D'Errigo Paola, Seccareccia Fulvia, Rosato Stefano, Manno Valerio, Badoni Gabriella, Fusco Danilo, Perucci Carlo A
National Centre of Epidemiology, Surveillance and Health Promotion - Istituto Superiore di Sanità, Rome, Italy.
Eur J Cardiothorac Surg. 2008 Mar;33(3):325-33. doi: 10.1016/j.ejcts.2007.12.001. Epub 2008 Jan 16.
To compare the risk-adjustment model empirically derived from the 'Italian CABG Outcome Project' with that of the additive and logistic EuroSCORE in terms of accuracy, predictive power and ability to rank hospital performances.
The Italian CABG model, the logistic and additive EuroSCORE were applied to the Italian CABG population; the observed deaths/expected deaths (O/E) ratios, as obtained by the three models, were computed for each Italian cardiac surgery centre and for six classes of risk-stratified patients. The performance of the three models in predicting the 30-day mortality was formally assessed for calibration (Hosmer-Lemeshow test) and discrimination (ROC area). According to the three models, risk-adjusted mortality rates (RAMR = O/E x Italian CABG population mortality rate) were estimated for each centre; possible differences were detected in the identification of hospitals with mortality rates higher and lower than average.
The Italian CABG model uses fewer variables than the EuroSCORE system (14 vs 17) and exhibits the best performance in terms of discrimination and calibration. Contrary to the other tested models, the logistic EuroSCORE shows a significant Hosmer-Lemeshow test (chi(H-L)(2)=19.30, p<0.0001), indicating unsatisfactory calibration, and a clear predicted death overestimation in each of the considered risk classes (O/E = 0.4). When a proper recalibration procedure is applied, the logistic EuroSCORE performance parameters achieve acceptable levels. The Italian CABG model identified seven centres as having higher than average mortality, while the EuroSCORE identified the same seven centres plus one other. The Italian CABG model identified eight centres with lower than average mortality, five of which were identified by the additive EuroSCORE and four of which were identified by the logistic EuroSCORE. The additive EuroSCORE identified four more and the logistic EuroSCORE three more low mortality centres.
Although this analysis reveals a satisfactory concordance between results from the three models, a detailed comparison shows that the Italian CABG model uses fewer variables and performs better than the others. Nevertheless, when properly recalibrated, the EuroSCORE model can be exported to the Italian population and used to rank hospital performance and evaluate preoperative risk of patients undergoing open-heart surgery.
比较从“意大利冠状动脉搭桥术结果项目”实证得出的风险调整模型与加法和逻辑回归欧洲心脏手术风险评估系统（EuroSCORE）在准确性、预测能力和对医院表现进行排名的能力方面的差异。
将意大利冠状动脉搭桥术模型、逻辑回归和加法EuroSCORE应用于意大利冠状动脉搭桥术人群；计算每个意大利心脏外科中心以及六类风险分层患者通过这三种模型得出的观察死亡数/预期死亡数（O/E）比率。正式评估这三种模型预测30天死亡率的校准情况（Hosmer-Lemeshow检验）和区分能力（ROC曲线下面积）。根据这三种模型，估算每个中心的风险调整死亡率（RAMR = O/E×意大利冠状动脉搭桥术人群死亡率）；检测在识别死亡率高于和低于平均水平的医院方面可能存在的差异。
意大利冠状动脉搭桥术模型使用的变量比EuroSCORE系统少（14个对17个），并且在区分能力和校准方面表现最佳。与其他测试模型不同，逻辑回归EuroSCORE显示出显著的Hosmer-Lemeshow检验结果（χ²(H-L)=19.30，p<0.0001），表明校准不令人满意，并且在每个考虑的风险类别中预测死亡数明显高估（O/E = 0.4）。当应用适当的重新校准程序时，逻辑回归EuroSCORE的性能参数达到可接受水平。意大利冠状动脉搭桥术模型识别出七个死亡率高于平均水平的中心，而EuroSCORE识别出相同的七个中心外加一个。意大利冠状动脉搭桥术模型识别出八个死亡率低于平均水平的中心，其中五个由加法EuroSCORE识别出，四个由逻辑回归EuroSCORE识别出。加法EuroSCORE多识别出四个低死亡率中心，逻辑回归EuroSCORE多识别出三个。
尽管该分析表明三种模型的结果之间存在令人满意的一致性，但详细比较表明，意大利冠状动脉搭桥术模型使用的变量更少，并且比其他模型表现更好。然而，经过适当重新校准后，EuroSCORE模型可以应用于意大利人群，用于对医院表现进行排名以及评估接受心脏直视手术患者的术前风险。
