Psychometric validation of the Psoriasis Symptom Scale, Functional Assessment of Chronic Illness Therapy–Fatigue and pain‐Visual Analogue Scale in patients with generalized pustular psoriasis

Abstract Background Generalized pustular psoriasis (GPP) is a rare, chronic, inflammatory skin disease associated with considerable patient burden. The Psoriasis Symptom Scale (PSS), Functional Assessment of Chronic Illness Therapy–Fatigue (FACIT‐Fatigue) and pain‐Visual Analogue Scale (pain‐VAS) are patient‐reported outcomes (PROs) that have not yet been validated in patients with GPP. Objectives To evaluate the psychometric properties of the PSS, FACIT‐Fatigue and pain‐VAS using data from Effisayil 1, a randomised trial of spesolimab in patients with moderate‐to‐severe GPP. Methods Inter‐item correlations and confirmatory factor analysis (CFA) were performed using Week 1 data. Internal consistency was assessed with Cronbach's α coefficient using baseline and Week 1 data. Test–retest reliability was assessed using intraclass correlation coefficients (ICCs); change data for the GPP Physician Global Assessment total score and pustulation subscore were used to define a stable population. Convergent validity was assessed at baseline and Week 1 using Spearman's rank‐order correlations. Known‐groups validity was measured by analysis of variance using Week 1 data. Ability to detect change from baseline to Week 1 was evaluated by analysis of covariance. Results Inter‐item and item‐to‐total correlations were moderate or strong for most PSS and FACIT‐Fatigue items. CFA demonstrated the unidimensionality of the PSS and FACIT‐Fatigue, with high factor loadings for most items (PSS range, 0.75–0.94; FACIT‐Fatigue range, 0.11–0.93) and acceptable fit statistics. Both scores demonstrated internal consistency (Cronbach's α, 0.71 and 0.95, respectively). The PSS, FACIT‐Fatigue and pain‐VAS demonstrated test–retest reliability (ICCs ≥0.70) and good evidence of convergent validity. Furthermore, the PROs could differentiate between known groups of varying symptom severity (range, p < 0.0001–0.0225) and detect changes in symptom severity from baseline to Week 1 (range, p < 0.0001–0.0002). Conclusions Overall, these results support the reliability, validity and ability to detect change of the PSS, FACIT‐Fatigue and pain‐VAS as PROs in patients with GPP.