Author + information
- Received November 1, 2002
- Revision received January 3, 2003
- Accepted January 24, 2003
- Published online July 2, 2003.
- ↵*Reprint requests and correspondence:
Dr. H. Leon Greene, AFFIRM Clinical Trial Center, Axio Research Corporation, 2601 Fourth Avenue, Suite 200, Seattle, Washington 98121, USA
Objectives This study evaluated the efficacy of antiarrhythmic drugs for the treatment of atrial fibrillation (AF).
Background The most effective and safest antiarrhythmic drug for the treatment of AF is unknown.
Methods The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study compared two treatment strategies—rate control and rhythm control—in patients with AF and risk factors for stroke or death. This substudy, performed in patients randomized to rhythm control, compared different antiarrhythmic drugs by randomly assigning the first drug treatment to: 1) amiodarone, 2) sotalol, or 3) a class I drug. The primary end point was the proportion of patients alive, in sinus rhythm, with no additional cardioversions and still taking the assigned drug at one year. Comparisons were made between patients eligible for each of three drug pairs.
Results At one year, in 222 patients randomized between amiodarone and class I agents, 62% were successfully treated with amiodarone, compared with 23% taking class I agents (p < 0.001). In 256 patients randomized between amiodarone and sotalol, 60% versus 38% were successfully treated, respectively (p = 0.002). In 183 patients randomized between sotalol and class I agents, 34% versus 23% were successfully treated, respectively (p = 0.488), although this portion of the substudy was stopped early when amiodarone was shown to be better than class I agents. Sinus rhythm was achieved in nearly 80% of patients at one-year follow-up with serial therapy. Adverse effects were common.
Conclusions Amiodarone was more effective at one year than either sotalol or class I agents for the strategy of maintenance of sinus rhythm without cardioversion.
The most effective and safest antiarrhythmic drug for the maintenance of sinus rhythm in patients with atrial fibrillation (AF) remains uncertain. Rhythm control of AF using antiarrhythmic drug therapy in some patients might offer clinically meaningful advantages, such as increased exercise tolerance, decreased symptoms, and prevention of atrial remodeling (1,2). Studies of a host of antiarrhythmic agents administered to maintain sinus rhythm in patients with AF have been previously reported (3–5). Efficacy rates in main taining sinus rhythm have been variably defined in diverse AF populations and approximate 50% at one year. Recent meta-analyses have emphasized that few studies have long-term follow-up, that most do not report meaningful mortality statistics, and that a clear superiority of one drug over another has not been demonstrated (4,5). All antiarrhythmic agents have the potential for serious or even lethal adverse events, which are seen more commonly in patients with left ventricular dysfunction, congestive heart failure, or ischemic heart disease (6–10).
The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study was a randomized clinical trial in patients with AF and a high risk of stroke or death, comparing two treatment strategies: 1) attempting to maintain sinus rhythm, principally with antiarrhythmic drugs (“rhythm control”); and 2) allowing AF to persist or recur while controlling the ventricular rate, principally with atrioventricular node blocking drugs (“rate control”) (11–13). Both arms of the study utilized anticoagulation. In the population studied, it was shown that rhythm control offered no advantage over rate control. In the group of patients assigned to the rhythm control strategy in AFFIRM, a substudy conducted at 80 of the 213 clinical sites compared the ability of different antiarrhythmic drugs to maintain sinus rhythm at one year by randomly assigning therapy to either amiodarone, sotalol, or a class I drug.
The AFFIRM study evaluated 4,060 patients recruited at 213 clinical sites in the U.S. and Canada from November 1995 through October 1999 (11–13). The study was approved by each site’s Institutional Review Board, and all patients gave written, informed consent. The primary end point for the main study was death from any cause. Follow-up in the main study and this substudy ended on October 31, 2001 (13).
