射血分数轻度降低心衰患者LVEF恢复的预测模型构建

Construction of a Prediction Model for LVEF Recovery in Patients with Heart Failure with Mildly Reduced Ejection Fraction

摘要: 目的：筛选射血分数轻度降低的心力衰竭(HFmrEF)患者左室射血分数(LVEF)恢复的独立临床预测因素，并构建列线图预测模型。方法：回顾性分析2020年1月至2024年12月青岛大学附属医院心血管内科收治的158例HFmrEF患者。LVEF恢复定义为随访LVEF ≥ 50%且较基线增加 ≥ 5%。采用最小绝对收缩和选择算子(LASSO)回归及多因素Logistic回归筛选独立预测变量，并基于筛选出的变量构建列线图预测模型。使用受试者工作特征(ROC)曲线、校准曲线及决策曲线分析(DCA)评估模型性能。结果：共纳入158例患者，其中66例(41.8%)在随访期间实现LVEF恢复。多因素Logistic回归分析结果显示，纽约心脏病协会(NYHA)心功能分级(OR = 0.219, 95% CI: 0.090~0.532, p < 0.001)、基线收缩压(OR = 1.028, 95% CI: 1.010~1.047, p = 0.003)、左室舒张末期内径(LVEDD) (OR = 0.499, 95% CI: 0.258~0.966, p = 0.039)以及节段性室壁运动异常(RWMA) (OR = 0.310, 95% CI: 0.140~0.684, p = 0.004)是LVEF恢复的独立预测因素。基于上述变量构建的列线图模型区分度良好，ROC曲线下面积(AUC)为0.813，校准曲线显示预测概率与实际观测概率高度一致。结论：NYHA分级、收缩压、LVEDD及RWMA是影响HFmrEF患者LVEF恢复的独立预测因素。本研究构建的列线图模型具有良好的区分度和校准度，有助于临床早期识别具有恢复潜力的患者并指导个体化治疗。

Abstract: Objective: To identify independent clinical predictors of left ventricular ejection fraction (LVEF) recovery in patients with heart failure with mildly reduced ejection fraction (HFmrEF) and to construct a nomogram prediction model. Methods: A retrospective analysis was conducted on 158 HFmrEF patients admitted to the Department of Cardiology, The Affiliated Hospital of Qingdao University from January 2020 to December 2024. LVEF recovery was defined as a follow-up LVEF ≥ 50% with an absolute increase of ≥5% from baseline. Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariable logistic regression were used to screen independent predictive variables, and a nomogram prediction model was constructed based on the selected variables. The performance of the model was evaluated using the Receiver Operating Characteristic (ROC) curve, calibration curve, and Decision Curve Analysis (DCA). Results: A total of 158 patients were included, of whom 66 (41.8%) achieved LVEF recovery during the follow-up period. Multivariable logistic regression analysis showed that New York Heart Association (NYHA) functional class (OR = 0.219, 95% CI: 0.090~0.532, p < 0.001), baseline systolic blood pressure (OR = 1.028, 95% CI: 1.010~1.047, p = 0.003), left ventricular end-diastolic diameter (LVEDD) (OR = 0.499, 95% CI: 0.258~0.966, p = 0.039), and regional wall motion abnormality (RWMA) (OR = 0.310, 95% CI: 0.140~0.684, p = 0.004) were independent predictors of LVEF recovery. The nomogram model constructed based on these variables demonstrated good discrimination, with an Area Under the Curve (AUC) of 0.813. The calibration curve showed high consistency between the predicted probability and the actual observed probability. Conclusion: NYHA classification, systolic blood pressure, LVEDD, and RWMA are independent predictive factors affecting LVEF recovery in HFmrEF patients. The nomogram model constructed in this study possesses good discrimination and calibration, which aids clinicians in the early identification of patients with recovery potential and guides individualized treatment.