Novel nomograms to predict risk and prognosis in hospitalized patients with severe fever with thrombocytopenia syndrome.

Background: Severe fever with thrombocytopenia syndrome (SFTS) is an emerging and life-threatening infectious disease caused by SFTS virus. Although recent studies have reported the use of nomograms based on demographic and laboratory data to predict the prognosis of SFTS, no study has included viral load, which is an important factor that influences the prognosis, when compared with other risk factors. Therefore, this study aimed to develop a model that predicts SFTS prognosis before it reaches the critical illness stage and to compare the predictive ability of groups with and without viral load. Methods: Two hundred patients with SFTS were enrolled between June 2018 and August 2023. Data were sourced from the first laboratory results at admission, and two nomograms for mortality risk were developed using multivariate logistic regression to identify the risk variables for poor prognosis in these patients. We calculated the area under the receiver operating characteristic curve (AUC) for the two nomograms to assess their discrimination, and predictive abilities were compared using net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Results: The multivariate logistic regression analysis identified four independent risk factors: age, bleeding manifestations, prolonged activated partial thromboplastin time, and viral load. Based on these factors, a final nomogram predicting mortality risk in patients with SFTS was constructed; in addition, a simplified nomogram was constructed excluding the viral load. The AUC [0.926, 95% confidence interval (CI): 0.882-0.970 vs. 0.882, 95% CI: 35 0.823-0.942], NRI (0.143, 95% CI, 0.036-0.285), and IDI (0.124, 95% CI, 0.061-0.186) were calculated and compared between the two models. The calibration curves of the two models showed excellent concordance, and decision curve analysis was used to quantify the net benefit at different threshold probabilities. Conclusion: Two critical risk nomograms were developed based on the indicators for early prediction of mortality risk in patients with SFTS, and enhanced predictive accuracy was observed in the model that incorporated the viral load. The models developed will provide frontline clinicians with a convenient tool for early identification of critically ill patients and initiation of a better personalized treatment in a timely manner.