Validation of Bleeding Risk Prediction Scores for Patients With Major Bleeding on Direct Oral Anticoagulants

Background: Direct oral anticoagulants (DOACs) offer many benefits over vitamin K antagonists (VKAs) but still carry a significant risk of major bleeding. Bleeding risk prediction scores such as the Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile International Normalized Ratio, Elderly, and Drugs/Alcohol (HAS-BLED), Hepatic or Renal Disease, Ethanol Abuse, Malignancy, Older Age, Reduced Platelet Count or Function, Re-Bleeding, Hypertension, Anemia, Genetic Factors, Excessive Fall Risk, and Stroke (HEMORR2HAGES), Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA), Registro Informatizado Enfermedad TromboEmbólica (RIETE), and CHEST scores were validated or evaluated for use with VKAs and parenteral anticoagulants, but evidence for use with DOACs is lacking. Objective: This study aims to evaluate bleeding risk prediction scores for DOAC patients presenting with major bleeding. Methods: A retrospective analysis of patients presenting from 2015 to 2018 was performed. Patients were separated into bleed and nonbleed groups. The primary objective was to assess the diagnostic accuracy of the bleeding risk prediction scores utilizing the receiver operating characteristic (ROC) curve. Results: A total of 126 patients were included in the analyses. The areas under the curve (AUC) for the ROC curves of the HAS-BLED, HEMORR2HAGES, ATRIA, RIETE, and CHEST scores were 0.645, 0.675, 0.580, 0.638, and 0.667, respectively. Conclusion and Relevance: The HAS-BLED, HEMORR2HAGES, RIETE, and CHEST scores were found to have sufficient diagnostic accuracy for predicting risk of major bleeding in our study population; however, no score was identified as having an AUC greater than 0.7. Caution may be considered when utilizing these scores for patients on DOACs.