目的:探讨基于风险因素构建的预测模型,预测糖尿病患者感染SARS-CoV-2 (新型冠状病毒)的预后。方法:回顾性分析了239例2022年12月至2023年1月重庆医科大学附属第二医院收治的确诊为SARS-CoV-2感染的糖尿病住院患者。通过电子病历系统收...目的:探讨基于风险因素构建的预测模型,预测糖尿病患者感染SARS-CoV-2 (新型冠状病毒)的预后。方法:回顾性分析了239例2022年12月至2023年1月重庆医科大学附属第二医院收治的确诊为SARS-CoV-2感染的糖尿病住院患者。通过电子病历系统收集患者的相关资料。其中死亡43例,好转出院196例。将患者分为死亡组及存活组,采用单因素Logistic回归分析筛选相关候选因子,再通过多因素Logistic回归分析构建预测模型,使用Bootstrap方法进行内部验证,利用校准曲线、DCA曲线对模型性能进行评估。结果:通过Logistic回归分析,最终有5个因素纳入预测模型,分别为C反应蛋白(CRP) [OR 1.01 (95% CI 1.02)]、白介素6 [OR 1.01 (95% CI 1.01)]、白介素10 [OR 1.04 (95% CI 1.08)]、血氯[OR 0.87 (95% CI 0.99)]、血钠[OR 1.31 (95% CI 1.55)]。经验证该模型在内部验证队列中表现良好。结论:我们通过Logistic回归分析构建的糖尿病患者感染SARS-CoV-2的预后预测模型能够可视化预测结果,并且具有较好的预测效能,对临床制定干预决策有指导意义。Objective: To explore a predictive model based on risk factors to predict the prognosis of diabetic patients infected with SARS-CoV-2. Methods: 239 diabetic inpatients with SARS-CoV-2 infection admitted to the Second Affiliated Hospital of Chongqing Medical University from December 2022 to January 2023 were analyzed retrospectively. The relevant data of the patients was collected by the electronic medical record system. The patients were divided into a death group and a survival group, and the relevant candidate factors were screened using univariate logistic regression analysis, and then the prediction model was constructed by multi-factor logistic regression analysis. The model was internally verified by the Bootstrap method, and the performance of the model was evaluated by the calibration curve and DCA curve. Results: Through logistic regression analysis, five factors were included in the prediction model: C-reactive protein (CRP) [OR 1.01 (95% CI 1.02)], IL-6 [OR 1.01 (95% CI 1.01)], IL-10 [OR 1.04 (95% CI 1.08)], and blood chlorine [OR 0.87 (95% CI 0.99)] and serum sodium [OR 1.31 (95% CI 1.55)]. It is proved that the model performs well in the internal verification queue. Conclusion: The prognosis prediction model of diabetic patients infected with SARS-CoV-2 constructed by Logistic regression analysis can visually predict the results, and has a good predictive efficiency, which is of guiding significance for clinical intervention decision-making.展开更多
文摘目的:探讨基于风险因素构建的预测模型,预测糖尿病患者感染SARS-CoV-2 (新型冠状病毒)的预后。方法:回顾性分析了239例2022年12月至2023年1月重庆医科大学附属第二医院收治的确诊为SARS-CoV-2感染的糖尿病住院患者。通过电子病历系统收集患者的相关资料。其中死亡43例,好转出院196例。将患者分为死亡组及存活组,采用单因素Logistic回归分析筛选相关候选因子,再通过多因素Logistic回归分析构建预测模型,使用Bootstrap方法进行内部验证,利用校准曲线、DCA曲线对模型性能进行评估。结果:通过Logistic回归分析,最终有5个因素纳入预测模型,分别为C反应蛋白(CRP) [OR 1.01 (95% CI 1.02)]、白介素6 [OR 1.01 (95% CI 1.01)]、白介素10 [OR 1.04 (95% CI 1.08)]、血氯[OR 0.87 (95% CI 0.99)]、血钠[OR 1.31 (95% CI 1.55)]。经验证该模型在内部验证队列中表现良好。结论:我们通过Logistic回归分析构建的糖尿病患者感染SARS-CoV-2的预后预测模型能够可视化预测结果,并且具有较好的预测效能,对临床制定干预决策有指导意义。Objective: To explore a predictive model based on risk factors to predict the prognosis of diabetic patients infected with SARS-CoV-2. Methods: 239 diabetic inpatients with SARS-CoV-2 infection admitted to the Second Affiliated Hospital of Chongqing Medical University from December 2022 to January 2023 were analyzed retrospectively. The relevant data of the patients was collected by the electronic medical record system. The patients were divided into a death group and a survival group, and the relevant candidate factors were screened using univariate logistic regression analysis, and then the prediction model was constructed by multi-factor logistic regression analysis. The model was internally verified by the Bootstrap method, and the performance of the model was evaluated by the calibration curve and DCA curve. Results: Through logistic regression analysis, five factors were included in the prediction model: C-reactive protein (CRP) [OR 1.01 (95% CI 1.02)], IL-6 [OR 1.01 (95% CI 1.01)], IL-10 [OR 1.04 (95% CI 1.08)], and blood chlorine [OR 0.87 (95% CI 0.99)] and serum sodium [OR 1.31 (95% CI 1.55)]. It is proved that the model performs well in the internal verification queue. Conclusion: The prognosis prediction model of diabetic patients infected with SARS-CoV-2 constructed by Logistic regression analysis can visually predict the results, and has a good predictive efficiency, which is of guiding significance for clinical intervention decision-making.
