Abstract
BACKGROUND: Severe Fever with Thrombocytopenia Syndrome (SFTS) is an acute infectious disease with high mortality. This study aimed to develop a quantitative scoring system for grading SFTS severity using dynamic clinical data. METHODS: A retrospective study included 547 confirmed SFTS patients from two hospitals. Clinical data were collected over a 14-day course (divided into four phases). Patients were grouped into survivors (n = 451) and non-survivors (n = 96). Statistical analyses, including Kaplan-Meier curves and log-rank tests, were performed. A prospective external validation cohort of 44 newly diagnosed patients was subsequently enrolled to validate the scoring system using C-statistic, calibration curves, and decision curve analysis (DCA). RESULTS: Of 547 patients, 96 (17.55%) were non-survivors. Multivariate logistic regression identified six independent prognostic factors across phases: age, platelet (PLT), aspartate aminotransferase (AST), and creatinine (Cr) (days 5-7); age, red blood cell distribution width (RDW), Cr, and lactate dehydrogenase (LDH) (days 8-10); Cr and LDH (days 11-14). A scoring system (0-11 points) was developed, stratifying patients into low (0-3), intermediate (4-7), and high (8-11) risk groups, with adverse outcome rates of 1.04, 22.92, and 76.04%, respectively. Kaplan-Meier curves showed significant prognostic differences (log-rank p < 0.001). External validation (44 cases) confirmed excellent performance: AUC 0.810-0.952, good calibration (Hosmer-Lemeshow p > 0.05), and net clinical benefit (DCA Eavg 0.068-0.098, Emax 0.422-0.559). CONCLUSION: A dynamic SFTS severity scoring system was developed and validated. Internal and external validation confirmed its reliability and clinical utility, providing a simple, practical tool for timely assessment and early intervention.