P4619Comparison of the CRUSADE, ACUITY-HORIZONS, and ACTION bleeding risk scores for predicting in-hospital bleeding in acute myocardial infarction patients undergoing primary PCI
Abstract Background Considering clinical importance of bleeding complications in patients with acute myocardial infarction (AMI), bleeding risk stratification is a key part of the management of these patients. CRUSADE, ACTION and ACUITY-HORIZONS bleeding risk scores are available for predicting in-hospital major bleeding events in patients with acute myocardial infarction. Purpose We aimed to evaluate performance of the three above mentioned risk scores for predicting in-hospital bleeding events defined according to The Bleeding Academic Research Consortium (BARC) criteria. Methods From a prospective electronic registry of a high-volume catheterization laboratory in a period from January 2009 to December 2017, a total of 6505 consecutive patients with acute myocardial infarction who underwent pPCI were included in analysis. Calibration and discrimination of the three risk models were evaluated by the Hosmer-Lemeshow (H-L) goodness-of-fit test and C-statistic, respectively. Results Overall there were 372 (5.7%) bleeding events out of which 117 (1.8%) fulfilled stage BARC 3 or higher bleeding criteria. All three scores showed good model calibration as assessed by the H-Ls test and very good discriminative power for BARC 3 of higher bleeding events detection as assessed by C-statistics (Table 1 & Figure 1): Bleeding events stage BARC 3 or higher were statistically highly related with higher in-hospital mortality (13.7% vs. 3.5%; p<0.000). Table 1 Risk score H-L H-L p AUC 95% CI p CRUSADE 11.46 0.177 0.761 0.750–0.771 vs. ACUITY = ns vs. ACTION <0.000 ACUITY-HORIZONS 10.47 0.236 0735 0.724–0.745 vs. ACTION = ns ACTION 5.74 0.677 0.701 0.698–0.712 Figure 1 Conclusions All three evaluated scores showed very good discriminative capacity for predicting BARC 3 or higher bleeding events in patients undergoing pPCI for AMI.