The first antiarrhythmic drug substudy
This substudy was a second randomization of patients assigned to the rhythm control arm of AFFIRM, performed at 80 study sites (11,12). Patients had to be eligible to receive treatment randomized from at least two of the three antiarrhythmic drug choices: amiodarone, sotalol, or a class I agent (Table 1). Patients previously failing one class I drug were not necessarily excluded from receiving a different class I drug, unless the patient had experienced torsade de pointes ventricular tachycardia as the reason for the intolerance of the first class I drug. All eligibility decisions were made by the enrolling physician, in discussion with the patient. Although a patient might not have met any strict criteria for contraindications for a drug, that drug might have been eliminated from consideration for many reasons. For example, if quinidine had been associated with unacceptable diarrhea and had not been effective at that dose, it would not have been used again. Likewise, some patients may have had such severe lung disease that the physician may have been unwilling to use amiodarone. As a registry was not kept either for the main trial or for this substudy, we do not know how many patients were considered for enrollment in the First Antiarrhythmic Drug Substudy, the reasons that some patients were eliminated from consideration, or the precise reasons that some drugs were considered to be inappropriate for the substudy.
At the time of the randomization telephone call to the Clinical Trial Center, the investigator reported whether the patient was eligible and agreed to randomization in the First Antiarrhythmic Drug Substudy. Patients randomized to the rhythm control arm of the main AFFIRM study then received a second randomization of the antiarrhythmic drug to be used in this substudy.
The substudy was designed to be three separate comparisons between two drugs: amiodarone versus class I agents, amiodarone versus sotalol, and sotalol versus class I agents. If a patient was eligible for all three drug arms, the randomization was performed among all three drug groups, and the patient’s data were added to the pool of data in two two-way comparisons. For example, a patient eligible for amiodarone, sotalol, and a class I agent might be randomized to amiodarone. His or her data would be included in the amiodarone arm of both the amiodarone versus class I comparison and in the amiodarone versus sotalol comparison. Thus, a patient may be represented in more than one two-way comparison, and the sum of patients in all six arms of the three two-way comparisons would exceed the total number of patients participating in the substudy. A patient eligible for only two drug arms would be randomized between those two arms and his or her data included only in that single two-way comparison.
Although amiodarone could be administered as a low dose (200 mg/day) or high dose (400 mg/day), it counted as only a single drug for this substudy analysis. If patients were assigned to a class I drug, the choice among available class I drugs was made by the treating physician. If the patient or physician declined participation in this substudy, the choice of antiarrhythmic drug for the main study was then made by the physician.
Recurrence of AF while on the first antiarrhythmic drug, followed by spontaneous reversion to sinus rhythm, did not constitute a drug failure. However, it is likely that the drug would have been discontinued and thus considered to be a failure if the patient had severe symptoms with the recurrence. Recurrence of AF while on the first antiarrhythmic drug withoutspontaneous reversion to sinus rhythm was classified as a drug failure for the purposes of this substudy. Discontinuation of the first class I drug, with substitution of another class I drug, was also considered to be a failure. However, the main AFFIRM protocol encouraged continued use of the drug in the presence of only one such “failure.” Any subsequent drug choice was left to the discretion of the investigator.
Antiarrhythmic drug use
The AFFIRM protocol instructed the investigators to exercise particular caution in the administration of all drugs to patients with ischemic heart disease and/or left ventricular dysfunction (Table 1) (11). Hospitalization and electrocardiographic (ECG) monitoring for initiation of treatment was strongly encouraged for all patients with structural heart disease. The class I-C agents, propafenone and flecainide, were not to be given to patients with a history of congestive heart failure or structural heart disease (ventricular myocardial disease, left ventricular hypertrophy, coronary artery disease, myocardial ischemia, or myocardial infarction). Patients who had previously experienced serious ventricular arrhythmias (ventricular tachycardia or ventricular fibrillation) were not eligible to receive class I-A agents or sotalol (11,12). Dofetilide was not approved for use by the U.S. Food and Drug Administration until near the end of the study; therefore, it was not included in the randomization scheme for this substudy.
Patients were followed at 2, 4, 8, and 12 months, and at 4-month intervals thereafter, the standard follow-up intervals for the main AFFIRM study. Although the primary end point of the First Antiarrhythmic Drug Substudy was measured at one year, follow-up for secondary end points and for the main study occurred at 4-month intervals until the end of the study—October 31, 2001. Electrocardiograms were obtained at 4 and 12 months, per protocol, or as clinically necessary, and the known or estimated date of the first recurrence of AF was recorded.
