Characterization of Different Subsets of Atrial Fibrillation in General Practice in France
The ALFA Study
Background—The clinical presentation and causes of atrial fibrillation (AF) in the 1990s may differ from AF seen 2 to 3 decades ago. It was the objective of this prospective study to characterize various clinical presentations and underlying conditions of patients with AF observed in general practice in France.
Methods and Results—The study population comprised 756 patients (19 to 95 years of age) with electrocardiographically documented AF subdivided into paroxysmal (<7 days), chronic (last episode >1 month) and recent onset AF(persistent >7 days and<1 month). Symptoms were present in 670 patients (88.6%). The relative prevalences of paroxysmal, chronic, and recent onset AF were 22.1%, 51.4%, and 26.4%, respectively. Cardiac disorders, present in 534 patients (70.6%), included hypertension (39.4%), coronary artery disease (16.6%), and myocardial diseases (15.3%) as the most common. Rheumatic valvular disease represented a common cause in women (25.0%) but not in men (8.0%). The paroxysmal group differed by a high percentage of palpitations (79.0%) and a low percentage of underlying heart disease (53.9%). With a mean follow-up of 8.6±3.7 months, 28 patients (3.7%) died, including 6 fatal cerebrovascular accidents. Among the 728 patients who survived, congestive heart failure occurred in 30 patients (4.1%), and embolic complications occurred in 13 patients (1.8%). In the paroxysmal AF group, 13 patients (8.0%) developed chronic AF and 51 (31.3%) had AF recurrences. At the time of follow-up, 53 patients (14.3%) from the chronic AF group and 108 patients (55.7%) from the recent onset AF group were in sinus rhythm.
Conclusions—This large-scale study establishes the current demographic profile of out-of-hospital patients with AF and highlights some of the changes that have occurred in the past decades, including a particular shift in cardiac causes toward nonrheumatic AF. This study also demonstrates significant differences between various subsets of AF.
There has been, in recent years, an increased awareness of the morbidity and mortality associated with atrial fibrillation (AF).1 2 3 4 Symptoms of AF, embolic complications, and left ventricular dysfunction are responsible for frequent physician visits and hospitalizations at a high cost.5 Few studies have attempted to characterize the various clinical presentations and causes of AF.6 7 Data on the causes of AF, derived from studies published 2 to 3 decades ago,8 9 10 11 12 13 14 15 may not be applicable to the AF population seen in the 1990s (particularly regarding the prevalence and type of underlying heart diseases).
The purpose of this prospective study was to characterize causes of different subsets of AF observed in general practice in France and to evaluate the outcome of these patients after 6 to12 months of follow-up.
This prospective study, entitled ALFA (Etude en Activité Libérale de la Fibrillation Auriculaire), was undertaken under the auspices of the College of French Cardiologists and involved 206 cardiologists established in general office practice and distributed in all regions of France. This survey involved only general cardiologists in office practice, representative of the vast majority of cardiologists in France. Each cardiologist agreed to enroll, to report on, and to follow 6 consecutive patients with documented AF.
Electrocardiographic diagnosis of AF was made according to Bellet’s definition.16 AF was subdivided into 3 types: paroxysmal, chronic, or recent onset. Paroxysmal AF was diagnosed in a patient with a history of recurrent episodes of AF lasting >2 minutes and <7 days. Patients with a first episode of AF lasting <7days or cardioverted within 7 days were also classified in this group. Chronic AF was defined as AF present for >1 month. Recent onset AF was defined as persistent (nonself-terminating) AF lasting ≥7 days and <1 month. A first symptomatic attack of AF lasting ≥7 days and <1 month, an asymptomatic or mildly symptomatic AF of recent discovery, or an AF episode for which the onset could not be determined were classified in this group. Should the physician opt for cardioversion (either pharmacological or electrical) of AF lasting >7 days before 1 month, the patient was classified in the recent onset AF group.
Hypertension was diagnosed when a history of hypertension in a treated patient for this condition or a diastolic blood pressure (≥95 mm Hg) or both were present. Hypertensive heart disease was diagnosed in a hypertensive patient if there was echocardiographic evidence of left ventricular hypertrophy17 and/or a history of left ventricular failure with normal systolic function. Coronary artery disease was diagnosed as definitive when the patient had a documented history of myocardial infarction and/or significant (≥70%) obstructive lesion on coronary angiogram and/or a history of coronary revascularization. It was diagnosed as probable when typical anginal syndrome was associated with evidence of reversible myocardial ischemia using a noninvasive method.
