【摘要】 目的：探讨体外循环（CPB）心脏瓣膜术后肝功能损伤的危险因素，根据这些危险因素建立预测模型，以期为术后肝功能损伤的预测、诊断与治疗提供理论依据。方法：回顾性分析我院心脏大血管外科2018年7月至2020年10月，CPB下行心脏直视瓣膜置换或成形手术的160例成人患者，统计其人口学资料及相关的临床资料。按照术后48h内转氨酶或胆红素任意一项超过正常即为肝功能损伤，将患者分为肝损伤组与非肝损伤组，先进行单因素分析，再对筛选出较有意义的变量进行多因素Logistic分析建立回归模型，并使用受试者工作特征曲线（ROC曲线）评价预测模型。结果：(1) 160例心脏瓣膜患者术后出现肝功能损伤128例，发生率为80%，其中1级损伤48例，2级损伤71例，3级损伤9例，无4级损伤；院内死亡2例，死亡率为1.2%。(2)单因素分析提示：术前丙氨酸氨基转移酶（ALT）水平、术前总胆红素（TBIL）水平、三尖瓣反流程度、CPB时间、主动脉阻断时间与CPB心脏瓣膜术后肝功能损伤有关（P <0.05）。多因素Logistic回归分析显示：术前TBIL水平（OR=1.152,95%CI:1.014～1.309,P=0.03）、CPB时间（OR=1.072,95%CI:1.025～1.121,P=0.002）、主动脉阻断时间（OR=0.945,95%CI:0.900～0.992,P=0.021）是CPB心脏瓣膜术后肝功能损伤的危险因素。(3)双瓣置换或成形患者比单瓣患者更容易发生术后肝功能损伤（P <0.05）；术后发生肝功能损伤患者术后住院时间更长（P <0.05）。(4) ROC的曲线下面积（AUC）为0.814(95%CI:1.025～1.121,P <0.001)，截断点为0.5，模型的敏感度为53.1%，特异度为96.9%。结论：(1)术前TBIL水平、CPB时间、主动脉阻断时间是CPB心脏瓣膜术后肝功能损伤的独立危险因素。(2)双瓣患者比单瓣患者更容易发生术后肝功能损伤；术后发生肝功能损伤患者术后住院时间更长。(3)预测模型的建立对预测术后肝功能损伤的发生有一定价值。更多还原

【Abstract】 Objective: To investigate the risk factors of liver function injury after cardiac valve surgery with cardiopulmonary bypass(CPB), and to establish a prediction model according to these risk factors in order to provide a theoretical basis for the prediction, diagnosis and treatment of postoperative liver function injury. Methods: A total of 160 adult patients who underwent open-heart valve replacement or plasty under CPB from July 2018 to October 2020 in the Department of Cardiovascular Surgery of our hospital were retrospectively analyzed, and their demographic data and relevant clinical data were statistically analyzed.According to any transaminase or bilirubin exceeding normal within 48 hours after surgery, it is considered as liver injury, and the patients were divided into liver injury group and non-liver injury group. Univariate analysis was performed first, and then multivariate Logistic analysis was performed for the variables screened to be more significant to establish a regression model. Receiver operating characteristic curve(ROC curve)was used to evaluate the prediction model. Results:(1) 128 of 160 patients with heart valves had postoperative liver function injury, with an incidence of 80%, including 48 cases of grade 1 injury, 71 cases of grade 2injury, 9 cases of grade 3 injury, and no grade 4 injury; 2 patients died in the hospital, with a mortality rate of 1.2%.(2) Univariate analysis suggested that preoperative alanine aminotransferase(ALT) level, total bilirubin(TBIL) level, degree of tricuspid valve, CPB time, and aortic cross-clamp time were associated with liver function injury after CPB heart valve surgery(P<0.05). Multivariate Logistic regression analysis showed that preoperative TBIL level(OR=1.152, 95% CI:1.014-1.309, P=0.03), CPB time(OR=1.072, 95% CI:1.025-1.121, P=0.002), and aortic cross-clamp time(OR=0.945, 95% CI:0.900-0.992, P=0.021) were independent risk factors for liver function injury after CPB heart valve surgery.(3)Patients with double-valve replacement or plasty were more likely to have postoperative liver function injury than patients with single-valve(P<0.05); patients with postoperative liver function injury had postoperative hospital stay(P<0.05).(4) the area under the curve(AUC) of ROC was 0.814(95% CI: 1.025-1.121,P<0.001), the cutoff point was 0.5, and the sensitivity of the model was 53.1% and the specificity was 96.9%.Conclusions:(1) preoperative TBIL level, CPB time, and aortic cross-clamp time were independent risk factors for liver function injury after CPB heart valve surgery.(2) Patients with double-valve are more likely to have postoperative liver function injury than patients with single-valve; patients with postoperative liver function injury have postoperative hospital stay.(3) the establishment of a prediction model is of some value in predicting the occurrence of postoperative liver function injury.更多还原