高龄慢性阻塞性肺疾病稳定期患者近期急性加重风险预测模型构建  

Construction of risk prediction model for acute exacerbation in near future of elderly COPD patients in stable stage

作　　者：韩慧[1] 徐平[1] 宋卫东[1] HAN Hui;XU Ping;SONG Weidong(Department of Respiratory and Critical Care Medicine,Peking University Shenzhen Hospital,Shenzhen 518000,China)

机构地区：[1]北京大学深圳医院呼吸与危重症医学科,广东深圳518000

出　　处：《实用医学杂志》2023年第22期2984-2988,共5页The Journal of Practical Medicine

基　　金：深圳市科技计划项目(编号:201103004)。

摘　　要：目的探讨高龄慢性阻塞性肺疾病(COPD)稳定期患者近期急性加重风险影响因素并构建预测模型。方法回顾性纳入2018年1月至2022年6月于我院诊治高龄COPD稳定期患者共654例,根据随访6个月内是否出现急性加重分为加重组(134例)和未加重组(520例),采用logistic分析高龄COPD稳定期患者近期急性加重风险独立影响因素,构建基于上述独立影响因素指标的预测模型并进一步评估该模型用于近期急性加重风险预测临床效能。结果单因素分析结果显示,性别、体质量指数、COPD病程、FEV1、FVC、近1年急性加重次数及持续时间均可能与高龄COPD稳定期患者近期急性加重发生有关(P<0.05);logistic回归模型多因素分析结果显示,性别、体质量指数、FEV1、FVC、近1年急性加重次数及持续时间均是高龄COPD稳定期患者近期急性加重风险独立影响因素(P<0.05);根据自变量偏回归系数构建模型方程式:近期急性加重风险[Logit(p/1-p)]=-2.981+0.642×性别-0.517×体质量指数+0.316×近1年急性加重次数+0.028×近1年急性加重持续时间-0.621×FEV1。ROC曲线分析结果显示,上述模型用于高龄COPD稳定期患者近期急性加重风险预测曲线下面积为0.84(95%CI:0.79~0.92),最佳cut-off值为0.23,灵敏度和特异度分别为82.71%,70.85%。结论高龄COPD稳定期患者近期急性加重发生与性别、体质量指数、FEV1、FVC、近1年急性加重次数及持续时间等因素有关;基于上述独立影响因素构建预测模型能够准确预测患者近期急性加重发生风险。Objective To investigate the risk factor of acute exacerbation in near future of elderly COPD patients in a stable stage and construct prediction model to provide more reference for early identification of high-risk groups of unplanned hospitalization and formulation of more active and effective intervention programs.Methods 654 elderly patients with COPD in stable stage were retrospectively chosen in the period from January 2018 to June 2022 in our hospital and grouped according to whether acute exacerbation occurred within 6 months of follow-up or not into with exacerbation group(134 cases)and non-exacerbation group(520 cases).Univariate and multivariate methods were used to analyze the independent influencing factors of short-term acute exacerbations risk in elderly COPD stable patients,and the prediction model based on the above independent influencing factors was con-structed and the clinical efficacy of prediction model for short-term acute exacerbations risk was further evaluated.Results Univariate analysis showed that gender,body mass index,cough,FEV1,FVC,the frequency and dura-tion of acute exacerbations in recent 1 year may be related to the occurrence of acute exacerbations in elderly patients with COPD in stable stage(P<0.05).Multivariate analysis results of Logistic regression model showed that gender,body mass index,FEV1,FVC,the number and duration of acute exacerbations in recent 1 year were independent factors influencing the risk of acute exacerbations in elderly patients with COPD in stable stage(P<0.05).The model equation was constructed according to the independent variable partial regression coefficient:near-term acute exacerbation risk[Logit(p/1-p)]=-2.981+0.642×sex-0.517×body mass index+0.316×the number of acute exacerbations in recent 1 year+0.028×duration of acute exacerbation in recent 1 year-0.621×FEV1.ROC curve analysis results showed that the area under the curve for predicting the risk of acute exacerbation in near future of elderly COPD patients in stable stage was 0.84(95%CI:0

关 键 词：慢性阻塞性肺疾病 稳定期 急性加重 预测模型 

分 类 号：R563.9[医药卫生—呼吸系统]