Relapse and mortality following cardioversion of new-onset vs. recurrent atrial fibrillation and atrial flutter in the elderly

Transcription

1 European Heart Journal Advance Access published February 2, 2006 European Heart Journal doi: /eurheartj/ehi753 Clinical research Relapse and mortality following cardioversion of new-onset vs. recurrent atrial fibrillation and atrial flutter in the elderly Ahmad A. Elesber 1, Ana Gabriela Rosales 2, Regina M. Herges 2, Win-Kuang Shen 1, Brenda S. Moon 1, Joseph F. Malouf 1, Naser M. Ammash 1, Virend Somers 1, David O. Hodge 2, Bernard J. Gersh 1, Stephen C. Hammill 1, and Paul A. Friedman 1 * 1 Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA; and 2 Division of Biostatistics, Mayo Clinic, Rochester, MN, USA Received 6 July 2005; revised 15 December 2005; accepted 13 January 2006 KEYWORDS Atrial fibrillation; Atrial flutter; Cardioversion; Mortality; Relapse Introduction Atrial fibrillation is the most common sustained clinical arrhythmia. 1 Management strategies include restoration and maintenance of normal sinus rhythm or control of the ventricular rate while fibrillation persists in the atria. Recently completed trials in North America and Europe compared pharmacologic rhythm control with pharmacologic rate control in predominantly older patients who had risk factors for stroke, and the results showed no improvement in quality-of-life, incidence of stroke, or mortality with a strategy of maintaining normal rhythm. 2,3 Important limitations with the rhythm control strategy in these trials included the limited ability of anti-arrhythmic drugs to maintain sinus rhythm and the possible underutilization of anticoagulation in patients assumed to be in sinus rhythm. 2,3 However, in an Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) substudy 4 and in other studies, 5 7 outcomes were improved in the * Corresponding author. Tel: þ ; fax: þ address: friedman.paul@mayo.edu Aims Trials of rate control vs. rhythm control for atrial fibrillation or flutter included few patients with new-onset arrhythmia. Our objective was to assess the relapse rate and the effect of the relapse of newonset atrial arrhythmias on mortality after direct-current cardioversion (DCCV). Methods and results A cohort of 351 patients with atrial fibrillation (new onset in 179) and 126 patients with atrial flutter (new onset in 78) was followed-up after DCCV. Cox proportional hazard models were used. Median age was 74.6 years. Mean follow-up for relapse was 7.7 months; for death, 29.4 months. Patients with new-onset atrial flutter [adjusted hazard ratio (HR) ¼ 1] were more likely to maintain sinus rhythm than the patients with recurrent atrial flutter (adjusted HR ¼ 2.5, P, 0.01), new-onset atrial fibrillation (adjusted HR ¼ 2.4, P, 0.01), or recurrent atrial fibrillation (adjusted HR ¼ 2.7, P, 0.01). Patients with new-onset atrial fibrillation were as likely to have relapses as patients with recurrent atrial fibrillation or flutter. Relapse of atrial arrhythmia after DCCV was associated with increased mortality (adjusted HR ¼ 3.1, P, 0.01). Conclusion DCCV is more successful in maintaining sinus rhythm in patients with new-onset atrial flutter than in patients with new-onset atrial fibrillation. Relapse of atrial arrhythmia after cardioversion is associated with increased mortality. subset of patients in whom normal rhythm was maintained, although it is unclear whether sinus rhythm was a direct cause of the improved outcomes or a marker. The recent trials included few patients presenting with an initial episode of atrial fibrillation, so that the optimal management of this population is still unknown. The major conclusions of the trials suggest that attempts to maintain sinus rhythm are not warranted in the majority of patients who have symptoms that are well-controlled by rate control alone and who can tolerate anticoagulants. However, it is not clear whether this strategy should be applied to patients who present with a first episode of arrhythmia. It has been suggested that a trial of cardioversion without anti-arrhythmic medications may be warranted in patients who have had a recent onset of atrial fibrillation or atrial flutter, although data are lacking. 8 To better understand the therapeutic management of patients presenting with a new-onset sustained atrial arrhythmia, we compared the arrhythmia relapse rate and mortality among patients who had new-onset atrial fibrillation or atrial flutter undergoing cardioversion with those of patients who had recurrent atrial fibrillation or atrial flutter in which at least one & The European Society of Cardiology All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Page 2 of 7 A.A. Elesber et al. episode was documented before the index evaluation. Patients with atrial flutter were included because this arrhythmia is common and is associated with clinically significant morbidity. 9 Methods Mayo Cardioversion Unit All patients who undergo electrical cardioversion at Saint Marys Hospital in Rochester, Minnesota, undergo a standardized evidence-based treatment protocol. This includes assessment of anticoagulation status, use of transoesophageal echocardiography when indicated, and use of standardized cardioversion techniques. 10,11 At cardioversion, clinical information is abstracted and prospectively entered into a database by a specialized nurse. The database includes information on patients demographics, medical history, details on current and past episodes of atrial arrhythmias, medications used, and echocardiographic evaluation. Most patients underwent either transoesophageal or transthoracic echocardiography within 1 month of cardioversion. Left and right atrial sizes were estimated visually by the echocardiographer and classified into four categories: normal, mildly enlarged, moderately enlarged, and severely enlarged. Mitral valve disease was defined as the presence of moderate or severe mitral valve regurgitation or of moderate or severe mitral valve stenosis as characterized by standard echocardiographic techniques. Aortic valve disease was defined as the presence of moderate or severe aortic valve regurgitation or of moderate or severe aortic valve stenosis as characterized by standard echocardiographic techniques. Atrial arrhythmias: definitions Atrial arrhythmias were defined according to electrocardiographic findings. Atrial fibrillation was defined as a supraventricular tachyarrhythmia characterized by the replacement of P-waves with rapid oscillations or fibrillatory waves that varied in size, shape, and timing and that were associated with an irregular ventricular response when atrioventricular conduction was intact. Atrial flutter was defined as an organized, regular monomorphic atrial rhythm with an atrial rate between 240 and 350 bpm. 9 Atrial fibrillation or atrial flutter was categorized as new onset if an episode had not been documented before the index presentation or as recurrent if a previously documented episode had resolved either spontaneously or after cardioversion before the index presentation. Relapse was defined as an arrhythmic event after index cardioversion. The arrhythmic event must have been documented by electrocardiography and could have been either atrial fibrillation or atrial flutter. On the basis of these definitions, patients were categorized into one of the four groups: new-onset atrial fibrillation, recurrent atrial fibrillation, new-onset atrial flutter, and recurrent atrial flutter. Patients Consecutive patients undergoing cardioversion for persistent atrial fibrillation or atrial flutter between May 2000 and December 2001 were eligible for study if they did not meet the exclusion criteria, which included the following: younger than 60 years, congenital heart disease, thyrotoxicosis, antecedent surgical maze procedure, history of ablation for atrial arrhythmia or ablation after cardioversion but before clinical arrhythmia relapse, and an index atrial arrhythmia that developed within 1 month after any surgical procedure. Patients were also excluded if cardioversion failed, if sinus rhythm was maintained for less than 1 h after cardioversion, or if they had electrocardiographic documentation of both atrial fibrillation and atrial flutter in the 3 days before cardioversion. If a patient had more than one cardioversion on separate presentations, only the first episode was included in the study. Only patients who had follow-up electrocardiography were included in the analysis. Finally, patients who denied research authorization were excluded. The study was approved by the Mayo Foundation Institutional Review Board. Follow-up Information on relapse of atrial arrhythmia was collected from review of medical records, which included information from visits to Mayo Clinic and to other medical facilities. No formal follow-up for recurrence was set with the participants, but follow-up was scheduled as judged necessary by the physician and the patient. Data were recorded in a centralized computerized system containing complete records, including detailed history and diagnosis for all outpatient encounters (including emergency department visits, house calls, and nursing home visits), laboratory values and electrocardiograms, and inpatient care information. 12 Relapse or absence of relapse was documented with electrocardiographic findings. Information on all-cause mortality was collected from the Social Security Death Index, which is considered highly specific and unbiased. 13 Because information on relapse rates and mortality was collected by different methods, the follow-up times were different for the two clinical endpoints, relapse and mortality. Statistical analysis Continuous variables are reported as mean + SD or as median and inter-quartile range, and categorical variables as percentage. Continuous data were compared between the four study groups by the Kruskal Wallis test, and categorical data were compared by the x 2 test. Specific two-group comparisons of continuous variables were assessed by the Wilcoxon rank sum test. Univariable and multivariable associations of clinical and echocardiographic variables were assessed with Cox proportional hazards model for the two endpoints, relapse of arrhythmia and mortality. The assumption of proportional hazards for each endpoint was assessed by plotting the scaled Schoenfeld residuals for each independent variable against rank of time. These plots did not show an indication for a nonrandom pattern of association. In addition, correlations between the scaled Schoenfeld residuals for each variable and the rank of time were performed and found to be non-significant. In the multivariable models, known important predictors of the specific endpoint were forced into the model to assess the association of the variable of interest. The multivariable model for relapse of arrhythmia was created by forcing the following into the model: age, hypertension, congestive heart failure, left atrial enlargement, mitral valve disease, and anti-arrhythmic medication at dismissal. Three dummy variables were added to this model to evaluate the association of the four types of arrhythmia at baseline. The new-onset atrial flutter group was used as a reference group. For the mortality model, the relapse of arrhythmia at follow-up was added to the model after adjusting the model by age, congestive heart failure, hypertension, and coronary artery disease. The relapse of arrhythmia was analysed as a time-dependent covariate, and the echocardiographic variables were analysed as dichotomous variables in all models. The results of these analyses were summarized as hazard ratios with 95% confidence intervals. Endpoint-free survivals were estimated using the Kaplan Meier method. Twosided P-values of,0.05 were considered statistically significant. Bonferroni-adjusted P-values were applied for comparisons of more than two groups. All analyses were done using SAS version 8 (SAS Institute, Cary, North Carolina). Results Patient characteristics Out of a total of 1016 cardioversions between May 2000 and December 2001, 477 were included in the study. For the

3 Cardioversion of atrial arrhythmias Page 3 of 7 47 patients who had direct-current cardioversion (DCCV) on two separate occasions on separate days, only the first episode was entered in the study. Other reasons for exclusion included congenital heart disease (31 patients), surgical maze procedure (five patients), atrial ablation (17 patients), post-operative atrial fibrillation or atrial flutter (175 patients), failed DCCV (46 patients), lack of research authorization (12 patients), lack of follow-up relapse information (66 patients), presence of both atrial fibrillation and atrial flutter at the index presentation (58 patients), and age younger than 60 years (82 patients). Important baseline characteristics are presented in Table 1. There was no difference between the four groups studied (P. 0.10) with regard to past medical history of diabetes mellitus (14.7% for total population), hypertension (67.8% for total population), and pulmonary disease (16.2% for total