Establishment and validation of a clinical risk scoring model to predict fatal risk in SFTS hospitalized patients

Abstract

Abstract Background Severe fever with thrombocytopenia syndrome (SFTS) is an emerging tick-borne infection with a high case fatality rate. Significant gaps remain in studies analyzing the clinical characteristics of fatal cases. Methods From January 2017 to June 2023, 427 SFTS cases were included in this study. A total of 67 variables about their demographic, clinical, and laboratory data were collected. Univariate logistic regression and the least absolute shrinkage and selection operator (LASSO) method was used to screen predictors from the cohort. Multivariate logistic regression was used to identify independent predictors and nomograms were developed. Calibration, decision curves and area under the curve (AUC) were used to assess model performance. Results The multivariate logistic regression analysis screened out the four most significant factors, including age > 70 years (p = 0.001, OR = 2.516, 95% CI 1.452–4.360), elevated serum PT (p 8.0 μmol/L) (p < 0.001, OR = 4.433, 95% CI 1.888–10.409). The AUC of the nomogram based on these four factors was 0.813 (95% CI, 0.758–0.868). The bootstrap resampling internal validation model performed well, and decision curve analysis indicated a high net benefit. Conclusions The nomogram based on age, elevated PT, high serum urea level, and high viral load can be used to help early identification of SFTS patients at risk of fatality.

Keywords

• SFTS
• Mortality
• Prediction model
• Logistic regression
• Nomogram