Development of a nomogram model for individualized prediction of the risk of nosocomial infection after heart valve replacement in elderly patients
Objective To develop a nomogram model for individualized prediction of the risk of nosocomial infection after heart valve replacement in the elderly patients,so as to provide a basis for the early screening of the high-risk pop-ulation and the formulating of targeted prevention strategies.Methods The clinical data of 334 elderly patients who underwent heart valve replacement surgery in Qilu Hospital of Shandong University during Jan.1,2020 and Dec.31,2021 were retrospectively collected,including 176 males and 158 females,aged 60-81 years,with an average of(65.86±4.34)years.The independent risk factors for nosocomial infection were identified with Lasso regression and multivariate Logistic regression,based on which,a nomogram model for individualized prediction of risk was construc-ted.The internal validation of the model was tested with Bootstrap self-sampling method(n=1,000).The predictive performance of the model was tested with C-index,area under the receiver operating characteristic(ROC)curve,cali-bration curve or decision curve.Results A total of 91 nosocomial infection occurred,with an incidence of 27.25%.Heart failure,stress hyperglycemia,indignant gastric tube,pulmonary hypertension,left ventricular ejection fraction(LVEF)and American Society of Anesthesiologists(ASA)were independent risk factors for nosocomial infection.Based on the above variables,a nomogram model was constructed,whose C-index after correction was 0.80.The cali-bration curve showed that the model could accurately predict the risk of nosocomial infection.The clinical decision curve showed good net benefit.Conclusion The nomogram model for the prediction of nosocomial infection risk in the elderly pa-tients after heart valve replacement has good differentiation,calibration and clinical effectiveness,which can help to screen the high-risk group and formulate targeted intervention strategies,so as to reduce the incidence of nosocomial infection.
NETL
NSTL
万方数据
维普