Inclusion Criteria and Data Collection
In order to be included in the study, a patient had to be in AF at the initial visit and/or have a history of AF with one or more AF episode documented on ECG, treated or untreated. The report form included patient characteristics, the type of AF according to the classification described, the presence and type of underlying heart disease, the current treatment, the results of thyroid function tests, 12-lead ECG, M mode and 2D echocardiogram findings, and, whenever indicated, the results of 24-hour ECG ambulatory monitoring. Doppler evaluation was not required. Although the treatment was left to the decision of each investigator, the investigators were provided with general guidelines regarding prevention of embolic complications using warfarin or equivalent agents.
A follow-up visit was required within 6 to 12 months after inclusion. Data were collected on a specially designed follow-up form. For paroxysmal AF, maintenance of sinus rhythm, recurrence of AF with frequency and duration of episodes, and class, using the classification system or evolution to chronic AF, were recorded. Complications, including embolic events, congestive heart failure, or other major events were recorded as was therapy instituted between the 2 examinations. Deaths were classified as cardiovascular (sudden unexpected [occurring within 1 hour of new symptom] and nonsudden), noncardiovascular, or unknown, using the classification reported by Julian et al.18
Data Quality Control and Analysis
All data forms were reviewed by a cardiologist trained in conducting clinical trials. A validation committee was asked to approve or reject the case-report form or to ask for missing information. A minimum of 1 on-site visit was done. In addition, on-site audits were also performed randomly or whenever ordered by the validation committee.
Normality of distributions was evaluated for each variable using the test of Shapiro and Wilk and the standardized coefficients of skew distribution and kurtosis.19 Nominal variables were compared by contingency table analysis. Continuous variables were compared by the Mann-Whitney test and the analysis of nonparametric variance of Kruskal-Wallis.20 Mortality and embolic event probabilities were estimated using the Kaplan-Meier method and a multivariate analysis according to the Cox model.
All data forms were collected in December 1994. Among the 1265 data forms collected, the Validation Committee found that 1043 fulfilled the entry criteria for the initial examination. Reasons for exclusion in 222 data forms included failure to provide an ECG of the patient in AF (n=215) and/or incomplete data (n=7). The present report includes the 756 patients who agreed to participate in the study and to be reevaluated at 6- to 12-month follow-up. The clinical characteristics of the study group did not differ significantly from the overall population of patients which entered this study.
Clinical Characteristics at Initial Examination
Total Patient Population
Characteristics of the 756 patients are shown in Table 1⇓. The age ranged from 19 to 95 years. The mean age was significantly (P<0.0001) higher in the female population (71.4±11.4 years) than in the male population (66.5±11 years). Evaluated by decades, the male/female ratio decreased after 60 years of age (Figure 1⇓). The mean time from first episode of AF was 47±63 months (range 0 to 15 years) and the median was 24 months. In 152 patients (20.2%), the time from first episode of AF was <1 month; in 140 patients (18.5%), it was 1 to 12 months; in 392 patients (51.9%), it was >12 months; and in 72 patients (9.5%), it was unknown. Underlying heart disease was present in 534 patients (70.6%) and was significantly (P<0.001) more common in women (76.9%) than in men (66.1%). Coronary artery disease in 126 patients (16.6%) was diagnosed as definite in 84 patients (46 of whom had a history of myocardial infarction and 38 had one or more significant coronary artery lesion at angiography) and probable in 42 patients (see definitions below). Rheumatic mitral valve disease included mitral stenosis in 38, mitral incompetence in 17, and combination in 22. Rheumatic aortic valve disease included incompetence in 12, stenosis in 4, and combination in 2. Twenty patients had both mitral and aortic valve disease. History of valvular surgery was found in 10 patients with mitral valve disease, 1 with aortic valve disease, and 2 with mitral and aortic valve disease. Nonrheumatic valvular heart disease, present in 25, included mitral valve prolapse in 19 patients (of which, 7 had mitral incompetence). Among the patient population with detectable heart disease, 156 patients (29.2%) were in class I of the NYHA, 258 (48.3%) in class II, 93 (17.4%) in class III, and 27 (5.0%) in class IV. Symptoms present in 670 patients (88.6%), as shown in Table 2⇓, were slightly although significantly (P<0.004) more frequent in the female population (92.5%) than in the male population (85.8%). Chronic AF was more common (51.4%) than paroxysmal (22.1%) or recent onset (26.4%) AF. No association was noted between gender and the type of AF.
