天津医药 ›› 2025, Vol. 53 ›› Issue (7): 751-755.doi: 10.11958/20250560

• 临床研究 • 上一篇    下一篇

基于术前肺动脉压评估心脏移植术后右心衰竭的发生风险

陈英1(), 郭长英1, 张静1, 李娟1, 陈丰毅2   

  1. 1 郑州市第七人民医院CSICU(邮编450000)
    2 郑州市第七人民医院心内科(邮编450000)
  • 收稿日期:2025-02-14 修回日期:2025-04-21 出版日期:2025-07-15 发布日期:2025-07-21
  • 作者简介:陈英(1976),女,副主任医师,主要从事心脏移植术后并发症方面研究。E-mail:15038072336@163.com
  • 基金资助:
    河南省医学科技攻关计划联合共建立项项目(LHGJ20200727)

The risk of right heart failure after heart transplantation based on preoperative pulmonary artery pressure assessment

CHEN Ying1(), GUO Changying1, ZHANG Jing1, LI Juan1, CHEN Fengyi2   

  1. 1 Department of CSICU, Zhengzhou Seventh People's Hospital, Zhengzhou 450000, China
    2 Department of Cardiology, Zhengzhou Seventh People's Hospital, Zhengzhou 450000, China
  • Received:2025-02-14 Revised:2025-04-21 Published:2025-07-15 Online:2025-07-21

摘要:

目的 基于术前肺动脉压建立列线图评估心脏移植术后右心衰竭发生风险并进行验证。方法 回顾性纳入接受心脏移植术患者184例,将患者分为训练集(126例)和验证集(58例)。依据术后是否发生右心衰竭将训练集患者分为右心衰组(60例)和非右心衰组(66例),比较2组患者临床资料的差异,对训练集患者运用Lasso-Logistic回归筛选术后发生右心衰竭的影响因素。根据筛选的影响因素绘制列线图,并利用受试者工作特征(ROC)曲线、校准曲线及临床决策曲线评估模型的预测效能。进一步在验证集中评价模型的临床应用效果。结果 Lasso-Logistic回归共筛选出总胆红素升高(OR=2.649,95%CI:1.339~5.239)、平均肺动脉压增大(OR=3.082,95%CI:1.608~5.910)、肺动脉阻力增大(OR=3.171,95%CI:1.710~5.879)、右心室流出道内径增宽(OR=2.681,95%CI:1.361~5.281)是心脏移植患者术后发生右心衰竭的危险因素;依此构建列线图模型,模型的曲线下面积(AUC)为0.846(95%CI:0.813~0.947),校准曲线显示模型拟合良好(Hosmer-Lemeshoe χ2=0.862,P=0.361)。临床决策曲线结果显示,高风险阈值概率在1%~95%时净获益率>0,该列线图模型能产生较好的临床效益。根据列线图模型预测结果,验证集58例患者中34例被预测为右心衰竭高风险,24例为低风险,实际诊断结果显示29例发生右心衰竭,29例未发生,Kappa系数为0.483(95%CI:0.261~0.705)。结论 术前肺动脉收缩压增大是心脏移植患者术后发生右心衰竭的独立危险因素,通过联合其他临床危险因素构建患者术后发生右心衰竭的列线图预测模型具有良好的预测效能。

关键词: 心脏移植, 心力衰竭, 肺动脉压, 预测模型, 列线图

Abstract:

Objective To evaluate the risk of right heart failure after heart transplantation by establishing nomogram based on preoperative pulmonary artery pressure. Methods A total of 184 patients undergoing heart transplantation were retrospectively collected and divided into the training group (126 cases) and the verification group (58 cases). Patients in the training set were divided into the right heart failure group (60 cases) and the non-right heart failure group (66 cases) according to whether right heart failure occurred after operation. The differences of clinical data between the two groups were compared, and the influencing factors of right heart failure occurred after operation in the training set were screened by Lasso-Logistic regression. According to the screened influencing factors, nomograms were drawn, and the predictive efficiency of the model was evaluated by using the receiver's operating characteristic (ROC) curve, calibration curve, receiver's operating characteristic (ROC) curve and clinical decision curve. Further vertification of the clinical application effect of centralized evaluation model was conducted. Results The Lasso-Logistic regression analysis identified the following independent risk factors for right heart failure after heart transplantation: elevated total bilirubin (OR=2.649, 95%CI: 1.339-5.239), increased mean pulmonary artery pressure (OR=3.082, 95%CI: 1.608-5.910), elevated pulmonary artery resistance (OR=3.171, 95%CI: 1.710-5.879), and widened right ventricular outflow tract diameter (OR=2.681, 95%CI: 1.361-5.281), all of which demonstrated statistical significance (P<0.05). The nomogram model was constructed accordingly. The AUC of the nomogram model was 0.846 (95%CI: 0.813-0.947). The calibration curve demonstrated good fit via the goodness-of-fit test (Hosmer-Lemeshow χ2=0.862, P=0.361). Clinical decision curve analysis revealed that the net benefit rate remained >0 when the high-risk threshold probability ranged from 1% to 95%, indicating favorable clinical utility of this nomogram model. Based on the model predictions, among 58 heart transplant patients in the validation cohort, 34 were classified as high-risk for right heart failure and 24 as low-risk. Actual diagnosis results showed 29 cases with right heart failure and 29 without. The Kappa coefficient reached 0.483 (95%CI: 0.261-0.705), demonstrating high consistency between model predictions and actual clinical outcomes. Conclusion Preoperative pulmonary systolic pressure increase is an independent risk factor for right heart failure after heart transplantation. A nomogram prediction model for right heart failure after heart transplantation is established by combining other clinical risk factors, and it has good prediction efficiency.

Key words: heart transplantation, heart failure, pulmonary artery pressure, prediction model, nomogram

中图分类号: