Intervention targeting clinicians did not increase anticoagulation use in patients with afib

Electronic messaging and academic detailing improved clinicians' comfort with prescribing anticoagulation for atrial fibrillation but did not increase anticoagulation use, a study found.

Researchers randomized outpatient clinicians in a 2.5:1 ratio to use an electronic profiling/messaging system combined with an academic detailing intervention or to receive no intervention. In the intervention, researchers emailed clinicians monthly reports of their anticoagulation percentage relative to peers for atrial fibrillation patients with elevated stroke risk (CHA₂DS₂-VASc ≥2). Researchers also sent electronic medical record-based messages to the clinicians shortly before they had an appointment with an anticoagulation-eligible but untreated atrial fibrillation patient.

To assess feasibility of the intervention, researchers tracked clinicians' review of messages, and to assess effectiveness, they measured the change in anticoagulation between intervention and control groups. Results were published Feb. 17 by Circulation: Cardiovascular Quality and Outcomes.

The study included 85 intervention clinicians taking care of 3,591 patients and 34 control clinicians taking care of 1,908 patients. More than 80% of intervention clinicians read the emails, and 98% of the time the clinician reviewed in-basket messages. Replies to messages identified patient refusal as the most common reason for patients not being on anticoagulation (11.2%).

At baseline, 70.8% of patients in the intervention group were on anticoagulation versus 74.0% of controls. At follow-up, the rates were 72.1% and 75.5%, respectively, for a net switch percentage of 1.3% for the intervention group versus 1.5% for the control group.

The study authors noted that clinicians who completed academic detailing were more likely than controls to go from not being comfortable prescribing anticoagulation in the vignette of a patient who had fallen three times in the previous six months to being comfortable (40% vs. 11%; P=0.02). There was also a trend toward increasing comfort with prescribing anticoagulation after gastrointestinal and subarachnoid bleeds (P=0.17).

The finding that patient refusal was the most common reason for a patient not being on anticoagulation suggested potential areas for future research, the study authors said. Patient-directed interventions or clinician-focused interventions that target patients who have declined anticoagulation may be necessary to raise the anticoagulation rates, they suggested.

More interaction between the academic detailers and the clinicians might also be helpful, they said. “Developing longer relationships with providers is another potential avenue to raise the rate of anticoagulation, particularly for providers prescribing at rates lower than their peers,” the authors wrote. “Finally, institutions with lower baseline rates of anticoagulation use may still benefit from our intervention as we implemented it, but further research is necessary to confirm a benefit.”

An accompanying editorial noted that patients' individual comorbid, personal, and socioeconomic factors may play an important role in anticoagulation rates. “The prescription of [oral anticoagulants] may be a function of the healthcare system,” the editorialists wrote. “Yet, as [this study] implies, it may also be a function of the clinical encounter and of the patient's situation.” They called for further appreciation of the challenges of care implementation, as well as “adequate funding, innovative methods, and bolder and more rigorous studies.”