Derivation and validation of adult Still Activity Score (SAS)

Abstract

Objectives: Adult-onset Still's disease (AOSD) is a multi-systemic, autoinflammatory disorder. Several activity scores have been proposed but none of them have been adopted universally. Our aim was to create a clinician-friendly activity scoring system by using simple clinical and laboratory parameters. Methods: AODS patients, according to Yamaguchi criteria, were included in this cross-sectional, multicenter study. Derivation and validation cohorts were constituted. Demographic, clinical, and laboratory evaluation at the study visit; patients’ and physicians’ global assessments of disease activity (both VAS/Likert scale) were recorded. To develop the score, an ordinal logistic regression model was used to determine independent predictors of physicians’ global assessments of disease activity. Clinically and statistically significant variables were weighted according to regression coefficients. Then, performance of the score was tested on the validation cohort. Results: A total of 197 consecutive AOSD patients (125 in derivation, 72 in validation cohorts) were included. Final Still Activity Score was fever (2 points), arthralgia (2 points, plus 1 point if arthritis was present in ≥ 2 joints), neutrophilia ≥ 65% (1 point) and ferritin ≥ 350 ng/mL (1 point) (maximum of 7 points). The SAS yielded an AUC value of 0.98 (0.96–1.00) in the derivation cohort and 0.91 (95%CI: 0.85–0.98) in the validation cohort to discriminate high AOSD activity from moderate-inactive AOSD. The correlation of SAS with PGA was 83% for the derivation cohort and 76% for the validation cohort. Conclusions: SAS has shown a good test performance to distinguish active AOSD patients from others. SAS may be a useful method for evaluating the disease activity of AOSD patients in daily practice.