Abstract
Background In elderly patients with venous thromboembolism (VTE), the decision to extend anticoagulation beyond 3 months must be weighed against the bleeding risk. We compared the predictive performance of 10 clinical bleeding scores (VTE-BLEED, Seiler, Kuijer, Kearon, RIETE, ACCP, OBRI, HEMORR2HAGES, HAS-BLED, ATRIA) in elderly patients receiving extended anticoagulation for VTE.
Methods In a multicenter Swiss cohort study, we analyzed 743 patients aged ≥65 years who received extended treatment with vitamin K antagonists after VTE. The outcomes were the time to a first major and clinically relevant bleeding. For each score, we classified patients into two bleeding risk categories (low/moderate vs. high). We calculated likelihood ratios and the area under the receiver operating characteristic (ROC) curve for each score.
Results Over a median anticoagulation duration of 10.1 months, 45 patients (6.1%) had a first major and 127 (17.1%) a clinically relevant bleeding. The positive likelihood ratios for predicting major bleeding ranged from 0.69 (OBRI) to 2.56 (Seiler) and from 1.07 (ACCP) to 2.36 (Seiler) for clinically relevant bleeding. The areas under the ROC curves were poor to fair and varied between 0.47 (OBRI) and 0.70 (Seiler) for major and between 0.52 (OBRI) and 0.67 (HEMORR2HAGES) for clinically relevant bleeding.
Conclusion The predictive performance of most clinical bleeding risk scores does not appear to be sufficiently high to identify elderly patients with VTE who are at high risk of bleeding and who may therefore not be suitable candidates for extended anticoagulation.
Keywords
anticoagulants - bleeding - elderly - risk scores - venous thromboembolism