The primary end point of this substudy was the percentage of patients at one year who were still alive, were taking the assigned drug, had not had external or internal electrical cardioversion or pharmacologic cardioversion (either a new drug or augmented doses of the drug administered long term), and were in sinus rhythm.
A secondary end point was the time to the first documented recurrence of AF, as determined by the patient presenting with symptoms and with confirmation of AF by electrocardiography or on routine electrocardiography at four months and one year. Another secondary end point was the prevalence of AF at four months and one year, as determined by a routine ECG taken at these time intervals. In addition, cardioversions were noted. Important descriptive end points were mortality, drug tolerance, and safety.
Randomization was stratified only by clinical site. All analyses were based on the intention-to-treat principle. Having to discontinue an assigned drug for any reason was considered to be a drug failure, and that patient would be counted as a failure in the arm to which he or she had been randomized. Patients were censored at last contact, at last vital status determination, or at withdrawal from the study. Covariates were compared in each group by chi-square tests, ttests, and nonparametric tests, without adjustments for multiple comparisons. Event rates and time-to-event curves were estimated by the Kaplan-Meier method (14)and compared using the log-rank test.
Data were reviewed every six months by the AFFIRM Data and Safety Monitoring Board, using the sequential monitoring technique of O’Brien and Fleming (15)and Lan and DeMets–type spending function (16). An alpha level of 0.0167 was used as a Bonferroni correction for the three comparisons (17). Success in maintaining sinus rhythm, as well as drug safety and adverse effects, was considered by the Board in deciding whether or not to terminate the substudy. Although some portions of the substudy were terminated early, this analysis includes all accumulated data on randomized patients.
This substudy began simultaneously with enrollment for the main AFFIRM study in November 1995 and enrolled a total of 410 patients, 20% of the subjects enrolled in the rhythm control arm of the main study (Tables 2 and 3). ⇓The randomization between amiodarone and class I drugs was halted at the third interim analysis on December 17, 1998, because ⇓the findings in this portion of the substudy were considered conclusive. Furthermore, because the Data and Safety Monitoring Board concluded that too few patients would then be eligible for the comparison between sotalol and class I agents, and accumulated data suggested that continuation of this comparison could not yield a significant result, randomization in this arm of the substudy was also terminated. The amiodarone versus sotalol randomization continued until the completion of AFFIRM enrollment on October 31, 1999. The Data and Safety Monitoring Board concluded at the seventh interim analysis on December 6, 2000, that data from this portion of the substudy were also conclusive.
At one year, only three patients in the substudy had withdrawn consent to participate (one assigned to amiodarone and two assigned to class I agents), and three patients were lost to follow-up (all in the sotalol assignment). These patients were censored at the time of withdrawal or loss to follow-up.
The time to achievement of sinus rhythm after randomization to the rhythm control arm of the study was not different in all three arms of this substudy. At randomization, 46% of patients randomized to amiodarone were in sinus rhythm, compared with 37% of patients randomized to class I agents and 37% of patients randomized to sotalol. Approximately 78% of all patients had sinus rhythm restored by 30 days after randomization, and 92% by 120 days after randomization.
Mean follow-up was 3.84 ± 1.30 years. Patients enrolled in the First Antiarrhythmic Drug Substudy were similar to the 4,060 patients enrolled in the AFFIRM study in general, and they were also similar to the 2,033 patients randomized to the rhythm control arm of the main AFFIRM study (12,13). Few significant or important differences existed between the comparison groups (Table 4).
Major end points
The ⇓primary and major end points are summarized in Tables 5 through 7. Only p values for the primary end point were adjusted for multiple comparisons. For deaths, the p value is based on the Kaplan-Meier estimates of the death rates at five years, as well as the log-rank statistic.
