肺功能3、4级慢性阻塞性肺疾病稳定期急性加重风险预测模型的建立与验证

Establishment and validation of risk prediction model for acute exacerbation of grade 3 and 4 chronic obstructive pulmonary disease in stable stage

目的构建肺功能3、4级慢性阻塞性肺疾病(慢阻肺)稳定期急性加重风险预测模型。方法回顾性分析前期已完成的多中心大样本随机对照试验中330例肺功能3、4级COPD稳定期患者随访52周的危险因素及急性加重情况。按7∶3比例随机分为建模组(n=224)和验证组(n=106)。采用单因素和多因素Logistic回归分析法筛选出慢性阻塞性肺疾病急性加重(AECOPD)相关的危险因素,构建急性加重风险预测模型并验证。结果多因素Logistic回归分析显示,前一年急性加重住院次数[OR=1.582,95%CI(1.252~2.161)]、气短评分[OR=1.823,95%CI 1.051~3.051]、FEV_(1)/FVC[OR=0.968,95%CI 0.940~1.002]、是否使用中药治疗[OR=0.395,95%CI 0.209~0.704]、是否使用西医常规治疗[OR=0.449,95%CI 0.206~0.736]是AECOPD的独立危险因素。基于筛选出的危险因素构建风险预测模型,建模组区分度ROC曲线下面积(AUC)为0.753,验证组AUC为0.732;模型的敏感度为73.77%,特异度为65.38%;DCA结果显示,在建模组中预测急性加重预测概率范围约0.21~0.97、在验证组中预测概率范围约0.20~0.79时,使用本模型有较好的收益。结论本研究所建立的AECOPD风险预测模型具有较好的区分度、校准度及临床有效性,可为中医防治AECOPD提供一定参考。

Objective To construct a risk prediction model for acute exacerbation of chronic obstructive pulmonary disease(COPD)in stable phase with pulmonary function grades 3 and 4.Methods The risk factors and acute exacerbations of 330 stable COPD patients with pulmonary function grade 3 and 4 in a multicenter large sample randomized controlled trial were retrospectively analyzed after 52 weeks of follow-up.They were randomly divided into a modeling group(n=224)and a verification group(n=106)at a ratio of 7∶3.Univariate and multivariate Logistic regression analysis were used to screen out the risk factors associated with acute exacerbation of chronic obstructive pulmonary disease(AECOPD),and the acute exacerbation risk prediction model was constructed and verified.Results Multivariate Logistic regression analysis showed that the number of acute exacerbation hospitalization in the previous year[OR=1.582,95%CI 1.252~2.161],shortness of breath(OR=1.823,95%CI 1.051~3.051),forced vital capacity in the first second/forced vital capacity(FEV_(1)/FVC)(OR=0.968,95%CI 0.940~1.002),whether to use traditional Chinese medicine treatment(OR=0.395,95%CI 0.209~0.704),the use of conventional western medicine treatment(OR=0.449,95%CI 0.206~0.736)were independent risk factors for AECOPD.A risk prediction model was constructed based on the screened risk factors.The area under the ROC curve(AUC)of discrimination in the modeling group was 0.753,and the AUC in the validation group was 0.732.The sensitivity of the model was 73.77%and the specificity was 65.38%.The DCA results showed that the prediction probability of acute exacerbation in the modeling group was 0.21~0.97 and that in the verification group was 0.20~0.79.Conclusion The AECOPD risk prediction model established in this study has good differentiation,calibration and clinical effectiveness,and can provide a certain reference for TCM prevention and treatment of AECOPD.