Psychometric Validation of the Psoriasis Symptoms and Impacts Measure (P-SIM), a Novel Patient-Reported Outcome Instrument for Patients with Plaque Psoriasis, Using Data from the BE VIVID and BE READY Phase 3 Trials.

ABSTRACT

INTRODUCTION: Plaque psoriasis can significantly impact patients' quality of life. We assessed psychometric properties of the Psoriasis Symptoms and Impacts Measure (P-SIM), developed to capture patients' experiences of signs, symptoms and impacts of psoriasis. METHODS: Pooled, blinded, 16-week data from 1002 patients in the BE VIVID and BE READY bimekizumab phase 3 trials were analysed. The suitability of the P-SIM missing score rule (weekly scores considered missing if ≥ 4 daily scores were missing) was assessed. Test-retest reliability was evaluated using intraclass correlation coefficients (ICCs). Convergent validity was assessed between P-SIM and relevant patient-reported outcome (PRO) (Dermatology Life Quality Index [DLQI], DLQI item 1 [skin symptoms], Patient Global Assessment of Psoriasis) and clinician-reported outcome (ClinRO) scores (Psoriasis Area and Severity Index [PASI], Investigator's Global Assessment [IGA]) at baseline and week 16. Known-groups validity was assessed, comparing P-SIM scores between patient subgroups predefined using PASI/IGA scores. Sensitivity to change over 16 weeks was evaluated; responder definition (RD) thresholds were explored. RESULTS: The missing score rule used did not impact P-SIM scores. Test-retest reliability analyses demonstrated excellent score reproducibility (ICC 0.91-0.98). Inter-item correlations at baseline and week 16 were strong (> 0.5), apart from "choice of clothing" with "skin pain" and "burning" at baseline (both 0.49). All P-SIM scores were moderately to strongly correlated with other outcomes, demonstrating convergent validity, apart from ClinROs (PASI, IGA) at baseline that had low variability. P-SIM scores discriminated known groups at week 16, confirming known-groups validity. Changes from baseline to week 16 in P-SIM and other clinically relevant outcomes were strongly correlated (> 0.5; weaker with ClinROs), establishing sensitivity to change. Anchor-based RD analyses determined a four-point P-SIM item score decrease as indicative of marked clinically meaningful improvement. CONCLUSION: P-SIM scores demonstrated good reliability, validity and sensitivity to change. A four-point RD threshold could be used to assess 16-week treatment effects. TRIAL REGISTRATION BE VIVID NCT03370133; BE READY NCT03410992.