Abstract 2217: Impact of the ACUITY and TIMI Major Bleeding Definitions on One-Year Mortality in Patients with Acute Coronary Syndromes
Background: Major bleeding is associated with increased mortality in patients with acute coronary syndromes (ACS). Although several major bleeding definitions exist, the predictive accuracy of different scales has rarely been compared. We therefore assessed the impact of major bleeding (MB), defined by the ACUITY and TIMI criteria, on one-year mortality in patients with ACS.
Methods: The ACUITY Trial was a prospective, randomized comparison of: bivalirudin (BIV), heparin or enoxaparin (H) + glycoprotein IIb/IIIa inhibition (GPI), and BIV+GPI in 13,819 moderate and high-risk ACS patients. MB (non-CABG-related) was defined by the ACUITY scale as: intracranial, intraocular, or retroperitoneal bleeding; access site bleeding with intervention; hematoma ≥5 cm; hemoglobin drop ≥3g/dL with source or ≥4g/dL without source; or transfusion (ACUITY-MB) and by the TIMI scale as: intracranial bleeding; hemoglobin drop >5g/dL; or hematocrit drop >15% (TIMI-MB). Separate adjusted Cox models assessed the impact of ACUITY-MB and TIMI-MB on one-year mortality, with major bleeding and myocardial infarction as time-updated covariates.
Results: Of 13,819 patients, ACUITY-MB occurred in 4.7% (n=644) and TIMI-MB in 1.5% (n=205). ACUITY-MB was less frequent for BIV vs. H+GPI (3.0% vs. 5.7 %, p<0.001). TIMI-MB was also less frequent for BIV vs. H+GPI (0.9% vs. 1.9%, p<0.001). Rates of ACUITY-MB and TIMI-MB were similar for H+GPI vs. BIV+GPI (p=ns for both). In a covariate-adjusted multivariable model including ACUITY-MB only, ACUITY-MB was an independent predictor of one-year mortality (hazard ratio [HR] 2.89, 95% confidence interval [CI] 2.24–3.72, p<0.0001). In a covariate-adjusted multivariable model including TIMI-MB only, TIMI-MB was also an independent predictor of one-year mortality (HR 3.60, 95% CI 2.53–5.13, p<0.0001).
Conclusions: ACUITY-MB and TIMI-MB are both independent predictors of one-year mortality in patients with ACS, and occur less frequently in patients treated with BIV compared to H+GPI. Although the hazard of death after TIMI-MB is numerically slightly greater, ACUITY-MB is three-fold more common than TIMI-MB and thus identifies more patients at risk from MB.