Predicting Adverse Events beyond Stroke and Bleeding with the ABC-Stroke and ABC-Bleeding Scores in Patients with Atrial Fibrillation: The Murcia AF Project
Anny Camelo‐Castillo, José Miguel Rivera‐Caravaca, Francisco Marı́n, Vicente Vicente, Gregory Y.H. Lip, Vanessa Roldán
Abstract
Abstract Background The ABC (age, biomarkers, and clinical history)-stroke and ABC-bleeding are biomarker-based scores proposed to predict stroke and bleeding, but non-specificity of biomarkers is common, predicting different clinical events at the same time. We assessed the predictive performance of the ABC-stroke and ABC-bleeding scores, for outcomes beyond ischemic stroke and major bleeding, in a cohort of atrial fibrillation (AF) patients. Methods We included AF patients stable on vitamin K antagonists for 6 months. The ABC-stroke and ABC-bleeding were calculated and the predictive values for myocardial infarction (MI), acute heart failure (HF), a composite of cardiovascular events, and all-cause deaths were compared. Results We included 1,044 patients (49.2% male; median age 76 [71–81] years). During 6.5 (4.3–7.9) years, there were 58 (5.6%) MIs, 98 (9.4%) acute HFs, 167 (16%) cardiovascular events, and 418 (40%) all-cause deaths. There were no differences in mean ABC-stroke and ABC-bleeding scores in patients with/without MI (p = 0.367 and p = 0.286, respectively); both scores were higher in patients with acute HF, cardiovascular events, or death (all p < 0.05). Predictive performances for the ABC-stroke and ABC-bleeding scores were similar, ranging from “poor” for MI (c-indexes ∼0.54), “moderate” for acute HF and cardiovascular events (c-indexes ∼0.60 and ∼0.64, respectively), and “good” for all-cause mortality (c-indexes > 0.70). Clinical usefulness whether assessed by ABC-stroke or ABC-bleeding was similar for various primary endpoints. Conclusion In AF patients, the ABC-stroke and ABC-bleeding scores demonstrated similar predictive ability for outcomes beyond stroke and bleeding, including MI, acute HF, a composite of cardiovascular events, and all-cause deaths. This is consistent with nonspecificity of biomarkers that predict “sick” patients or poor prognosis overall.