Outcomes of perioperative intra-aortic balloon pump support in patients undergoing aortic valve surgery: A single-centre, retrospective, observational study.

Patients undergoing aortic-valve surgery are prone to perioperative low-cardiac-output syndrome or progression to cardiogenic shock, which remains a challenging clinical problem. This study aimed to investigate the efficacy of perioperative intra-aortic balloon pump (IABP) implantation following aortic valve surgery and its function in patients with severe (ejection fraction [EF]<35%) and moderate (EF>35%) left ventricular (LV) insufficiency. A total of 120 patients who had undergone mitral valve surgery and IABP insertion under cardiopulmonary bypass at Guangdong Provincial People's Hospital between June 2017 and June 2021 were enrolled in this retrospective analysis. The patients were divided into early and late IABP insertion groups based on the timing of the insertion. The primary endpoint of the observation was the survival rate of patients in all groups. Between January 2017 and January 2021, 120 patients received an IABP in the perioperative period of aortic valve surgery. A control group of 193 patients with relatively severe disease was selected. The mean duration of cardiopulmonary bypass and aortic-cross clamping, time in the intensive care unit, and in-hospital time were longer, and the alanine aminotransferase levels were higher on postoperative days 1 and 3 In the IABP than in the no-IABP group (p<0.05). After propensity score matching, significantly elevated in-hospital mortality (p=0.008) and incidence of dialysis (p=0.008) were observed in the IABP group. In univariate and multivariate logistic assessments, for patients with LVEF >35%, no IABP and intraoperative IABPs were protective factors against mortality (odds ratio [OR]: 0.09, 95% confidence interval [CI]: 0.03-0.33], p=0.000; OR: 0.11, 95%CI: 0.03-0.47, p=0.003, respectively). In addition, we employed several machine learning models, including Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Decision Tree (DT), and Gaussian Naïve Bayes (GNB), which also yielded satisfactory predictive performance. There was no perioperative benefit of IABPs in patients with LVEF <35% after aortic valve surgery. In patients with LVEF >35%, IABP confers only limited peri-operative benefit when implanted late.