At one year, successful treatment was noted in 62% of amiodarone-treated patients, compared with 23% of patients given class I agents (p < 0.001). Also at one year, 60% of amiodarone-assigned patients were successfully treated, compared with 38% of patients assigned to sotalol (p = 0.002). No differences were noted in the success of sotalol compared with class I agents, with one-year point estimates of the major end point of 34% and 23%, respectively (p = 0.488). However, as noted previously, this portion of the substudy was stopped early, and the reduced sample size did not provide adequate power to detect a clinically meaningful difference.
The prevalence of AF at four months and one year was assessed by the rhythm recorded on the ECG at those intervals and did not take into account any other factors in the follow-up of these patients. The prevalence of AF was lower for amiodarone-treated patients than for class I-treated and sotalol-treated patients at these times.
In addition, time-to-event analyses based solely on the first recurrence of AF are illustrated in Figures 1 through 3. ⇓In these figures, patients are censored at death, loss to follow-up, or patient withdrawal from further participation in the entire study. ⇓Because this analysis considers results over time, and the primary analysis considers the data at one year only, the percentage without recurrence in these figures would be expected to be different from the estimates ⇓of the primary end point and prevalences of sinus rhythm listed in Tables 5 through 7. Amiodarone was more successful than class I agents in preventing a recurrence of AF (p = 0.0114) (Fig. 1). Amiodarone was also more successful than sotalol in maintaining sinus rhythm (p = 0.0003) (Fig. 2). No significant differences were noted in the ability of sotalol versus class I agents to maintain sinus rhythm (p = 0.8671) (Fig. 3).
Multivariate analyses were performed on the two two-way comparisons with adequate power for such an assessment, evaluating for the presence of the primary end point (death, recurrent AF, having cardioversion, or stopping the first assigned drug). For amiodarone versus class I drugs, the drug therapy was significant (p = 0.001), and a history of thyroid disease was associated with the primary end point (p = 0.021). For amiodarone versus sotalol, the drug therapy was significant (p = 0.003), and enrolling with a first episode of AF (p = 0.019) or having a history of thyroid disease (p = 0.048) was associated with the primary end point.
Figure 4shows that nearly 80% of patients were in sinus rhythm at one-year follow-up, irrespective of the initial randomized therapy. This result was obtained by electrical or pharmacologic cardioversions, by changing drugs, or both, for those patients whose AF recurred. Patients in the amiodarone arms were less likely to require cardioversion or drug changes.
This substudy was not designed to have sufficient statistical power for the detection of any differences in death rates among the different drug groups. However, we observed that more deaths occurred in patients randomized to class I agents than to amiodarone (Table 5). Most of these deaths occurred during two to five years of follow-up, after the randomized therapy had been changed. Assignment of the cause of death can be difficult (18). Nevertheless, in this amiodarone versus class I comparison, 26 deaths occurred in the class I arm (7 were classified as arrhythmic deaths); 24 of these deaths occurred after the assigned drug had been discontinued (usually for inefficacy). Of the 10 deaths in the amiodarone arm of this comparison, four were arrhythmic and five occurred after amiodarone was discontinued. None were due to amiodarone-induced pulmonary toxicity. No significant differences were noted in the death rates between amiodarone and sotalol (15 vs. 24) (Table 6) or between sotalol and class I agents (13 vs. 17) (Table 7), and again, most of these deaths also occurred after the initially assigned drug was changed or all antiarrhythmic drugs had been discontinued. In total, 69 deaths were noted in this substudy (11 in the first year). At the time of death, the number of patients taking no antiarrhythmic drugs was 27, amiodarone 24, quinidine 1, disopyramide 1, flecainide 2, moricizine 1, procainamide 3, propafenone 2, and sotalol 8.
Other end points and adverse effects
Electrical cardioversions in follow-up were recorded as a part of the AFFIRM protocol. By one year, 29% of patients randomized to amiodarone, 33% of patients randomized to sotalol, and 36% of patients randomized to class I agents had been cardioverted (p = 0.52).
