Measurement properties of the patient global assessment numerical rating scale in moderate-to-severe psoriasis

Background Patient global assessment (PtGA) has been recommended as one of the core domains in psoriasis clinical trials. Among multiple versions of PtGA, the single-question, 11-point PtGA numeric rating scale (NRS) remains to be validated in patients with plaque psoriasis. Objectives To evaluate the psychometric characteristics of an 11-point PtGA NRS for disease severity in patients with moderate-to-severe plaque psoriasis. Methods Data were analysed from 759 patients with moderate-to-severe psoriasis in the Shanghai Psoriasis Effectiveness Evaluation CoHort (SPEECH), a prospective, multicentre and observational registry assessing the comparative effectiveness and safety of biologics (adalimumab, ustekinumab, secukinumab or ixekizumab), conventional systemic therapies (acitretin or methotrexate) and phototherapy. Results The test-retest reliability of the PtGA NRS showed good agreement (intraclass correlation coefficient range 0.79-0.83). No floor or ceiling effects of PtGA NRS were observed. The PtGA NRS was significantly correlated with the Psoriasis Area and Severity Index (PASI), static Physician Global Assessment (sPGA), body surface area, Dermatology Quality of Life Index (DLQI) and Hospital Anxiety and Depression Scale. Relatively large correlations of PtGA NRS with PASI and the DLQI 'symptoms and feelings' domain (all correlations >= 0.4 except at baseline) supported convergent validity. The presence of psoriatic arthritis or joint symptoms had no significant association with the PtGA NRS. In multivariate regression analyses, the PtGA NRS at baseline was predicted by age, lesion extent, lesion intensity, patients' symptoms and feelings, and impact on work or school. The PtGA NRS displayed known-groups validity with the PASI, sPGA and DLQI score bands. The PtGA NRS was responsive to change in PASI and DLQI after treatment. Anchor- and distribution-based approaches supported -3 as the minimal important difference for PtGA NRS. An absolute PtGA NRS <= 2 during follow-up was concordant with the state of minimal disease activity based on a 90% reduction in PASI (PASI 90) or PASI 90 plus a DLQI of 0/1. Sensitivity analysis using subgroup comparison and multiple imputation model yielded consistent conclusions. Conclusions The PtGA NRS showed good reliability, validity and responsiveness in patients with psoriasis, and was feasible in clinical trials and daily practice.