老年慢性阻塞性肺疾病患者并发心房颤动的危险因素分析及列线图模型构建

Analysis of Risk Factors of Atrial Fibrillation in Elderly Patients with Chronic Obstructive Pulmonary Disease and Nomogram Model Construction

摘要: 目的：探讨老年慢性阻塞性肺疾病(Chronic Obstructive Pulmonary Disease, COPD)患者并发心房颤动(Atrial Fibrillation, AF)的独立危险因素，并初步建立老年COPD患者并发AF的列线图风险预测模型。方法：回顾性收集2016年至2020年于某三甲医院住院治疗的老年COPD患者病例资料531例，根据是否发生AF分为COPD并发AF病例组81例和对照患者450例。对两组患者的年龄、性别、吸烟、饮酒等一般情况，高血压病、冠心病、心力衰竭、糖尿病、急性呼吸衰竭、肺部感染、脑卒中、急性心肌梗死、消化道出血以及肾功能不全等合并疾病情况进行统计分析，并使用R语言开发构建列线图预测模型，根据曲线下面积(Area Under Curve, AUC)和Hosmer-Lemeshow (H-L)检验验证模型准确性。结果：两组患者的吸烟、饮酒情况，心力衰竭、糖尿病、急性呼吸衰竭、肺部感染、脑卒中、急性心肌梗死等合并疾病差异有统计学意义(P < 0.05)；多因素Logistic回归结果示：心力衰竭(OR = 1.706)、糖尿病(OR = 1.730)、急性呼吸衰竭(OR = 3.408)、肺部感染(OR = 2.859)、脑卒中(OR = 1.662)、急性心肌梗死(OR = 1.204)是老年COPD患者并发AF的独立危险因素。使用R语言开发列线图模型，受试者工作特征曲线下面积AUC为0.809，H-L检验P = 0.35 (P > 0.05)，发现该模型的区分度和校准度均较好。结论：心力衰竭、糖尿病、急性呼吸衰竭、肺部感染、脑卒中、急性心肌梗死等合并疾病是老年COPD患者并发AF的独立危险因素。本研究中列线图模型准确度较好，在临床工作中可提供一定的参考。

Abstract: Objective: To investigate the independent risk factors of Atrial Fibrillation (AF) in elderly patients with Chronic Obstructive Pulmonary Disease (COPD), and to establish a nomogram risk prediction model for AF in elderly patients with COPD. Methods: A total of 531 elderly patients with COPD hospitalized in a tertiary hospital from 2016 to 2020 were retrospectively collected. According to the occurrence of AF, they were divided into COPD complicated with AF case group (81 cases) and control group (450 cases). The age, gender, smoking, drinking and other general conditions of the two groups, hypertension, coronary heart disease, heart failure, diabetes, acute respiratory failure, pulmonary infection, stroke, acute myocardial infarction, gastrointestinal bleeding and renal insufficiency were statistically analyzed, and the nomogram prediction model was developed by R language. Area Under Curve (AUC) and Hosmer-Lemeshow (H-L) were used to test the accuracy of the model. Results: There were significant differences in smoking, drinking, heart failure, diabetes, acute respiratory failure, pulmonary infection, stroke, acute myocardial infarction and other comorbid diseases between the two groups (P < 0.05); multivariate logistic regression results showed that: heart failure (OR = 1.706), diabetes (OR = 1.730), acute respiratory failure (OR = 3.408), pulmonary infection (OR = 2.859), heart failure (OR = 1.706), diabetes (OR = 1.730), acute respiratory failure (OR = 3.408), pulmonary infection (OR = 2.859), stroke (OR = 1.662) and acute myocardial infarction (OR = 1.204) were independent risk factors for AF in elderly COPD patients. R language was used to develop the nomogram model. The AUC of the area under the curve was 0.809, and the H-L test was P = 0.35 (P > 0.05). Conclusion: Heart failure, diabetes, acute respiratory failure, pulmonary infection, stroke and acute myocardial infarction are independent risk factors of AF in elderly COPD patients. In this study, the accuracy of nomogram model is good, which can provide some reference in clinical work.