目的 探讨分析胰十二指肠切除术（PD）后肺部感染的危险因素并构建预测模型以指导临床早期干预。 方法 回顾性纳入2022年1月至2024年10月新疆医科大学第一附属医院行PD的220例患者，排除术前肺部感染、资料不全及围手术期死亡病例。收集术前、术中及术后临床资料，通过单因素分析筛选潜在危险因素，采用多因素logistic回归确定独立危险因素，并构建列线图预测模型。 结果 84例（38.2%）患者发生术后肺部感染。多因素分析显示，高BMI（OR= 1.12，95%CI：1.02 ~ 1.23，P < 0.05）、术后3 d低白蛋白（OR= 0.91，95%CI：0.83 ~ 0.99，P < 0.05）、术后腹腔引流管留置时间长（OR= 1.05，95%CI：1.01 ~ 1.09，P < 0.05）、术后卧床时间长（OR= 1.56，95%CI：1.22 ~ 1.99，P < 0.05）以及术后发生肺部以外的并发症（OR= 2.23，95%CI：1.05 ~ 4.75，P < 0.05）是PD术后肺部感染的独立危险因素。预测模型AUC为0.82（95%CI：0.76 ~ 0.88），校准曲线显示拟合良好，内部验证后AUC为0.80（95%CI：0.71 ~ 0.88），模型有较好的预测能力。 结论 高BMI、术后白蛋白水平低、术后腹腔引流管留置时间长、术后卧床时间长及术后发生肺部以外的并发症是PD术后肺部感染的独立危险因素，构建的风险预测模型对PD患者术后肺部感染的发生具有良好的预测效能。

Objective To investigate the risk factors associated with postoperative pulmonary infection after pancreaticoduodenectomy （PD） and construct a predictive model for guiding early clinical intervention. Methods A retrospective analysis was carried out on 220 patients who underwent PD at the First Affiliated Hospital of Xinjiang Medical University between January 2022 and October 2024. Cases with preoperative pulmonary infection， incomplete data， or perioperative mortality were excluded. Preoperative， intraoperative， and postoperative clinical data were meticulously collected. Univariate analysis was employed to screen potential risk factors， and multivariate logistic regression was utilized to identify independent risk factors. Subsequently， a nomogram prediction model was constructed. Results Pulmonary infection occurred in 84 patients （38.2%） following surgery. Multivariate analysis indicated that a high body mass index （BMI） （OR = 1.12， 95%CI： 1.02 ~ 1.23， P < 0.05）， low albumin levels on postoperative day 3 （OR = 0.91， 95% CI： 0.83 ~ 0.99， P < 0.05）， an extended retention time of the postoperative abdominal drainage tube （OR = 1.05， 95% CI： 1.01 ~ 1.09， P < 0.05）， a prolonged postoperative bed rest duration （OR = 1.56， 95% CI： 1.22 ~ 1.99， P < 0.05）， and the occurrence of non?pulmonary complications （OR = 2.23， 95% CI： 1.05 ~ 4.75， P < 0.05） were independent risk factors for pulmonary infection. The prediction model attained an area under the receiver operating characteristic curve （AUC） of 0.82 （95% CI： 0.76 ~ 0.88）， with the calibration curve showing a good fit. After internal validation， the AUC remained stable at 0.80 （95% CI： 0.71 ~ 0.88）， validating the robust predictive ability of the model. Conclusions Elevated BMI， low postoperative albumin levels， extended retention time of the abdominal drainage tube， prolonged duration of bed rest， and non?pulmonary complications are identified as independent risk factors for pulmonary infection following PD. The established risk prediction model demonstrates robust predictive capabilities， thereby furnishing a foundation for individualized risk assessment and targeted preventive strategies.