One of the many potential problems with warfarin-based anticoagulant therapy is the poor rate of adherence and persistence among patients who are prescribed the drug. Now a new observational study published in Archives of Internal Medicine raises the possibility that the problem may be even worse than many may have previously suspected, as discontinuation rates in clinical trials appear to be much lower than in the real world.
Tara Gomes and colleagues analyzed data from more than 125,000 new users of warfarin in Ontario, Canada and found very high rates of discontinuation over time:
- 8.9% never filled a second prescription
- 31.8% discontinued warfarin within 1 year
- 43.2% discontinued warfarin with 2 years
- 61.3% discontinued warfarin within 5 years
People at higher risk for stroke, as assessed by the CHADS2 score, were more likely to continue taking warfarin over the course of the study.
In an invited commentary, Whitney Maxwell and Charles Bennett write that the results are consistent with previous observational studies but that warfarin discontinuation can be appropriate and is often initiated by the physician. “Appropriateness of anticoagulation discontinuation is perhaps a more important outcome to evaluate rather than absolute discontinuation rates,” they write. An additional plausible explanation for the study finding is that few patients in Canada were treated at anticoagulation clinics.
Maxwell and Bennett write that any potential problem with anticoagulation discontinuation is not limited to warfarin. In the RE-LY trial, they note, more patients discontinued therapy in the dabigatran arm than in the warfarin arm, and a similar trend was observed with rivaroxaban in ROCKET AF.