目的：研究表明，我国糖尿病患者肺炎高发，但早期诊断、治疗效果以及病毒感染检测能力均有所欠缺，且临床上CURB-65评分和肺部感染严重程度指数（PSI）评分系统在T2DM患者中的准确性和适用性降低。本研究旨在研究需机械通气支持的肺部感染合并2型糖尿病患者的住院死亡相关风险因素，并构建列线图。

方法：采用单中心回顾性研究进行分析。主要结果为发生院内死亡。通过单因素logistic回归分析和多因素logistic回归分析确定独立风险因素，采用多因素logistic回归分析模型构建评分模型，并绘制决策曲线分析(DCA)曲线及临床决策曲线。

结果：由住院病历系统中筛选出2132例需机械通气支持的肺部感染患者，其中403例合并2型糖尿病。根据患者是否发生住院死亡，将403例患者分为两组：死亡组和生存组；分组后分析出18个参数具有统计学意义。通过多因素4个参数是死亡率的独立预测因素，包括：年龄、合并休克以及治疗期间监测到的肌酐和凝血酶原时间的最大值。基于这些变量，建立了风险预测评分。评分模型的ROC=0.821 95% CI: 0.768-0.874。

结论：本项研究发现年龄增高、合并血流动力学不稳定、肌酐升高及凝血酶原时间延长成为需机械通气支持的肺部感染合并2型糖尿病患者的住院死亡的独立风险因素并能够预测住院死亡发生风险。通过本研究开发的评分模型来明确需机械通气支持的肺部感染合并2型糖尿病患者的临床特征，帮助临床上识别高危患者，预测患者结局。

Objective: With a high incidence of pneumonia, low early diagnosis rate, low disease control rate, low pathogen detection rate, and high case fatality rate in Chinese patients with diabetes mellitus, the accuracy and applicability of CURB-65 score and pulmonary infection Severity Index (PSI) scoring system in clinical patients with T2DM are reduced. The objective of this study was to investigate the risk factors associated with in-hospital death in patients with pulmonary infection and type 2 diabetes requiring mechanical ventilation support and to construct a graph.

Methods: A single-center retrospective study (Jan. 1, 2015, to Dec. 31, 2020) was conducted to analyze the patients with pulmonary infection complicated with type 2 diabetes requiring mechanical ventilation in Peking Union Medical College Hospital. The main outcome was hospital death and death. The independent risk factors were determined by univariate logistic regression analysis and multivariate logistic regression analysis. The multivariate logistic regression analysis model was used to construct the scoring model, and the decision curve analysis (DCA) curve and clinical decision curve were drawn.

Results: A total of 2132 patients with pulmonary infection requiring mechanical ventilation were selected from the inpatient medical record system, and 403 of them were complicated with type 2 diabetes. According to the patient's treatment results, 403 patients were divided into two groups: death group and survival group; After grouping, 18 parameters were found to be statistically significant. Four parameters were independent predictors of mortality by multiple factors, including age, concomitant shock, and maximum creatinine and prothrombin time detected during treatment. Based on these variables, a risk prediction score was established. ROC=0.821 95%CI: 0.768-0.874.

Conclusions: In this study, we found that increasing age, hemodynamic instability, increased creatinine, and prothrombin duration were independent risk factors and predictors of in-hospital death in patients with pulmonary infection and type 2 diabetes requiring mechanical ventilation. The scoring model developed in this study was used to clarify the clinical characteristics of patients with pulmonary infection complicated with type 2 diabetes requiring mechanical ventilation support, to help clinically identify high-risk patients and predict the outcome of patients.