Only one case of torsade de pointes ventricular tachycardia was recognized in this substudy population (in a patient taking quinidine for more than one year). Furthermore, no cases of agranulocytosis or lupus syndrome induced by procainamide were reported in these substudy patients. Adverse effects causing discontinuation of the antiarrhythmic drugs during the first year were frequent (Table 8), occurring in 12.3% of patients taking amiodarone, 11.1% taking sotalol, and 28.1% taking class I agents. Among those patients randomized to amiodarone, pulmonary toxicity was diagnosed in two patients (1.3%) by one year, three patients (2.0%) by two years, and no additional patients by three years. Gastrointestinal adverse events were a common reason for stopping class I drugs.
This study is a prospective, randomized comparison of the efficacy and safety of antiarrhythmic agents available for the suppression of AF. On the basis of the primary substudy end point, amiodarone was better than either sotalol or class I agents in separate comparisons. Serious adverse drug effects were uncommon within the one-year time frame of the primary end point in this AFFIRM substudy, although adverse effects that led to discontinuation of a drug were frequent.
The baseline clinical characteristics of this population, including the duration of the qualifying episode of AF, the extent of left ventricular dysfunction, and the size of the left atrium, gave them not only a high risk for stroke and death but also a high risk for recurrence of AF. All patients in the antiarrhythmic drug group comparisons had similar clinical characteristics, except for minor imbalances in age, history of congestive heart failure, proportion of patients presenting with a first episode of AF, and beta-blocker and calcium channel blocker use. Given the large number of comparisons tested, it would be expected that some would be significant simply by chance.
Measures of antiarrhythmic drug efficacy
Antiarrhythmic drugs are frequently compared for their ability to maintain sinus rhythm. In general, most antiarrhythmic drugs are able to maintain sinus rhythm in only ∼50% of patients at one year. For amiodarone, this percentage appears to be greater (19). However, this AFFIRM substudy had a complex composite end point that incorporated survival, adverse effects that required discontinuation of the drug, and the ability to control cardiac rhythm without cardioversion. Although the results of this substudy are clear, it is not necessarily true that a single recurrence of AF should be called a categorical failure of an antiarrhythmic drug. The choice of an antiarrhythmic drug to suppress AF is sufficiently complex such that no single measure of efficacy or adverse effects can be used alone to determine optimal therapy. For instance, although amiodarone was the drug most likely to maintain sinus rhythm without cardioversion, its potential for serious adverse effects over a long term might preclude it being used as the first drug. Moreover, it might be preferable to accept an occasional recurrence of AF using a drug with easily reversible potential adverse effects than to use a drug with less likelihood of AF recurrence but with potential adverse effects that occur late and recede slowly. As demonstrated by this substudy, starting with an antiarrhythmic drug other than amiodarone and accepting serial therapy (cardioversion to restore sinus rhythm if needed and possibly switching to another antiarrhythmic drug) will result in sinus rhythm being present at one year in nearly 80% of patients. In short, when choosing a drug, there should be a balance between potential adverse effects and potential beneficial effects of maintaining sinus rhythm, and the frequency of AF recurrence must be considered.
Of note, deaths most often occurred after the one-year primary end point of this substudy was assessed, usually after drug therapy had been stopped or changed. Although the higher mortality rate in patients starting with class I drugs is intriguing, it is unexplained. Most of these deaths occurred after the class I drug was discontinued (for inefficacy or adverse effects) and while another drug, often amiodarone, was being used.
Comparison with other studies
Our substudy population was larger than most previous studies. Its patients were better characterized and had a higher risk for recurrence and adverse outcomes, such as stroke and death. Also, although other studies have usually compared one drug with placebo or one drug with another, this substudy was unique in evaluating multiple, randomly selected drugs to maintain sinus rhythm.
This AFFIRM substudy generally confirms the results of previous studies of the efficacy of antiarrhythmic drugs to maintain sinus rhythm. The Canadian Trial of Atrial Fibrillation (CTAF) investigators compared amiodarone with sotalol and propafenone for the maintenance of sinus rhythm in patients with a history of AF (19). The primary end point was the first recurrence of AF, as measured by transtelephonic monitoring. The mean follow-up period was 16 months. They showed that 69% of patients maintained sinus rhythm at one year with amiodarone, and amiodarone was superior to sotalol or propafenone for the maintenance of sinus rhythm in these patients.
The Pharmacological Intervention in Atrial Fibrillation (PIAF) study compared rate control with rhythm control in patients with a history of AF (20). It demonstrated that sinus rhythm could be maintained with amiodarone in 56% of patients at one year. The PIAF study had a one-year follow-up period, with the primary end point being improvement of symptoms. Symptomatic improvement was similar in the rate control and rhythm control groups.
In PIAF, 25% of patients had stopped amiodarone at one year because of adverse effects (20). Pulmonary toxicity was not reported. In CTAF (with a follow-up of 16 months, slightly longer than that in PIAF), 18% of patients stopped amiodarone because of adverse events. Pulmonary toxicity was suspected in five patients (2.5%), although no deaths from pulmonary toxicity were noted (19). These results were similar to the experience in the AFFIRM substudy, with 3.8 years of mean follow-up.
The primary end point measured at one year might not, by itself, be the only appropriate criterion to evaluate antiarrhythmic drug therapy. Most patients with AF will require therapy for many years, even decades, and thus would be subject to longer term potential adverse effects of the drugs. It is possible that this “early” end point biases our results toward amiodarone, for which longer term adverse effects are a concern. However, it is one reasonable measure of drug effectiveness.
Not all antiarrhythmic drugs were tested in the AFFIRM substudy, nor were those tested used equally. Furthermore, class I drugs were considered as a group for purposes of the substudy, but they are not interchangeable.
Drug administration was not blinded, and adverse event reporting could have been biased.
The substudy required that a patient be able to take at least two rhythm control drugs. It is possible that efficacy would have been higher and toxicity lower for one or more drugs in a younger, healthier group—for example, patients with “lone” fibrillation and no risk factors for stroke, who were not eligible for AFFIRM.
It is difficult to know whether the doses administered of the different drugs were comparable across all patients to whom they were given. Written guidelines for dosing were part of the AFFIRM protocol, and they were adjusted within the limits of clinical judgment and patient tolerance. Thus, our results likely reflect what most experienced physicians would do in clinical practice. Furthermore, newer antiarrhythmic drugs, such as dofetilide, were not tested in this protocol.
Episodes of paroxysmal AF are frequently asymptomatic, even in patients who have some symptomatic episodes. Thus, a patient with sinus rhythm at one-year follow-up might be having asymptomatic recurrences that are undetected. Clinically, this phenomenon could be dangerous, especially if warfarin had been discontinued because it was thought that the patient had achieved successful rhythm control. This scenario may be particularly common with those antiarrhythmic drugs that also slow the ventricular rate during a recurrence of AF.
Accepted definitions for the three types of AF (paroxysmal, persistent, and permanent) (21)had not been articulated when this trial was in its planning stages (1). Thus, patients were not classified by this scheme. Our approximation that classified patients by the duration of known AF (<48 h or ≥48 h) of the qualifying episode is not precisely the same.
This AFFIRM substudy should be viewed in the context of the main AFFIRM results, which support rate control as an acceptable primary treatment. Among patients eligible to participate in AFFIRM in whom rhythm control drugs were randomly assigned, amiodarone was better than class I agents and sotalol for the maintenance of sinus rhythm at one year, defined as the patient being alive, still taking the drug, being in sinus rhythm, and needing no electrical cardioversions.
If amiodarone is not chosen as the primary antiarrhythmic drug, the use of another antiarrhythmic drug alone will yield about a 20% to 40% prevalence of sinus rhythm at one year without cardioversion. Sequential treatment with two or more antiarrhythmic drugs, with or without cardioversion, can achieve nearly 80% prevalence of sinus rhythm at one year.
The adverse effects of all drugs tested in this protocol were common over the first year.
If the rhythm control strategy is chosen for any given patient, antiarrhythmic drug treatment must remain individualized, considering any underlying disease, potential drug efficacy, and the potential risk of short-term and long-term exposure to that drug.
Amiodarone was generously supplied by Wyeth-Ayerst Laboratories (Philadelphia, Pennsylvania).
