Establishment of prediction model for postoperative nosocomial infection in lung cancer patients
OBJECTIVE To analyze the risk factors for the postoperative nosocomial infection in the lung cancer pa-tients undergoing video-assisted thoracic surgery(VATS)and establish the risk prediction model.METHODS To-tally 150 lung cancer patients who were treated in Linping Campus of the Second Affiliated Hospital of Zhejiang University School of Medicine from Jun 2021 to Dec 2022 were recruited as the research subjects and were divided into the infection group and the no infection group according to the status of infection after VATS.Logistic re-gression analysis was performed for the risk factors for the nosocomial infection,the prediction model was estab-lished,and receiver operating characteristic(ROC)curves were drawn to analyze the predictive efficiency of the model.RESULTS Totally 34 of 150 lung cancer patients had the postoperative infection,with the incidence rate 22.67％.The result of multivariate logistic regression analysis showed that Ⅲ-Ⅳ stage of TNM,chronic obstruc-tive pulmonary disease(COPD)and operation duration no less than 3 hours were the risk factors for the postoper-ative nosocomial infection in the lung cancer patients.The logistic prediction model was established based on the above risk factors:Logistic(P)=-6.797+0.778 × TNM stage+0.873 × COPD+1.324 × operation duration;the sensitivity of the model was 81.62％in prediction of the postoperative nosocomial infection in the lung cancer patients,with the accuracy 78.56％,the positive predictive value 83.42％,the area under curve(AUC)0.743.CONCLUSION TNM stage,operation duration and complication with COPD are the risk factors for the postoper-ative nosocomial infection in the lung cancer patients.The model that is established based on the above risk factors has high predictive value.
Lung cancerThoracoscopic surgeryPostoperative infectionPrediction modelPredictive value
NETL
NSTL
万方数据
维普
