Atrial fibrillation and anticoagulation management: a wake-up call to practitioners, patients, and policymakers

Atrial fibrillation (AF) afflicts millions of patients worldwide1–3. While heart rate and rhythm control have been the cornerstone of medical management, improvements in mechanical interventions such as radiofrequency catheter ablation offer patients hope for a long-term cure. Despite these treatment options, the lurking threat of embolic stroke and its devastating sequelae often mandates the prescribing of antithrombotic therapy. Today clinicians are better able to define embolic risk with CHADS₂ and CHA₂DS₂-VASc and bleeding risk with HAS-BLED clinical prediction tools4–6. Furthermore, pharmacogenetic testing and dosing algorithms offer promise in unraveling the mysteries of warfarin dosing by improving anticoagulant control and time in therapeutic range7. Finally, the pharmaceutical industry has introduced three novel oral anticoagulants; dabigatran, rivaroxaban, and apixaban8. These new agents offer a rapid onset of anticoagulant activity, a predictable dose response, minimal drug–drug/drug–food interactions, and do not require laboratory monitoring. In clinical trials, each has performed as well or better than warfarin in reducing stroke and systemic embolism in patients with AF9–11. With all these advances, the future of anticoagulant management looks brighter.

In this issue of JME, Stokes et al.1 share a somber glimpse of ‘real-world’ AF outcomes, but also an opportunity for future improvement. Briefly, the authors captured a group of non-valvular AF patients from the American Medco medical and pharmacy claims database. They followed this group longitudinally for thrombosis and bleeding end-points using ICD-9 coding. Bleeding events were defined as major bleeding excluding intracranial hemorrhage (MBEIH). To calculate event rates per 100 patient years, the authors estimated warfarin exposure based on prescription filling data. Despite their ‘real world’ AF population being comprised primarily of lower risk patients compared to the clinical trial (78.5% had CHADS₂ scores of ≤2 in their study vs 69.8% in ARISTOTLE), the stroke rate was high, 5.3 events per 100 patient years overall, and 3.2 and 5.7 events per 100 patient years in those with CHADS₂ scores of 1 and 2, respectively. These event rates are considerably higher than what is predicted with the CHADS₂ scoring system (2.8 and 4.0 events per 100 patient years, respectively). MBEIH was also alarmingly high at 10 events per 100 patient years.

The risk reductions from the ARISTOTLE Trial, where apixaban was employed for stroke prophylaxis, were then applied to the ‘real world’ population to determine event rates had apixaban been used instead of warfarin. The application of the apixaban risk reduction to the ‘real world’ population avoided 1.1 stroke events per 100 patient years and avoided 2.1 MBEIH events per 100 patient years. The authors estimated that apixaban use would generate an incremental medical cost avoidance for stroke of $493 in a patient year and cost avoidance of $752 for MBEIH in a patient year. The combined cost avoidance for stroke and MBEIH was $1245 during a patient year. Overall, the medical cost avoidance increased as patient stroke risk increased.

What are the implications for our readers? Stroke and bleeding complications occur frequently in the ‘real world’ setting, probably more frequently than most realize. Warfarin anticoagulation management is often sub-optimal. Stokes et al. have sent an AF alert, perhaps not a fire alarm, but a wake-up call. Performance surrounding AF management can be improved.

Practitioners should pause and examine their prescribing practices, quality metrics, and performance outcomes. They should evaluate these new medication options and assist others in this process. Warfarin has established a long track record (since 1954) of efficacy at a low cost. Its use has been facilitated by point-of-care self-testing, extended laboratory testing intervals, and centralized anticoagulation clinics. Comfort, clinical inertia, and ‘prior authorizations or preapprovals’ all represent obstacles to change. However, the novel agents are now being considered by many professional medical organizations as the ‘preferred option’ when anticoagulation is required in AF patients12–14. The novel agents are at least as effective as warfarin, more convenient for patients without laboratory testing, and a superior safety profile (∼ a 50% reduction in intracranial hemorrhage). While safety of the new agents has been challenged because there is no antidote, their short half-lives lead to reversibility. For those clinicians enrolling patients in anticoagulation clinics, computer software programs make it relatively easy to determine thrombosis and bleeding rates as well as time in therapeutic range. Since many clinics use these measures as quality indicators, clinicians should be able to obtain estimate of their event rates in routine practice. Maintaining warfarin-treated patients within a therapeutic range 55–66% of the time, as was seen in novel anticoagulant clinical trials, is a daunting task, even with specialized anticoagulation management services9–11,15. Evaluating existing performance and outcomes may be impetus for changing prescribing habits.

Patients and administrative decision-makers need to have these economic study results translated into a language they can understand. For patients, the economic scope must be broadened beyond the medication price and, preferably, before arriving at the community pharmacy prescription counter. Including discussion on the stream of future medical costs that may be avoided is warranted and should be part of medication counseling and disease education. Patients become better educated in weighing value vs purchase price. Healthcare delivery is moving toward accountable care organizations, where quality is measured in outcomes and total costs are tabulated across an assigned population. Since providers are responsible for results, decision-makers for drug purchasing and access policies must consider the acute and long-term care costs of debilitating thromboembolic stroke and intracranial hemorrhage when evaluating the ‘cost-effectiveness’ of a novel agent. They must be willing to remove barriers and make these medications accessible for routine use when there is a choice between a novel agent or warfarin.

Although the pharmaceutical industry has not cured the underlying disease, it has once again introduced an improvement in an existing class of medications; in this instance, oral anticoagulants. While the purchase price of these new agents is higher, the downstream savings from reduced complications and convenience appears far greater when compared to our existing option. Practitioners, policy-makers, and patients, when responding to the AF alert, should stop and give thoughtful consideration on where they want to spend their healthcare resources.

Transparency

Declaration of funding

The author received no payment in preparation of this manuscript.

Declaration of financial/other relationships

The author declares no conflicts of interest.