population); to echocardiographic findings of moderate or severe mitral valve disease (18.8% for total population), moderate or severe aortic valve disease (4.3% for total population), moderate or severe left atrial enlargement (48.8% for total population), and moderate or severe right atrial enlargement (33.9% for total population); and to dismissal on the following medications: dihydropyridine calcium channel blockers (8.1% for total population), nondihydropyridine calcium channel blockers (21.2% for total population), digoxin (28.7% for total population), angiotensin-converting enzyme-inhibitors or angiotensin-ii receptor blockers (52.9% for total population), diuretics (44.6% for total population), and warfarin (91.5% for total population). Table 1 Characteristic Follow-up for mortality was complete for 475 patients (99.6%). The mean follow-up time for relapse or absence of relapse of atrial fibrillation or atrial flutter was months (median, 3.5 months). The mean follow-up time for survival was months (median, 30.3 months). Relapse of atrial fibrillation and atrial flutter At 1 year of follow-up, 63% of the patients in the new-onset atrial flutter group maintained normal sinus rhythm when compared with only 30% of patients in the new-onset atrial fibrillation group, 33% of patients in the recurrent atrial flutter group, and 35% of patients in the recurrent atrial fibrillation group (Figure 1). Interestingly, all groups had relapses of both types of arrhythmias but in a different pattern. In patients with new-onset atrial flutter, 71% of the relapses were atrial flutter and 29% were atrial fibrillation. In patients with recurrent atrial flutter, 61% of the relapses were atrial flutter and 39% were atrial fibrillation. In patients with new-onset atrial fibrillation, 5% of the relapses were atrial flutter and 95% were atrial fibrillation. Finally, in patients with recurrent atrial fibrillation, 8% of the relapses were atrial flutter and 92% were atrial fibrillation. There was no difference between the four groups with respect to the need for pacemaker implantation at followup for either symptomatic atrial dysrhythmias (e.g. sick sinus syndrome or atrioventricular blocks) or for atrioventricular node ablation for symptomatic treatment of the atrial Characteristics of patients undergoing cardioversion for persistent atrial flutter or atrial fibrillation a Recurrent AF (n ¼ 172) New-onset AF (n ¼ 179) Recurrent AFL (n ¼ 48) New-onset AFL (n ¼ 78) Total (n ¼ 477) P-value b Age (years), mean (SD) 74.4 (8.25) 74.5 (8.04) 73.9 (7.47) 75.8 (6.51) 74.6 (7.83) 0.33 Follow-up for recurrence 3.1 ( ) 2.8 ( ) 3.7 ( ) 6.2 ( ) 3.5 ( ) (months), median (IQR) Recurrence of AF or AFL (%) 128 (74.4) 127 (70.9) 36 (75.0) 28 (35.9) 319 (66.9) Follow-up for death 30.5 ( ) 31.1 ( ) 29.9 ( ) 28.6 ( ) 30.3 ( ) (months), median (IQR) Death (%) 24/170 (14.1) 28 (15.6) 11 (22.9) 17 (21.8) 80 (16.8) Ejection fraction (%), 53.9 (12.55) 51.3 (14.76) 48.2 (16.49) 49.0 (15.86) 51.5 (14.48) 0.10 mean (SD) Gender,0.01 c Female (%) 77 (44.8) 68 (38.0) 15 (31.2) 17 (21.8) 177 (37.1) Male (%) 95 (55.2) 111 (62.0) 33 (68.8) 61 (78.2) 300 (62.9) PMH of CHF (%) 27 (15.7) 46 (25.7) 13/46 (28.3) 19 (24.4) 105/475 (22.1) 0.08 CAD (%) 55/170 (32.4) 52/177 (29.4) 22/47 (46.8) 35/77 (45.5) 164/471 (34.8) 0.02 Estimated onset of 0.08 present arrhythmia Number of patients ,48 h (%) 18 (11.3) 13 (8.7) 5 (21.8) 7 (11.5) 43 (10.5) 48 h to 1 mo (%) 55 (34.6) 42 (28.0) 20 (51.3) 24 (39.3) 141 (34.5).1 mo (%) 86 (54.1) 95 (63.3) 14 (35.9) 30 (49.2) 225 (55.0) Anti-arrhythmic 103/171 (60.2) 43/176 (24.4) 34 (70.8) 15/76 (19.7) 195/471 (41.4),0.01 medication (%) b-blockers (%) 85 (49.4) 82 (45.8) 17 (35.4) 25 (32.1) 209 (43.8) 0.04 AF, atrial fibrillation; AFL, atrial flutter; CAD, coronary artery disease; CHF, congestive heart failure; IQR, interquartile range; PMH, past medical history. a Denominators in data cells indicate incomplete data and sample sizes smaller than those in the column headings. Categorical data are presented as number of patients and percentage of sample. b Continuous data were compared among the four study groups by the Kruskal Wallis test; categorical data, by the x 2 test. c The reported P-value refers to the overall comparison among the four study groups. However, the specific comparison of recurrent AF vs. new-onset AFL was also significant.

