Measurement Properties of PROMIS Measures Relevant to Chronic Skin Diseases.

Abstract

Patient-Reported Outcomes Measurement Information System (PROMIS) is a generalized set of patient-reported outcome measures (PROMs) that can be readily integrated into the dermatologic clinic setting, but the measurement properties of these PROMs among patients with a range of chronic skin diseases have yet to be established. To evaluate the structural validity, internal consistency, construct validity, and responsiveness of PROMIS short-form measures for anxiety, depression, social satisfaction, and social isolation as well as the construct validity and responsiveness of single-item pain and itch intensity measures. In this cohort study, adults 18 years and older with a self-reported chronic skin disease (eg, atopic dermatitis, acne, psoriasis) were recruited in February 2025 from across the US via an online survey platform. Enrolled participants completed a baseline survey of PROMIS and several other relevant measures and then a follow-up survey 3 months later. Measures collected include the PROMIS measures, Dermatology Life Quality Index, Skindex-29, a baseline patient global assessment, and a change in skin disease anchor question at follow-up. Structural validity, internal consistency, construct validity, and responsiveness were evaluated for the short-form measures. Construct validity and responsiveness were assessed for the single-item measures. Of 549 included participants with chronic skin diseases, 287 (52.3%) were female, 262 (47.7%) were male, and the median (IQR) age was 36 (30-45) years. A total of 249 patients (45.1%) had atopic dermatitis, 246 (44.8%) had acne, and 113 (20.6%) had psoriasis. The anxiety, depression, social satisfaction, and social isolation short-form measures demonstrated adequate structural validity (ie, all with comparative fit index and/or Tucker-Lewis index values greater than 0.95; standardized root mean residual values less than 0.08) and internal consistency (Cronbach α values greater than 0.90). All measures showed sufficient construct validity on known-groups and convergent validity testing (eg, magnitude of Pearson correlations between measures evaluating similar underlying constructs with values of 0.30 or greater). When comparing responsiveness based on standardized response means (SRMs), the pain (SRM, 0.41) and itch intensity (SRM range, 0.32-0.44) measures showed similar responsiveness as the Skindex-29 symptoms domain (SRM, 0.47), and the social satisfaction (SRM, -0.20) and isolation (SRM, 0.27) measures showed similar responsiveness to the Skindex-29 functioning domain (SRM, 0.25). Responsiveness of the anxiety (SRM, 0.11) and depression (SRM, 0.16) measures were lower than the Skindex-29 emotions domain (SRM, 0.43). In this study, PROMIS measures were valid for assessing the physical, emotional, and social effects of chronic skin diseases, and measures for physical and social impact tracked improvements with treatment.