatric Anger scale, of which 482 completed the PedsQLTM. Structural validitywas sufficient as no items displayedmisfit (S−X2 = 22.9–40.3, p >.001). The Anger scale score correlated moderately (Pearson’s r =.64) with the PedsQLTM EF subscale score. Dutch mean T score was 44.20 (SD = 11.39), with cut-offs of >52.2 for moderate and ≥62.3 for severe symptoms. Reliable measurements were obtained at the population mean and >2SD in the clinically relevant direction. CAT outperformed all other measures in efficiency. The PROMIS pediatric Anger scale displayed sufficient psychometric properties within the Dutch population and reference data are available.