欧洲心脏手术危险评分系统评分在心脏术后重症监护中的预测作用

Predictive effect of European system for cardiac operative risk evaluation for intensive care in cardiac surgery

目的探讨欧洲心脏手术危险评分系统(EuroSCORE)评分在心脏术后重症监护中的预测作用。方法对492例成人心脏病手术的患者进行术前EuroSCORE评分,根据additive EuroSCORE评分结果将患者分为低危组、中危组和高危组,比较3组间术后死亡率、严重并发症发生率、术后心脏外科重症监护室(CSICU)入住时间及医疗费用。采用受试者工作特征(ROC)曲线分析及Spearman相关系数分析EuroSCORE评分与死亡率、严重并发症、CSICU入住天数及医疗费用的相关性。结果低、中、高危组间死亡率、严重并发症发生率、CSICU入住天数及费用间差异有统计学意义(P<0.05)。Additive及logistic EuroSCORE分值对死亡率的ROC下面积分别为0.803和0.793,对术后严重并发症的ROC下面积分别为0.824和0.829。Additive和logistic EuroSCORE与术后CSICU入住天数Spearman相关系数分别为0.673和0.683(P<0.01),与CSICU医疗费用Spearman相关系数为0.654和0.657(P<0.01)。结论EuroSCORE评分高估了术后死亡率,但仍可有效应用于心外重症监护以评估术后风险。

Objective To evaluate the predictive value of European System for cardiac operative risk evaluation （EuroSCORE） in intensive care of heart surgery. Methods According to the results of additive EuroSCORE, 492 cases of adult cardiac surgery patients were divided into low-risk group, medium-risk group and high-risk group. The mortality rate of postoperative complications, length of stay and cost in cardiac surgical intensive care unit （CSICU） were compared among three groups. Spearman correlation analysis was used to test the associations between the length of stay, the costs in CSICU and EuroSCORE. The discriminatory power of the additive and logistic EuroSCORE were evaluated by calculating the areas under receiver operating characteristics （ROC） curves. Results There were statistical significances of the differences in the observed mortality rate, the severe postoperative complication rate, the length of stay and the cost in CSICU among groups （P 〈0.05 or 0.01）. The area under ROC curve were 0.803 （additive） and 0.793 （logistic） for postoperative mortality, and 0.824 and 0.829 for the severe complications. The spearman correlation coefficients about additive and logistic EuroSCORE with length of stay in CSICU were 0.673 and 0.683, and costs were 0.654 and 0.657（P 〈0.01）. Conclusion Although EuroSCORE system overestimates the mortality of patients, it still performs well with good discrimination in predicting the postoperative risk.