Influencing Factors of Pulmonary Infection after Thoracoscopic Surgery in Patients with Lung Cancer and Construction of A Risk Prediction Nomogram Model

LIU Jianlei; WEN Zuochuan; WANG Yongping; TIAN Hao; WANG Hui

doi:10.3639/j.issn.2095-3097.2025.07.025

Abstract：Objective To analyze the influencing factors of pulmonary infection after thoracoscopic surgery in patients with lung cancer and construct a risk prediction nomogram model.Methods A total of 204 patients with lung cancer who un-derwent thoracoscopic surgery at Beijing Daxing District People's Hospital from January 2022 to December 2024 were divided into an infection group(n=39)and a non-infection group(n=165)based on whether pulmonary infection occurred after sur-gery.Univariate and multivariate logistic stepwise regression analyses were used to identify influencing factors of pulmonary infection.A nomogram prediction model for pulmonary infection after thoracoscopic surgery in lung cancer patients was con-structed,and its predictive value was evaluated using the receiver operating characteristic(ROC)curve.Results Among the 204 lung cancer patients who underwent thoracoscopic surgery,39 patients(19.12%)developed pulmonary infections,with a total of 53 strains of pathogens detected,predominantly Gram-negative bacteria(62.26%).Univariate analysis showed that the infection group had higher proportions of older age,diabetes history,TNM clinical stage Ⅲ,intraoperative blood loss≥200 mL,operation time≥3 h,and mechanical ventilation time≥12 h compared to the non-infection group.Additionally,C-re-active protein levels and white blood cell counts were higher,while albumin levels were lower in the infection group(P<0.05).Multivariate analysis showed that older age,TNM clinical stageⅢ,mechanical ventilation time≥12 h,and elevated C-re-active protein levels were independent risk factors for postoperative pulmonary infection(P<0.05).The area under the curve(AUC)(95%CI)of the nomogram model for predicting pulmonary infection was 0.902(0.851-0.953),with a specificity of 68.33%and a sensitivity of 92.08%.Conclusion The nomogram model constructed based on the risk factors of pulmonary infection has good predictive performance and can be used as an effective auxiliary means for clinical risk assessment.