4 Page 4 of 7 A.A. Elesber et al. arrhythmia after pharmacologic failure (11.11% for the overall group at a median follow-up of 7.7 months). Results of the univariable model are summarized in Table 2. Predictors of relapse of atrial fibrillation or atrial flutter were the following: presentation with new-onset atrial fibrillation, recurrent atrial fibrillation, or recurrent atrial flutter as opposed to presentation with new-onset atrial flutter; lack of past medical history of coronary artery disease; and a larger left atrium. Use of anti-arrhythmics did not have a significant effect on the relapse rate in the univariable analysis. The following variables did not affect the relapse rate of atrial fibrillation or atrial flutter: sex, hypertension, past medical history of congestive heart failure, pulmonary Figure 1 Percentage of patients who maintained normal sinus rhythm during follow-up. Patients were categorized into four groups based on the arrhythmia [atrial fibrillation (AF) or atrial flutter (AFL)] and on whether the arrhythmia had been documented before the index presentation ( recurrent ) or not ( new onset ): new-onset AF, recurrent AF, new-onset AFL, and recurrent AFL. disease, or diabetes mellitus; the estimated time of onset of the index atrial arrhythmia; left ventricular ejection fraction; mitral or aortic valve disease; and the use after cardioversion of warfarin, digoxin, diuretics, calcium channel blockers, b-blockers, or angiotensin receptor blockers and angiotensin-converting enzyme-inhibitors (not all data shown). Results of the multivariable analysis are summarized in Table 3. Patients with new-onset atrial flutter had significantly fewer relapses of atrial arrhythmia after cardioversion than the other three groups. This held after correcting for age, past medical history of hypertension or congestive heart failure, left atrial enlargement, moderate or severe mitral valve disease, and use of anti-arrhythmic medications at discharge. Patients with new-onset atrial fibrillation were no more likely to maintain sinus rhythm after cardioversion than the patients with recurrent atrial fibrillation (Figure 1). Survival Results of the univariable analysis are summarized in Table 4. The predictors of increased mortality were relapse of atrial fibrillation or atrial flutter at follow-up; older age; past medical history of coronary artery disease, congestive heart failure, or pulmonary disease; moderate to severe congestive heart failure at presentation (New York Heart Association Functional Class III or IV); a lower ejection fraction; presence of moderate or severe mitral disease; use of non-dihydropyridine calcium channel blockers, digoxin, or diuretics after cardioversion; and no use of b-blockers after cardioversion. The following factors did not have a statistically significant influence on survival: left or right atrial size, aortic valve disease, use of anti-arrhythmic medications, and use of angiotensin Table 2 Univariable associations with relapse of atrial fibrillation or atrial flutter after index cardioversion (366 events) Type of arrhythmia at baseline Recurrent atrial fibrillation ,0.01 New-onset atrial fibrillation ,0.01 Recurrent atrial flutter ,0.01 New-onset atrial flutter reference 1.00 Age Male vs. female Past medical history Coronary artery disease Hypertension CHF Pulmonary disease Diabetes mellitus Severe CHF at presentation (NYHA class III or IV) Estimated onset of atrial arrhythmia continuous Echocardiographic variables Ejection fraction Right atrial enlargement moderate or severe Left atrial enlargement moderate or severe Dismissal medication Anti-arrhythmic medication Warfarin NYHA, New York Heart Association.

