Whether you are a resident on the cardiology, surgery, or general medicine service, encountering patients with atrial fibrillation is common. Many patients, particularly after cardiac surgery, go in and out of atrial fibrillation so often that residents caring for them often ignore the blinking lights and loud alarms from telemetry machines after a while. However, atrial fibrillation has adverse consequences– postoperative atrial fibrillation and its sequelae cost the US healthcare system up to $1 billion each year. While anticoagulation can reduce the risk of the dreaded complication of stroke, it is also important to treat atrial fibrillation to prevent structural consequences of rapid ventricular response, such as tachycardia-mediated cardiomyopathy. Two strategies – controlling the heart rate (rate control) and converting the patient’s heart to sinus rhythm (rhythm control) – are used as medical treatments for atrial fibrillation.
Taking an historical perspective, in the 1980s and 1990s, when large clinical trials demonstrated the large stroke risk of atrial fibrillation, a slew of antiarrythmic drugs came onto the market. The logic was simple– if atrial fibrillation led to devastating consequences, it made sense to try to restore sinus rhythm. However, antiarrythmic drugs came with potentially dangerous side effects. Some practitioners thought that decreasing the risk of rapid ventricular response through rate control might confer a similar benefit without the potentially harmful side effects of rhythm control. The landmark Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial, published in NEJM in 2002, found that in 4000 nonsurgical patients with prior atrial fibrillation, rhythm control offered no difference in survival compared to rate control, but resulted in more hospitalizations and more adverse events. The smaller Rate Control versus Electrical Cardioversion for Persistent Atrial Fibrillation (RACE) trial, also published in NEJM in 2002, reported similar findings in 522 patients with persistent atrial fibrillation.
However, as in most of medicine, it is difficult to generalize. This is particularly true after cardiac surgery, when up to 50% of patients develop atrial fibrillation. While a joint 2014 American College of Cardiology (ACC), American Heart Association (AHA), and Heart Rhythm Society (HRS) guideline recommended rate control with beta-blockers for postoperative atrial fibrillation, considerable variation in practice continues.
In this week’s NEJM, Gillinov et al. report the results of a multi-center (23 centers) randomized controlled trial designed to answer this question. Investigators enrolled patients who were undergoing elective cardiac surgery for either coronary artery disease or valvular disease and developed persistent or recurrent postoperative atrial fibrillation. Patients with a history of prior atrial fibrillation were excluded. From 2014-2015, 523 patients with postoperative atrial fibrillation were randomized to receive either a rate control or rhythm control strategy. The rate control arm received medications to slow the heart rate to a target of less than 100 beats per minute. The rhythm control arm received a standard loading dose of amiodarone for pharmacologic rhythm control. If atrial fibrillation persisted for 24-48 hours after randomization, patients received cardioversion. If any patient in either arm had atrial fibrillation for longer than 24-48 hours after randomization, they received anticoagulation.
The investigators found that atrial fibrillation was present after cardiac surgery in 30% of patients – and in 50% of patients who had received combined coronary artery bypass graft and valve surgery. Adults (mean age 68.8 years) receiving rate control had similar numbers of hospital days (6.4 vs 7.0, respectively; P=0.76); length of stay for the index hospitalizations (5.5 vs 5.8 respectively; P=0.88); and rates of rehospitalization (P=0.88). At 60-days, 93.2% of individuals in the rate control group remained in sinus rhythm, compared to 97.9% of patients in the rhythm control group (P=0.02). There were no significant differences in the overall rates of serious adverse events, strokes, or serious bleeding in the rate- or rhythm-control groups. Similar numbers of patients in both arms received anticoagulation prior to discharge; while rate control resulted in slower resolution to sinus rhythm, the duration of anticoagulation was similar between both arms (median 45 days).
There are several limitations to the study, including the large rates of crossover between treatment strategies, as well as treatment discontinuation in the two arms. However, the study’s greatest strength lies in numbers—it is a large (N=523) randomized trial of post-operative atrial fibrillation management. The findings of Gillinov et al. suggest that whereas rhythm control may offer faster time to sinus rhythm, it comes with the potential side effects of antiarrythmic drugs, in this case, amiodarone, with no shorter duration of anticoagulation. Since rhythm control strategies are complex and offered no definitive clinical benefit, many would interpret the study’s results as in clear favor of rate control. Hugh Calkins, M.D., Director of the Cardiac Arrhythmia Service at the Johns Hopkins Hospital, writes in an accompanying editorial: “The all too frequent knee-jerk reaction to cardiovert patients with postoperative atrial fibrillation with hopes of reducing stroke risk, the need for anticoagulation, and the duration of hospitalization seems hard to justify given the well-known risks and costs associated with such interventions and the absence of clear beneficial effects of the rhythm control strategy.” The Gillinov et al. study clarifies the risks and benefits of either treatment strategy in post-operative atrial fibrillation and argues for shared decision-making with patients prior to heart surgery.