Responsiveness to change and interpretability of the simplified psoriasis index

Abstract

The Simplified Psoriasis Index (SPI) is a summary measure of psoriasis with separate components for current severity (weighted for functionally or psychosocially important sites), psychosocial impact, and past behavior. The current severity components of the professionally assessed SPI (proSPI-s) and self-assessed SPI (saSPI-s) have each been shown to be valid and reliable. Their responsiveness to change and equivalence to the current standard (Psoriasis Area and Severity Index, PASI) were investigated. Responsiveness and minimum clinically important differences (MCIDs) were derived from PASI changes from baseline at weeks 4 (n=100) and 10 (n=65) in patients commencing therapy for psoriasis. Receiver operating characteristic (ROC) analysis confirmed that both measures detected responsiveness well (area under the curve (AUC)=0.72-0.96). On ROC and PASI-based anchor analysis, MCIDs equated to mean absolute and percentage changes of 5 and 60% (proSPI-s), and 7 and 70% (saSPI-s). Satisfactory response as defined by 75% reduction in PASI equated to 85 and 95% reductions in proSPI-s and saSPI-s, respectively. PASI-equivalent cutoff scores for mild (PASI<10) and severe (PASI>20) psoriasis were <9 and >18 for proSPI-s (n=300) and <10 and >20 for saSPI-s (n=200; AUC=0.86-0.96). These studies further support the validity of SPI for use in routine clinical practice.