5 Cardioversion of atrial arrhythmias Page 5 of 7 receptor blockers and angiotensin-converting enzyme inhibitors (not all data shown). The multivariable analysis (Table 5) demonstrated that after correcting for advanced age and past medical history of coronary artery disease, congestive heart failure, or hypertension, the relapse of atrial fibrillation or atrial flutter after cardioversion was a significant predictor of death. The type of atrial arrhythmia at baseline (recurrent atrial fibrillation, new-onset atrial fibrillation, recurrent atrial flutter, or new-onset atrial flutter) was not (Figure 2). Table 3 Multivariable model for recurrence of atrial fibrillation or atrial flutter after index cardioversion (454 patients; 305 events) Age Hypertension PMH of CHF Left atrial enlargement Mitral valve disease Anti-arrhythmic at dismissal Atrial fibrillation Recurrent ,0.01 New onset ,0.01 Atrial flutter Recurrent ,0.01 New onset 1.00 Table 4 Discussion Our study showed that patients with new-onset atrial flutter are more likely to maintain sinus rhythm after DCCV than the patients with new-onset atrial fibrillation, patients with recurrent atrial fibrillation, and patients with recurrent atrial flutter even though less anti-arrhythmic medication was used in the new-onset atrial flutter group. The relapse rate of atrial arrhythmias was similar between patients with new-onset atrial fibrillation and patients with recurrent atrial fibrillation or recurrent atrial flutter. Relapse of atrial arrhythmias after DCCV was associated with increased mortality. Management of new-onset atrial fibrillation and atrial flutter In recent prospective randomized trials, pharmacologic management with a rhythm control strategy offered no advantage over rate control, and rhythm control had the disadvantage of greater drug toxicity. 2,3,14 However, patients with new-onset arrhythmia either were a small minority or were excluded from these studies, 2,3,14 and it has been unclear whether atrial fibrillation should be permitted to persist after an initial episode or whether cardioversion without rhythm-controlling medications should be given for the first episode. 8 Our patient population with new-onset atrial fibrillation had a very high risk of arrhythmia relapse after cardioversion (70.1% at 12 months) (Figure 1), which was similar to that of patients with a history of recurrent Univariable associations with death after index cardioversion of atrial fibrillation or atrial flutter (80 events) Type of arrhythmia at baseline Recurrent atrial fibrillation New-onset atrial fibrillation Recurrent atrial flutter New-onset atrial flutter reference 1.00 Relapse at follow-up (time-dependent) ,0.01 Age ,0.01 Male vs. female Past medical history Coronary artery disease Hypertension CHF ,0.01 Pulmonary disease Diabetes mellitus Severe CHF at presentation (NYHA class III or IV) ,0.01 Estimated onset of atrial arrhythmia continuous Echocardiographic variables Ejection fraction Mitral valve disease Dismissal medication b-blockers Non-dihydropyridine calcium channel blockers ,0.01 Digoxin Anti-arrhythmic ACE-inhibitor and ARB Diuretics ,0.01 Warfarin ACE, angiotensin-converting enzyme; ARB, angiotensin-ii receptor blocker.

6 Page 6 of 7 A.A. Elesber et al. Table 5 Multivariable model for death after index cardioversion of atrial fibrillation or atrial flutter (466 patients; 79 events) Age ,0.01 History of CHF ,0.01 Hypertension CAD ,0.01 Relapse at follow-up (time-dependent) ,0.01 Patients with new-onset atrial flutter had a 62.7% likelihood of remaining in normal rhythm 1 year after cardioversion, which was significantly greater than that for patients with atrial fibrillation or recurrent atrial flutter. This was true even after correcting for baseline characteristics including age and comorbidities. The high rate of maintenance of normal rhythm occurred even though this group used anti-arrhythmic medications less frequently. This finding suggests that cardioversion in the absence of anti-arrhythmic medications might be a reasonable strategy for new-onset atrial flutter. Careful surveillance is warranted, however, and discontinuation of indicated anticoagulation should be avoided, in light of the long-term risk of thrombo-embolism or subsequent atrial fibrillation. 