医学
列线图
血流动力学
血管成形术
心脏病学
内科学
肺动脉高压
作者
Xin Li,Qian Zhang,Qin Luo,Qing Zhao,Tao Yang,Qixian Zeng,Qi Jin,Anqi Duan,Zhihua Huang,Meixi Hu,Sicheng Zhang,Luyang Gao,Changming Xiong,Zhihui Zhao,Zhihong Liu
标识
DOI:10.1016/j.rmed.2023.107440
摘要
Background Balloon pulmonary angioplasty (BPA) can effectively alleviate pulmonary hypertension in patients with chronic thromboembolic pulmonary hypertension (CTEPH). Identifying predictors of unfavorable hemodynamic response to BPA is essential to guide clinical practice. Therefore, our objective was to construct and validate a nomogram to facilitate clinicians predicting hemodynamic response to BPA. Methods Patients with CTEPH and underwent BPA from May 2018 to April 2022 were retrospectively collected. Favorable hemodynamic response to BPA was defined as a mean pulmonary arterial pressure ≤30 mmHg and/or a reduction in pulmonary vascular resistance ≥30 % at follow-up. Results A total of 155 patients were included. At baseline, patients with favorable hemodynamic response had significantly lower proportion of occlusive lesions (11.11 % vs. 26.32 %, P = 0.017), higher diffusing capacity of the lungs for carbon monoxide (63.77 % ± 14.10 % vs. 59.11 % ± 11.78 %, P = 0.039), and better cardiac morphology than counterparts. LASSO regression and random forest were used to construct prediction models respectively. The LASSO regression model demonstrated better predictive ability and accuracy than the random forest model, as evidenced by higher area under curve (0.745 vs. 0.740) and lower Brier score (0.192 vs. 0.195). A nomogram was constructed based on the LASSO regression model, consisting of right ventricular end-diastolic diameter/left ventricular end-diastolic diameter, number of treated pulmonary vessels and proportion of occlusion lesions. High predictive ability of the LASSO model was preserved in validation (C index 0.744). Conclusions The current study constructed a nomogram with high accuracy in predicting BPA hemodynamic outcome, which could facilitate decision-making in clinical practice.
科研通智能强力驱动
Strongly Powered by AbleSci AI