EuroSCORE II is a poor predictor of peroperative outcomes in octogenarians after cardiac surgery

Results We included 192 octogenarians (83±3y, 53% male, EF 61 ±13%. ES I=10,1 [7,9-17,2] , ES II=3,6 [2,6-5,5]. Surgeries included CABG (35,9%), Valve (41,1%), CABG + valve (44, 22,9%). The observed in-hospital outcomes were: early reoperation (5,2%), MACCE (12,5%), in-hospital mortality (12,0%). ES II was higher in CABG + valve group compared to isolated CABG and valve groups (6,1±4,8% vs 4,2±3,6% e 4,5±3,4%, p=0,028). Corresponding observed mortality rates were 20,5%, 11,6%, and 7,6% (p=0,108). Clinical performance of ES II was poor. O/E ratio was: entire cohort (2,55), CABG (2,76), Valve (1,69), CABG + valve (3,36), p 15). Calibration of ES II in these strata were 0; 2,88; 2,50; e 2,6, respectively (Hosmer-Lemeshow p<0,05). Accuracy of ES II was a bit higher than ES I, but both scores showed poor discriminatory power (AUC ROC 0,686 e 0,647, respectively).