2,9 Figure 2 Survival curves for all-cause mortality. atrial fibrillation at the time of index cardioversion. Our data suggest that a trial of cardioversion without antiarrhythmic drugs is not very effective in elderly patients. Only small proportions of our patients with new-onset atrial fibrillation received anti-arrhythmic drugs with or after cardioversion. This is consistent with the clinical strategy of cardioversion without drugs for new-onset atrial fibrillation. Prospective studies are needed to assess whether the strategy of using anti-arrhythmic medications after cardioversion in symptomatic elderly patients with new-onset atrial fibrillation is effective. The fact that patients with new-onset atrial fibrillation had a similar rate of relapse as patients with recurrent atrial fibrillation underlies the importance of using anticoagulation even in patients with new-onset atrial fibrillation. Anticoagulation in this group of patients should be discontinued only after strong consideration of the risks and benefits. Importantly, our population was similar to patients included in the AFFIRM and RACE trials with respect to age, sex, and cardiac comorbidities as well as other medical comorbidities, and probably reflects the characteristics of the majority of patients with atrial fibrillation. 15 The long-term prevalence of atrial fibrillation in the AFFIRM rhythm-control arm was lower than that in our study population. The inclusion of patients with paroxysmal atrial fibrillation, the greater use of anti-arrhythmic medications, and the repeated use of cardioversions most likely account for this finding. However, that strategy was the cause of more frequent side effects, hospitalizations, and, potentially, costs incurred by patients in the rhythmcontrol arm. Mortality Previous studies have shown an association between atrial fibrillation and mortality in patients with structural heart disease. 5 7 In contrast to these studies, few patients in the present study had heart failure, and the median ejection fraction was normal. Despite studying a generally healthier population, we observed for the first time that the relapse of atrial arrhythmia after cardioversion is associated with increased mortality. It is not clear whether this increased mortality resulted from haemodynamic or pro-arrhythmic effects of the atrial arrhythmia, 16,17 or whether it reflected an association of relapse with other yet undetermined risk factors for mortality. Irrespective of the potential mechanisms, atrial arrhythmia that recurs after cardioversion deserves special attention because of the associated increased risk of death. We found an association between the use of a diuretic, digoxin, or a non-dihydropyridine calcium channel blocker after cardioversion and an increase in mortality rates in the univariable analysis. The design of our study does not permit us to determine whether this finding was because of a cause-and-effect relationship or whether it was because of an association of the use of these medications with sicker patients. Limitations Our conclusions cannot be extended to younger patients. Patients with symptoms that persist after rate control of atrial fibrillation or atrial flutter also likely need a trial of cardioversion. Patients in whom rhythm-controlling nonpharmacologic therapy is offered may warrant a trial of cardioversion, although such therapies are infrequently offered to elderly patients at high risk of stroke. 18 The present study was not randomized. However, use of a consecutive series of patients in a large cardioversion practice was used to minimize selection bias and to accurately reflect clinical practice. It is well established that atrial fibrillation may recur without symptoms and that the observed recurrence rate most likely underestimates the actual recurrence. However, as there is no reason to believe that an asymptomatic recurrence is more likely in any one of the four groups studied, this should not affect the results. The fact that the four groups had similar rates of congestive heart failure symptoms at presentation, as assessed by use of the New York Heart Association classification, supports this concept.

7 Cardioversion of atrial arrhythmias Page 7 of 7 Conclusion The relapse rate of atrial arrhythmias in patients with newonset atrial flutter after DCCV is significantly lower than in patients with new-onset atrial fibrillation, recurrent atrial fibrillation, or recurrent atrial flutter even though less anti-arrhythmic medication was used in the new-onset atrial flutter group. The relapse rate among patients with new-onset atrial fibrillation after DCCV is similar to that among patients with recurrent atrial fibrillation and recurrent atrial flutter. Careful attention is warranted when atrial arrhythmias recur in patients after successful cardioversion because they have a higher mortality rate than in patients who maintain sinus