Construction and Evaluation of a Sepsis Risk Prediction Model for Urinary Tract Infection

Abstract

Background: Urinary tract infection (UTI) is one of the common causes of sepsis. However, nomograms predicting the sepsis risk in UTI patients have not been comprehensively researched. The goal of this study was to establish and validate a nomogram to predict the probability of sepsis in UTI patients.Methods: Patients diagnosed with UTI were extracted from the Medical Information Mart for Intensive Care III database. These patients were randomly divided into training and validation cohorts. Independent prognostic factors for UTI patients were determined using forward stepwise logistic regression. A nomogram containing these factors was established to predict the sepsis incidence in UTI patients. The validity of our nomogram model was determined using multiple indicators, including the area under the receiver operating characteristic curve (AUC), correction curve, Hosmer-Lemeshow test, integrated discrimination improvement (IDI), net reclassification improvement (NRI), and decision-curve analysis (DCA).Results: This study included 6,551 UTI patients. Stepwise regression analysis revealed that the independent risk factors for sepsis in UTI patients were congestive heart failure, diabetes, liver disease, fluid electrolyte disorders, APSIII, neutrophils, lymphocytes, red blood cell distribution width, urinary protein, urinary blood, and microorganisms. The nomogram was then constructed and validated. The AUC, NRI, IDI and DCA of the nomogram all showed better performance than traditional APSIII score. The calibration curve and Hosmer-Lemeshow test results indicate that the nomogram was well-calibrated. Improved NRI and IDI values indicate that our nomogram scoring system is superior to other commonly used ICU scoring systems. The DCA curve indicates that the DCA map of the nomogram has good clinical application ability.Conclusion: This study identified the independent risk factors of sepsis in UTI patients and used them to construct a prediction model. The present findings may provide clinical reference information for preventing sepsis in UTI patients.