Validation and Comparison of the CRUSADE and ACTION scores to predict the risk of bleeding in patients with acute coronary syndrome without ST segment elevation

Abstract

Objective: to validate and compare the CRUSADE and ACTION bleeding risk scores in patients with acute coronary syndrome without ST-segment elevation (NSTE-ACS) undergoing early invasive coronary stratification in a highly complex center. Methods: an observational, analytical and prospective study of scale validation in a cohort of patients with NSTEACS who underwent early invasive coronary stratification in a cardiovascular center. The CRUSADE and ACTION risk scales were calculated with the baseline characteristics. The occurrence of major bleeding was defined by intracerebral, retroperitoneal hemorrhage, hematocrit drop ? 12% or need for transfusion when the hematocrit was> 3 g/dL or ? 28% or <28% with clinical bleeding document or bleeding not related to revascularization surgery. Results: in this analysis, the highest observed bleeding rate was 4.8%. The presence of major bleeding was associated with a higher probability of mortality during six-month follow-up (OR 5.316, 95% CI 1.77 -15.92, p = 0.002). The calibration of the risk scales was adequate according to the Hosmer-Lemeshow statistic (p> 0.05). Both scales showed good discrimination without statistically significant differences in the intrahospital setting (AUC ROC CRUSADE ACTION 0.77 vs 0.80 p = 0.19) one month (AUC ROC CRUSADE ACTION 0.70 vs 0.75 p = 0.08) and 6 months (AUC ROC CRUSADE ACTION 0.71 vs 0.76 p = 0.09), respectively. Conclusions: the CRUSADE and ACTION risk scales are useful tools for the prediction of bleeding risk in patients with NSTEACS who are undergoing early invasive coronary stratification. These findings suggest the use of the ACTION scale over the CRUSADE risk scale