rhythm. Acknowledgements There was no grant support for this manuscript. Conflict of interest: none declared. References 1. Fuster V, Ryden LE, Asinger RW, Cannom DS, Crijns HJ, Frye RL, Halperin JL, Kay GN, Klein WW, Levy S, McNamara RL, Prystowsky EN, Wann LS, Wyse DG, Gibbons RJ, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Hiratzka LF, Jacobs AK, Russell RO, Smith SC Jr, Klein WW, Alonso-Garcia A, Blomstrom-Lundqvist C, de Backer G, Flather M, Hradec J, Oto A, Parkhomenko A, Silber S, Torbicki A. ACC/AHA/ESC Guidelines for the Management of Patients With Atrial Fibrillation: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines and Policy Conferences (Committee to Develop Guidelines for the Management of Patients With Atrial Fibrillation) Developed in Collaboration With the North American Society of Pacing and Electrophysiology. Circulation 2001;104: The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Investigators. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002; 347: Van Gelder IC, Hagens VE, Bosker HA, Kingma JH, Kamp O, Kingma T, Said SA, Darmanata JI, Timmermans AJ, Tijssen JG, Crijns HJ, for the Rate Control versus Electrical Cardioversion for Persistent Atrial Fibrillation Study Group. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002;347: The AFFIRM Investigators. Relationships between sinus rhythm, treatment, and survival in the Atrial Fibrillation Follow-Up Investigation of Rhythm Management (AFFIRM) Study. Circulation 2004;109: Published online ahead of print March 08, Deedwania PC, Singh BN, Ellenbogen K, Fisher S, Fletcher R, Singh SN, for the Department of Veterans Affairs CHF-STAT Investigators. Spontaneous conversion and maintenance of sinus rhythm by amiodarone in patients with heart failure and atrial fibrillation: observations from the Veterans Affairs Congestive Heart Failure Survival Trial of Antiarrhythmic Therapy (CHF-STAT). Circulation 1998;98: Pedersen OD, Bagger H, Keller N, Marchant B, Køber L, Torp-Pedersen C, for the Danish Investigations of Arrhythmia and Mortality ON Dofetilide Study Group. Efficacy of dofetilide in the treatment of atrial fibrillation-flutter in patients with reduced left ventricular function: a Danish Investigations of Arrhythmia and Mortality ON Dofetilide (DIAMOND) substudy. Circulation 2001;104: Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. J Am Coll Cardiol 1998;32: Falk RH. Management of atrial fibrillation: radical reform or modest modification? NEnglJMed2002;347: Halligan SC, Gersh BJ, Brown RD Jr, Rosales AG, Munger TM, Shen WK, Hammill SC, Friedman PA. The natural history of lone atrial flutter. Ann Intern Med 2004;140: Grimm RA, Black IW, Klein AL. Transesophageal echocardiography before cardioversion [letter]. NEnglJMed1993;329: Malouf JF, Ammash NM, Chandrasekaran K, Friedman PA, Khandheria BK, Gersh BJ. Critical appraisal of transesophageal echocardiography in cardioversion of atrial fibrillation. Am J Med 2002;113: Ozcan C, Jahangir A, Friedman PA, Patel PJ, Munger TM, Rea RF, Lloyd MA, Packer DL, Hodge DO, Gersh BJ, Hammill SC, Shen WK. Long-term survival after ablation of the atrioventricular node and implantation of a permanent pacemaker in patients with atrial fibrillation. N Engl J Med 2001;344: Cowper DC, Kubal JD, Maynard C, Hynes DM. A primer and comparative review of major US mortality databases. Ann Epidemiol 2002; 12: Carlsson J, Miketic S, Windeler J, Cuneo A, Haun S, Micus S, Walter S, Tebbe U, for the STAF Investigators. Randomized trial of rate-control versus rhythm-control in persistent atrial fibrillation: the Strategies of Treatment of Atrial Fibrillation (STAF) study. J Am Coll Cardiol 2003; 41: Kannel WB, Abbott RD, Savage DD, McNamara PM. Coronary heart disease and atrial fibrillation: the Framingham Study. Am Heart J 1983; 106: Stein KM, Euler DE, Mehra R, Seidl K, Slotwiner DJ, Mittal S, Markowitz SM, Lerman BB, for the Jewel AF Worldwide Investigators. Do atrial tachyarrhythmias beget ventricular tachyarrhythmias in defibrillator recipients? J Am Coll Cardiol 2002;40: Daoud EG, Weiss R, Bahu M, Knight BP, Bogun F, Goyal R, Harvey M, Strickberger SA, Man KC, Morady F. Effect of an irregular ventricular rhythm on cardiac output. Am J Cardiol 1996;78: Pappone C, Rosanio S, Augello G, Gallus G, Vicedomini G, Mazzone P, Gulletta S, Gugliotta F, Pappone A, Santinelli V, Tortoriello V, Sala S, Zangrillo A, Crescenzi G, Benussi S, Alfieri O. Mortality, morbidity, and quality of life after circumferential pulmonary vein ablation for atrial fibrillation: outcomes from a controlled nonrandomized long-term study. J Am Coll Cardiol 2003;42: