Atrial Fibrillation Follow-up Investigation of Rhythm Management - AFFIRM


Comparison of initial strategies of rate control and rhythm control for the treatment of atrial fibrillation in older adults.

Study Design

  • Randomized

Patients Screened: 7401
Patients Enrolled: 4060
Mean Follow Up: target 5 years; mean 3.5 years
Mean Patient Age: Mean age: 69.7 years
Female: 39
Mean Ejection Fraction: mean overall 54.7

Patient Populations:

Age ≥65 years or other risk factors for stroke or death; atrial fibrillation likely to be recurrent; atrial fibrillation likely to cause illness or death; long-term treatment for atrial fibrillation was warranted; anticoagulant therapy was not contraindicated; the patient was eligible to undergo trials of at least two drugs in both treatment strategies; and treatment with either strategy could be initiated immediately after randomization.

Primary Endpoints:

All-cause mortality

Secondary Endpoints:

Composite of death, disabling stroke, disabling anoxic encephalopathy, major bleeding, and cardiac arrest.

Drug/Procedures Used:

The randomized, controlled trial included 4,060 adults (n=2,027 in rate control arm and n=2,033 in rhythm control arm) at 213 sites with a mean age of 69.7 years who had AF documented by an electrocardiogram. To be eligible, patients had to have 6 or more hours of AF during the previous 6 months with a qualifying episode within the 12 weeks prior to enrollment. The three principal drugs used in the patients randomized to rate control were digoxin (51% of patients), beta blockers (49%), or calcium channel antagonists (41%). In the rhythm control arm, patients were treated using predominantly amiodarone (39%), sotalol (33%), and propafenone (10%). Ablation and pacemakers were also used in the rhythm control arm, if necessary. The rate of warfarin use was 85 to 95% in the rate control arm and approximately 70% in the rhythm control arm. Patients in the rhythm arm were permitted to discontinue anticoagulation therapy after normal rhythm had been achieved and maintained for 1 month.

Principal Findings:

The prevalence of sinus rhythm declined over time, and at 5 years was approximately 60% in the rhythm control arm and nearly 40% in the rate control arm. Successful rate control was achieved in more than 80% of patients in the rate control arm by year 5. The primary endpoint of AFFIRM was all-cause mortality. Mortality was 25.9% in the rate control arm and 26.7% in the rhythm control arm (HR=1.15, p=0.08). The mortality curves appeared to diverge after approximately 1.5 to 2.0 years. Ischemic strokes occurred in 5.5% of the rate control arm and 7.1% of the rhythm control arm, a nonsignificant difference (p=0.79). Most strokes occurred in patients in whom warfarin therapy was discontinued. There were no statistically significant differences in the secondary composite endpoint (32.7% vs 32.0%, p=0.33), functional status, or quality of life. The hospitalization rate was higher in the rhythm control arm (80% vs 73%, p<0.001).


The data show that there is no survival benefit to the strategy of rhythm control in elderly patients with atrial fibrillation. Indeed there were trends toward higher rates of death and stroke among patients treated with rhythm control. Rate control appears to be an acceptable primary strategy in patients with atrial fibrillation. Long term anticoagulation appears warranted in all patients with risk factors for stroke. Risk factors for ischemic stroke in nonvalvular AF include prior thromboembolism, CHF, HTN, age, DM, female gender, BP greater than 160 mmHg and LV dysfunction. As stated in the ACC/AHA guidelines, maximum protection against ischemic stroke in AF is probably achieved with an international normalized ratio (INR) of 2 to 3, whereas an INR range of 1.6 to 2.5 appears to be associated with incomplete efficacy, estimated at approximately 80% of that achieved with higher intensity anticoagulation. ACC/AHA guidelines recommend that physicians individualize the selection of the antithrombotic agent based on assessment of the absolute risks of stroke and bleeding and the relative risk and benefit for a particular patient. It should be noted that while rate control reduces proarrhythmia risks, it may be associated with persistent symptoms of palpitations, dyspnea, chest pain, fatigue, and fainting. Patients may not find these symptoms tolerable. It is notable that patients who could not tolerate these symptoms were not enrolled in the trial, and this study therefore does not apply to those patients who cannot tolerate rate control. It is possible, but as yet untested, that ablation may be a solution for those patients who do not tolerate rate control. Ablation, if successful, has the additional advantage of reducing the need for anticoagluation. The long term recurrence rate with pulmonary vein ablation is not yet known.


N Engl J Med 2002;347:1825-33.

Keywords: Stroke, Digoxin, Heart Conduction System, Warfarin, Syncope, Risk Factors, Electrocardiography, Fibrinolytic Agents, Dyspnea, Prevalence, Thromboembolism, Quality of Life, Chest Pain, Propafenone, Sotalol, Hospitalization

< Back to Listings