PATIENT KNOWLEDGE AND PREFERENCES SURROUNDING INTERRUPTION OF ANTICOAGULATION THERAPY FOR MINOR PROCEDURES

Purpose: There are over 250,000 incidents of and more than 70% of patients experiencing anticoagulation interruption within 10 years of therapy initiation. This study evaluated the impact of nurse delivered education on patient knowledge of risks and benefits of anticoagulation, assessed patient preferences for the trade-off of risks for stroke versus bleeding and examined alignment between patient and physician preferences.

Methods: We conducted a prospective study of patients in an anti-coagulation clinic and surveyed both physicians and nurses. We collected basic demographics then asked patients to complete the validated oral anti-coagulation knowledge (OAK) test, as well as two questions about what they should do relative to stopping their warfarin prior to a dental or surgical procedure.

After completing the questionnaires, patients were randomized to receive either standard care or additional nurse led education concerning best practices for anti-coagulation maintenance. When patients returned for their next visit, they completed the OAK test and additional 2 questions again as well as an exercise assessing patient preference for risk of bleeding versus stroke and impact on their preferences for interrupting their anti-coagulation therapy. Providers were asked to assess their preferences for stroke versus bleeding risk via a link emailed to them. 

Results: The study sample included 73 patients, 31 physicians and 194 nurses. 69% of the education versus 35% of standard care patients improved their OAK score from the pre-to the post-test (p<.001). Both patients and providers were willing to accept bleeding rates much higher than those that occurring clinically to avoid a 1% risk of stroke, with patients accepting a moderate to severe bleeding risk up to 49% before electing to interrupt anti-coagulation therapy and over 90% of physicians and nearly 90% of nurses electing against discontinuing anti-coagulation faced with a 1% risk of a moderate to severe bleed, consistent with what would be expected clinically.

Conclusion: Patients are moderately knowledgeable about anti-coagulation therapy, but there is room for improvement. Regardless of knowledge, patients can express their preference for competing adverse events. Neither patient nor provider preferences are well reflected by current practice or by guidelines for anti-coagulation interruption pointing to the need to modify guidelines to reflect not only evidence, but patient preferences for incorporating that evidence in clinical practice.