Mastroiacovo Giorgio, Bonomi Alice, Ludergnani Monica, Franchi Matteo, Maragna Riccardo, Pirola Sergio, Baggiano Andrea, Caglio Alice, Pontone Gianluca, Polvani Gianluca, Merlino Luca
IRCCS Centro Cardiologico Monzino, Department of Cardiovascular Surgery, Milan, Italy.
Department of Statistics, IRCCS Centro Cardiologico Monzino, Milan, Italy.
Eur J Cardiothorac Surg. 2022 Jan 1;64(3). doi: 10.1093/ejcts/ezad294.
The European System for Cardiac Operation Risk Evaluation II (EuroSCORE II) is the most common tool used to evaluate the perioperative risk of mortality after cardiac surgery in Europe, and its use is currently recommended by the relevant guidelines. However, recently, its role has been questioned: Several papers have suggested that these algorithms may no longer be adequate for risk prediction due to an overestimation of adult cardiac surgical risk. Our goal was to validate the EuroSCORE II in the prediction of 30-day in-hospital mortality in patients undergoing open cardiac surgery in a high-volume hospital.
In this retrospective cohort study, we included all patients who underwent cardiac surgery from January 2016 to May 2022 within the departments of cardiac surgery of the Monzino Cardiology Centre in Milan, Italy. We evaluated the discrimination power of the EuroSCORE II by using the receiver operating characteristic curve and the corresponding area under the curve. We performed calibration plots to assess the concordance between the model's prediction and the observed outcomes.
A total of 4,034 patients were included (mean age = 65.1 years; 68% males), of which 674 (16.7%) underwent isolated coronary artery bypass grafting. The EuroSCORE II showed a good discrimination power in predicting 30-day in-hospital mortality (area under the curve = 0.834). However, for interventions performed in an elective setting, very low values of the EuroSCORE II overestimated the observed mortality, whereas for interventions performed in an emergency setting, EuroSCORE II values above 10 extensively underestimated the observed mortality.
Our study suggests that the EuroSCORE II seems not to be a reliable score in estimating the true risk of death, especially in high-risk patients.
欧洲心脏手术风险评估系统II（EuroSCORE II）是欧洲用于评估心脏手术后围手术期死亡风险最常用的工具，目前相关指南推荐使用该系统。然而，最近其作用受到了质疑：几篇论文表明，由于对成人心脏手术风险的高估，这些算法可能不再足以用于风险预测。我们的目标是在一家大型医院对接受心脏直视手术患者的30天院内死亡率预测中验证EuroSCORE II。
在这项回顾性队列研究中，我们纳入了2016年1月至2022年5月在意大利米兰蒙齐诺心脏病中心心脏外科接受心脏手术的所有患者。我们通过使用受试者工作特征曲线及相应的曲线下面积来评估EuroSCORE II的辨别能力。我们绘制校准图以评估模型预测与观察结果之间的一致性。
共纳入4034例患者（平均年龄=65.1岁；68%为男性），其中674例（16.7%）接受了单纯冠状动脉旁路移植术。EuroSCORE II在预测30天院内死亡率方面显示出良好的辨别能力（曲线下面积=0.834）。然而，对于择期进行的手术，EuroSCORE II的极低值高估了观察到的死亡率，而对于急诊进行的手术，EuroSCORE II值高于10则广泛低估了观察到的死亡率。
我们的研究表明，EuroSCORE II在估计真正的死亡风险方面似乎不是一个可靠的评分系统，尤其是在高危患者中。
