Telemonitoring based feedback improves adherence for non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation

Abstract

Introduction: Effective therapy with non-vitamin K antagonist oral anticoagulants (NOACs) requires strict therapy adherence given half-lives of these drugs of about 12 hours. Adherence to other cardiovascular chronic drug therapy is often suboptimal. Systematic monitoring of anticoagulation or medication intake is not performed in NOAC patients. Data on interventions to monitor and/or improve adherence to NOAC therapy are almost absent. Purpose: To investigate the effect of in-person feedback, based on telemonitoring of medication intake, on adherence to NOACs in patients with atrial fibrillation (AF). Methods: Forty eight AF patients (24 male; mean age 72 ±9 years; 24 on a once daily (OD) NOAC (rivaroxaban) and 24 patients on a twice daily (BID) NOAC (apixaban)) were enrolled in a randomised, single-blind, crossover, controlled trial. Adherence to NOACs was measured using the electronic Medication Event Monitoring System (MEMS, WestRock, Switzerland). Patients were randomised to 3 months each of a purely observation phase and a feedback phase, in random order. Adherence data was checked daily on weekdays through telemonitoring. During the feedback phase, patients received a phone call in case of an ‘unprotected day’. Taking adherence (i.e. proportion of prescribed doses taken), regimen adherence (i.e. proportion of days with the correct number of doses taken) and number of unprotected days were calculated based on the MEMS data. An ‘unprotected day’ was defined as three or more consecutive missed doses for a BID NOAC and one or more missed doses for a OD NOAC. Patients were also contacted when they took excess doses during the prior 24 hours. Pill counts were performed after each period of three months. Results: No patient stopped OAC treatment, although one was switched to VKA after three months due to a venous thrombus (i.e. persistence = 98%). Already under active telemonitoring observation, adherence was very high, with a taking adherence of 97.4% and a regimen adherence of 93.8%. Nevertheless, adherence was further improved through direct feedback: taking adherence increased with an absolute 1.6% to 99% (p<0.001) and regimen adherence with 3% to 96.8% (p=0.001). Pill count increased from 97.9% to 99.0% (p=0.004). The number of unprotected days in a 3 month period decreased from 2.6 to 1.5 (p=0.125). Both during the observation and the feedback phase, taking adherence was higher with the OD NOAC (p<0.001 and p=0.018, respectively) although unprotected days were similar (p=0.272 and p=0.251, respectively). Conclusion: Electronic monitoring revealed an unexpectedly high adherence to NOAC therapy in an elderly unselected population. This may be related to highly motivated patients but certainly also to the sense of being watched (‘radar effect’). Nevertheless, telemonitoring follow-up and direct feedback could further optimise adherence. This may be a valuable approach in selected patients deemed poorly adherent in clinical practice.