H. L. Greene, MD, FACC, Axio Research Corp., Seattle, Washington; A. L. Waldo, MD, FACC, Case Western Reserve University, Cleveland, Ohio.
Writing group members (listed alphabetically)
S. D. Corley, MS, Axio Research Corp., Seattle, Washington; B. Coutu, MD, Hopital Notre-Dame, Montreal, Quebec, Canada; M. J. Domanski, MD, FACC, National Heart, Lung, and Blood Institute, Bethesda, Maryland; R. T. Faillace, MD, FACC, Deborah Heart and Lung Center, Browns Mills, New Jersey; J. A. Gomes, MD, FACC, Mt. Sinai Medical Center, New York, New York; L. J. Haywood, MD, FACC, Los Angeles County and USC Medical Center, Los Angeles, California; R. A. Josephson, MD, FACC, Summa Health System and Northeast Ohio Universities College of Medicine, Akron, Ohio; M. H. Kim, MD, FACC, Northwestern University, Evanston, Illinois; S. G. Kim, MD, FACC, Montefiore Medical Center, Bronx, New York; S. Lee, MD, FACC, George Washington University Medical Center, Washington, DC; R. B. Leman, MD, FACC, Medical University of South Carolina, Charleston, South Carolina; P. E. Linz, MD, FACC, Naval Medical Center San Diego, San Diego, California; M. D. Meissner, MD, FACC, Harper Hospital and Sinai-Grace Hospital, Detroit, Michigan; M. C. Mickel, MS, Axio Research Corp.; J. D. Nelson, RN, Axio Research Corp.; D. L. Packer, MD, FACC, Mayo Foundation, Rochester, Minnesota; S. M. Pai, MD, FACC, Loma Linda University Medical Center, Loma Linda, California; E. V. Palileo, MD, FACC, Loyola University, Maywood, Illinois; J. J. Schutzman, MD, FACC, Methodist Hospital of Indiana, Indianapolis, Indiana; S. Swiryn, MD, FACC, Northwestern University, Evanston, Illinois; R. J. Weiss, MD, FACC, Androscoggin Cardiology Associates, Auburn, Maine; D. G. Wyse, MD, PhD, FACC, University of Calgary, Calgary, Alberta, Canada; G. M. Ziady, MD, FACC, University of Pittsburgh, Pittsburgh, Pennsylvania.
☆ The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Investigators and their affiliations are listed in reference 11. This study was supported by the National Heart, Lung, and Blood Institute (contract N01-HC-55139), National Institutes of Health, Bethesda, Maryland.
- atrial fibrillation
- Atrial Fibrillation Follow-up Investigation of Rhythm Management
- Received November 1, 2002.
- Revision received January 3, 2003.
- Accepted January 24, 2003.
- American College of Cardiology Foundation
- The National Heart, Lung, and Blood Institute Working Group on Atrial Fibrillation
- Nichol G.,
- McAlister F.,
- Pham B.,
- et al.
- Coplen S.E.,
- Antman E.M.,
- Berlin J.A.,
- Hewitt P.,
- Chalmers T.C.
- the Stroke Prevention in Atrial Fibrillation Investigators,
- Flaker G.C.,
- Blackshear J.L.,
- McBride R.,
- Kronmal R.A.,
- Halperin J.L.,
- Hart R.G.
- the Stroke Prevention in Atrial Fibrillation Investigators,
- Flaker G.C.,
- Blackshear J.L.,
- McBride R.,
- Kronmal R.A.,
- Halperin J.L.,
- Hart R.G.
- Lan K.K.G.,
- DeMets D.L.
- Miller R.G. Jr.
- Epstein A.E.,
- Carlson M.D.,
- Fogoros R.N.,
- Higgins S.L.,
- Venditti F.J. Jr.
- ↵Hohnloser SH, Kuck K-H, Lilienthal J, for the PIAF Investigators. Rhythm or rate control in atrial fibrillation—Pharmacological Intervention in Atrial Fibrillation (PIAF): a randomised trial. Lancet 2000;356:1789–94
- Fuster V.,
- Rydén L.E.,
- Asinger R.W.,
- et al.