天津医药

天津医药 ›› 2025, Vol. 53 ›› Issue (10): 1061-1065.doi: 10.11958/20251088

• 临床研究 • 上一篇    下一篇

慢性阻塞性肺疾病虚弱患者90 d内再入院的影响因素及预警模型的构建

刘欣1(), 谷静1, 杨洁2, 陈新苹1, 蔡传凤1, 张秀梅1,△()   

  1. 1 安徽医科大学第一附属医院老年呼吸科（邮编 230022）
    2 中国科学院合肥肿瘤医院肺部肿瘤科
  • 收稿日期:2025-03-14 修回日期:2025-08-01 出版日期:2025-10-15 发布日期:2025-10-12
  • 通讯作者: △E-mail：1843832021@qq.com
  • 作者简介:刘欣（1984），女，主治医师，主要从事老年呼吸系统疾病研究。E-mail：lx30wzx@163.com
  • 基金资助:
    安徽省高校科研项目(2022AH051128);中华医学会2022年度科研课题(ZHKY202211)

The influencing factors of readmission within 90 days for weak patients with chronic obstructive pulmonary disease and construction of an early warning model

LIU Xin1(), GU Jing1, YANG Jie2, CHEN Xinping1, CAI Chuanfeng1, ZHANG Xiumei1,△()   

  1. 1 Department of Senile Inhalation, the First Affiliated Hospital of Anhui Medical University, Hefei 230022, China
    2 Department of Pulmonary Oncology, Hefei Cancer Hospital, Chinese Academy of Sciences
  • Received:2025-03-14 Revised:2025-08-01 Published:2025-10-15 Online:2025-10-12
  • Contact: △E-mail：1843832021@qq.com

RichHTML

1

PDF (PC)

8

摘要：

目的 探究慢性阻塞性肺疾病（COPD）虚弱患者90 d内再入院的影响因素，构建预警模型并评估区分度与有效性。方法 纳入831例COPD且虚弱患者并根据其出院后90 d内是否再入院分为再入院组（290例）和对照组（541例）。收集并比较2组患者的电子病历数据，多因素Logistic回归分析筛选出COPD虚弱患者90 d内再入院的独立影响因素。构建预警模型，绘制列线图，通过受试者工作特征（ROC）曲线评估预警模型的区分度，通过决策曲线、校准曲线评估预警模型的临床实用性和校准度。结果 再入院组体质量指数（BMI）低于对照组，过去1年COPD急性加重住院次数≥2次占比、埃德蒙顿衰弱量表（REFS）评分、查尔森合并症指数（CCI）评分高于对照组（P<0.05）。多因素Logistic回归分析显示，BMI（OR=0.721，95%CI：0.636~0.818）、过去1年COPD急性加重住院次数（OR=3.040，95%CI：1.944~4.753），REFS评分（OR=1.726，95%CI：1.486~2.005）、CCI评分（OR=3.917，95%CI：3.079~4.983）均是COPD虚弱患者90 d内再入院的独立影响因素（P<0.05）。基于多因素Logistic回归分析结果构建列线图预警模型，ROC曲线下面积为0.847（95%CI：0.820~0.874），敏感度为64.50%，特异度为87.20%。预警模型校准曲线贴近对角线，校准度良好。决策曲线提示预警模型具有临床实用性。结论 基于COPD虚弱患者90 d内再入院的影响因素构建的预警模型具有良好的预测效果和临床实用性。

关键词: 肺疾病, 慢性阻塞性, 虚弱, 病人再入院, 影响因素分析

Abstract:

Objective To explore the influencing factors of readmission within 90 days in frail patients with chronic obstructive pulmonary disease (COPD), construct an early warning model and evaluate its discrimination and effectiveness. Methods A total of 831 COPD patients with frailty were included and divided into the readmission group (290 cases) and the control group (541 cases) based on whether they were readmitted within 90 days after discharge. The electronic medical record data were collected and compared between the two groups of patients. Multivariate Logistic regression analysis was used to screen independent influencing factors for readmission within 90 days in COPD weak patients. A predictive model was established. The column chart was drawn. The discriminability of the predictive model was evaluated through receiver operating characteristic (ROC) curves. The clinical practicality and the predictive model calibration were evaluated through decision curves and calibration curves. Results The body mass index (BMI) of the readmission group was lower than that of the non readmission group, and the proportion of acute exacerbation of COPD hospitalizations ≥ 2 times in the past year， the Edmonton frailty scale (REFS) score and the Chalson comorbidity index (CCI) were higher in the readmission group than those of the non readmission group (P<0.05). Multivariate Logistic regression analysis showed that BMI (OR=0.721, 95%CI: 0.636-0.818), the number of acute exacerbations of COPD hospitalizations in the past year (OR=3.040, 95%CI: 1.944-4.753), REFS score (OR=1.726, 95%CI: 1.486-2.005) and CCI score (OR=3.917, 95%CI: 3.079-4.983) were all independent influencing factors for readmission within 90 days in COPD frail patients (P<0.05). Based on the results of multiple factor Logistic regression analysis, the column chart warning model was constructed. The AUC value of the ROC curve was 0.847 (95%CI: 0.820-0.874), the specificity was 87.20% and the sensitivity was 64.50%. The calibration curve of the predictive model was close to the diagonal, and the calibrationa degree was good. The decision curve suggested that the prediction model had clinical practicality. Conclusion The early warning model constructed based on the influencing factors of readmission hospitalization within 90 days in weak patients with COPD has good predictive effect and clinical practicability.

Key words: pulmonary disease, chronic obstructive, asthenia, patient readmission, root cause analysis

中图分类号: