In patients receiving anticoagulation for deep vein thrombosis (DVT), a variety of reasons (including active bleeding or high-risk for bleeding) may lead into premature discontinuation of therapy (prior to completing 90 days). The relative frequency and clinical consequences of premature discontinuation in contemporary patients remain unknown.
We used the data from RIETE, an international registry of patients with venous thromboembolism (VTE), to identify patients with proximal (above knee) lower limb DVT who prematurely discontinued anticoagulation. We assessed the incidence of the composite outcome: pulmonary embolism (PE)-related death, sudden death, or recurrent VTE within the subsequent 30 days after discontinuation and compared the risk of these events vs. the risk in patients without premature discontinuation, once adjusted for demographics and clinical factors.
Of 26,335 patients with proximal DVT recruited from 2001 to 2018, 1322 (5.02%) prematurely discontinued anticoagulation. Thirty days after discontinuation, 12 (0.91%) patients suffered fatal PE (n = 8) or sudden death (n = 4) and 33 (2.50%) had non-fatal recurrent VTE (PE = 15; recurrent DVT = 18). In patients with premature discontinuation, the 30-day incidence of the composite outcome was 1.62 per 1000 patient-days (95%CI: 0.00-3.80). During the first week after discontinuation, the incidence rate was 4.09 per 1000 patient-days (95%CI: 0.65-7.52). The adjusted odds of the composite outcome was 7.88 times (95%CI: 6.39-9.72) higher in patients who discontinued prematurely than in those without premature discontinuation.
Premature discontinuation of anticoagulation occurred in 5% of patients with proximal DVT, and was associated an 8-fold increased odds for the composite outcome.

Copyright © 2020 Elsevier Ltd. All rights reserved.