Development and validation of a clinical and laboratory-based nomogram to predict mortality in patients with severe fever with thrombocytopenia syndrome.

Abstract

Background: Severe fever with thrombocytopenia syndrome (SFTS) is an emerging global infectious disease with a high mortality rate. Clinicians lack a convenient tool for early identification of critically ill SFTS patients. The aim of this study was to construct a simple and accurate nomogarm to predict the prognosis of SFTS patients.

Methods: We retrospectively analyzed the clinical data of 372 SFTS patients collected between May 2015 and June 2023, which were divided 7:3 into a training set and an internal validation set. We used LASSO regression to select predictor variables and multivariable logistic regression to identify independent predictor variables. Prognostic nomograms for SFTS were constructed based on these factors and analysed for concordance index, calibration curves and area under the curve (AUC) to determine the predictive accuracy and consistency of the model.

Results: In the training set, LASSO and multivariate logistic regression analyses showed that age, SFTSV RNA, maximum body temperature, pancreatitis, gastrointestinal bleeding, pulmonary fungal infection (PFI), BUN, and PT were independent risk factors for death in SFTS patients. There was a strong correlation between neurological symptoms and mortality (P < 0.001, OR = 108.92). Excluding neurological symptoms, nomograms constructed based on the other eight variables had AUCs of 0.937 and 0.943 for the training and validation sets, respectively. Furthermore, we found that age, gastrointestinal bleeding, PFI, bacteraemia, SFTSV RNA, platelets, and PT were the independent risk factors for neurological symptoms, with SFTSV RNA having the highest diagnostic value (AUC = 0.785).

Conclusions: The nomogram constructed on the basis of eight common clinical variables can easily and accurately predict the prognosis of SFTS patients. Moreover, the diagnostic value of neurological symptoms far exceeded that of other predictors, and SFTSV RNA was the strongest independent risk factor for neurological symptoms, but these need to be further verified by external data.