Background: More than 5 percent of hospitalizations in the United States are a result of poor adherence to medication therapy. This problem is particularly significant with medications that have narrow therapeutic windows, such as warfarin (Coumadin). This medication is highly effective when anticoagulation levels are maintained within the International Normalized Ratio (INR) of 2.0 to 3.0. However, there is an increased risk of bleeding with supratherapeutic levels and an increased risk of thromboembolic events with subtherapeutic levels. Because of these issues, poor control of anticoagulation therapy may be one of the leading causes of preventable adverse drug events in older ambulatory patients.

Despite this, the influence of patient adherence on anticoagulation control is not well studied. Because warfarin has a long half-life (nearly 40 hours), it is possible that typical problems with patient nonadherence have only a minimal impact on anticoagulation control. Kimmel and associates evaluated patient nonadherence to warfarin therapy and its impact on anticoagulation control.

The Study: The authors performed a prospective cohort, multicenter trial. Adult patients at three anticoagulation clinics with a target INR of 2.0 to 3.0 were included in the study. Patients with an abnormal INR at presentation and those with antiphospholipid antibodies were excluded. A trained interviewer collected information about factors that could influence the INR during warfarin therapy. To assess patient adherence, each participant was provided with the Medication Event Monitoring System, which records the exact time and date that a patient opens the pill bottle. Patients also completed questionnaires during each visit that documented any changes in medications, vitamin K, or alcohol intake; weight loss or weight gain; and whether they had instructions to temporarily stop warfarin therapy.

Results: There were 136 participants in the study, and 1,490 INR measurements were completed. Of these INR measurements, 40.4 percent were out of therapeutic range: 25.8 percent were below 2.0, and 14.6 percent were above 3.0. Furthermore, 92 percent of the participants incorrectly dosed their warfarin, which was defined as missing one or more bottle openings or opening the bottle more times than needed. Thirty-six percent of the participants missed more than 20 percent of their bottle openings, and 4 percent opened their bottles more times than needed 10 percent of the time. For each 10 percent increase in the missed bottle openings, the odds of underanticoagulation increased 14 percent. Those who missed one or two pills per week (i.e., missing 20 percent of bottle openings) had more than a twofold increase in the odds of underanticoagulation. Patients who had 10 percent or more extra bottle openings had a statistically significant increase in their risk of overanticoagulation.

Conclusion: Poor adherence to warfarin therapy can have a significant impact on anticoagulation control. Improving patient adherence to warfarin regimens is a major challenge facing health care professionals, and future efforts should be directed toward improving patient adherence with warfarin therapy or finding a safer alternative.