Predictive value of SinoSCORE on in-hospital mortality and postoperative complications after coronary artery bypass surgery / 中华胸心血管外科杂志

ABSTRACT

Objective To evaluate the performance of the Sino System for Coronary Operative Risk Evaluation ( SinoSCORE) on in-hospital mortality and postoperative complications in patients undergoing coronary artery bypass grafting (CABG) in a single heart center. Methods From January 2007 to December 2008, clinical information of 201 consecutive patients undergoing isolated CABG in our hospital was collected. The SinoSCORE was used to predict hospital mortality and major complications[sternal wound infection, postoperative renal failure, multiple organ dysfunction syndrome, perioperative intra-aortic balloon pumps ( IABP), etc.]after CABG among our study participants, which was initially designed as CBAG operative risk scoring system and included 11 risk factors. We estimated the predictable capability of SinoORE model by the means of analysing the calibration and discrimination characters of this risk scoring system. Calibration was evaluated with the method of Hosmer-Lemeshow goodness-of-fit test. Discrimination was tested by determining the area under the receiver operating characteristic (ROC) curve. The optimal cut-off point for SinoSCORE predicting major complications was obtained by the Youden index. Results Of all our study patients, the observed in-hospital mortality was 1.99% (4/201). The overall mean baseline age was ( 63.3 ± 9.2 ) years and 24.4% ( 49/201 ) were female. The predicted mortality cakulated by the SinoSCORE was 2.88% which was slightly higher than the actual mortality. SinoSCORE model slwed very high discriminatory ability and the good calibration power in predicting in-hospital mortality Hosmer-Lemeshow calibration testx2 =4. 304, P =0.744 and area under ROC was 0. 81 (95% CI 0.687 -0.932). As for the major postoperative complications after CABG,SinoSCORE model still achieved a satisfactory performance with the good predictive value for the main complications risk evaluation such as postoperative renal failure, multiple organ dysfunction syndrome and IABP. Hosmer-Lemeshow P =0.75, P =0. 75, P = 1.00; Areas under ROC respectively at was 0. 768 ( 95 % CI 0.613 - 0. 924 ). 0. 832 ( 95 % CI 0. 732 - 0. 932 )and 0. 737 (95% CI 0. 607 -0. 867 ). The optimal cut-off points for SinoSCORE model predicting each of the above three major postoperative complications was ultimately determined to be 4.5. Conclusion In our patient database, the SinoSCORE model proved a good preoperative risk model in predicting both in-hospital mortality and major complications after CABG, which provides a realistic estimation of hospital death and post-operative complications risk for patients undergoing CABG. Sinoscore model is a suitable operative risk estimation system for Chinese CABG patients.