A Risk Predictive Model for the Severity of Illness in Patients with Severe Fever with Thrombocytopenia Syndrome

Abstract

Abstract Background Severe fever with thrombocytopenia syndrome (SFTS) is a newly identified infectious disease characterized by a high mortality rate and wide prevalence. The objective of our study was to investigate the factors that influence the severity of SFTS and develop a predictive model applicable to primary care settings. Methods This retrospective study examined a cohort of 169 patients with SFTS who received medical attention from May 2015 to February 2022. Clinical and laboratory data were compared between the mild and severe groups. Independent risk factors contributing to the severity of the patients' condition were assessed using multifactorial logistic regression analysis. Subsequently, a nomogram was constructed based on the outcomes of the regression analysis. The predictive model's discrimination and calibration were evaluated using metrics such as the concordance index (C-index), ROC curve, and Hosmer-Lemeshow analysis. Results A total of 169 patients diagnosed with SFTS were included in this study. A comprehensive analysis of 19 factors was conducted, including AGE, neurological manifestations, PLT, NEUT%, MONO%, CK, CK-MB, LDH, ALT, AST, BUN, Cr, Ca, APTT, PCT, HCT, ALB, HCT-ALB, and HGB. Logistic regression analysis revealed that PLT (OR = 0.930, 95% CI = 0.892–0.970), CK (OR = 1.005, 95% CI = 1.001–1.008), APTT (OR = 1.042, 95% CI = 1.002–1.083), LDH (OR = 1.004, 95% CI = 1.000-1.007), and NEUT% (OR = 1.062, 95% CI = 1.020–1.106) were identified as independent risk factors for disease severity. The constructed nomogram exhibited excellent predictive performance in estimating severe disease (C-index = 0.927, AUC = 0.927, sensitivity = 84.4%, specificity = 87.3%, Hosmer-Lemeshow analysis SD = 0.00051, and quantile of absolute error = 0.036). Conclusions Decreased PLT, increased LDH, CK, APTT, and NEUT% serve as reliable predictors of severe disease progression in patients with SFTS. Utilizing these five predictors, a predictive line chart exhibits strong capability in accurately assessing the risk of severe disease during the course of the illness.