At the time of the first office visit, 550 patients (72.7%) were being treated with antiarrhythmic therapy, including sodium channel blockers in 147 (19.4%), class IA in 46, class IB in 2, class IC in 99, amiodarone in 173 (22.4%), sotalol in 17 (2.2%), beta-blockers in 79 (10.4%), calcium channel blockers in 36 (4.7%), and digitalis glycosides in 302 (39.9%). Regarding anticoagulant therapy, 276 patients (36%) were on warfarin or a similar agent, 177 (23.4%) were on aspirin, and 18 (2.4%) on heparin.
A satisfactory echocardiogram was available in 591 patients (78.1%). As seen in Table 3⇓, the patients with AF had a significantly larger left atrial diameter, lower ejection fraction, and lower fractional shortening than those in sinus rhythm.
Age ranged from 19 to 91 years among the patients with paroxysmal AF. Forty-five patients (26.9%) were ≤60 years of age. The time from the first episode of AF had a median of 24.0 months. As shown in Table 2⇑, palpitations constituted the main complaint. Frequency, duration of the longest episode of AF, and mode of termination of the attacks of paroxysmal AF are shown in Table 4⇓. Organic heart disease was present in only 90 patients (53.9%) (as detailed in Table 1⇑). Antiarrhythmic therapy, aimed at preventing recurrences, was used in 126 patients (75.4%). Seventy patients had one or more cardioversion, (pharmacological in 61 and electrical in 9). Fifty-six patients (33.5%) received an agent for control of heart rate. Forty-three patients (25.7%) received warfarin or a similar agent.
This group, which ranged in age from 29 to 95 years, was significantly (P<0.0002) older than the paroxysmal AF group (Table 1⇑). Dyspnea was a more common complaint than in the paroxysmal group (P<0.0001). The time from first episode of AF ranged from 1 to 420 months with a median of 39 months. The duration of current or last episode of AF had a median of 22.8 months (range 1 to 420). Paroxysmal AF was found to have preceded chronic AF in 141 patients (36.2%). Valvular heart disease and cardiomyopathies were significantly more common (P<0.01 and P<0.001, respectively) in this group than in the paroxysmal group (Table 1⇑). A history of congestive heart failure or diabetes was significantly more common (P<0.0001) than in other types of AF. A prior thromboembolic event had occurred in 42 patients (10.8%).
At the time of the first examination, 240 patients (61.7%) were receiving digoxin or digitalis; 40 patients (10.3%), beta-blockers; and 21(5.4%), calcium channel antagonists. In patients with chronic AF, 106 (27.2%) were receiving amiodarone and 10 (2.6%) were receiving sotalol. Sodium channel blockers were being used in 48 patients (12.3%), mainly because of associated ventricular arrhythmias. One or more previous pharmacological cardioversions were reported in 141 patients (36.2%). Electrical cardioversion was attempted in 117 (30.1%) patients, with 1 session in 68 patients (17.5%) and >1 session in 49 patients (12.6%). An oral anticoagulant therapy was prescribed in 203 patients (52.2%) from this group. The mean left atrial diameter (range 22 to 83 mm) was significantly (P<0.0001) larger than in other groups of AF. Left ventricular ejection fraction was significantly lower (P<0.01) than in the other groups (Table 3⇑).
Recent Onset AF
This group included 128 patients with first episode of symptomatic AF lasting >7 days and <1 month and 72 patients with first discovery of asymptomatic or mildly symptomatic AF (or for whom the onset of current episode could not be determined). Age ranged from 34 to 92 years. Dyspnea was the most common symptom (58.0%) in this group. Other symptoms are shown in Table 2⇑. The underlying heart diseases are described in Table 1⇑. In this group, 92 patients (46.0%) were receiving ≥1 pharmacological agent, including 35 patients (17.5%) on oral amiodarone for pharmacological cardioversion or as pretreatment before electrical cardioversion.
Treatment at the time of the last follow-up is shown in Table 6. Of the 740 patients for whom complete data on antiarrhythmic and anticoagulant treatment were available, 683 patients (92.3%) were on ≥1 antiarrhythmic agents. In 359 patients (48.5%), antiarrhythmic treatment was introduced or changed. Electrical cardioversion was performed in 91 patients between the 2 office visits, including 44 patients (11.6%) with chronic AF, 46 patients with recent onset AF, and 1 patient with paroxysmal nonself-terminating AF. Anticoagulant therapy is also presented in Table 5⇓. Warfarin or a similar agent were prescribed between the 2 office visits in 105 patients (14.2%), aspirin in 89 patients(12.0%), and heparin in 2 patients.
Outcome for the Total Patient Population
The clinical outcome was assessed with a mean follow-up of 8.6±3.7 months (range 6 to 22). During this period, 28 of the 756 patients (3.7%) died. The age of the patients who died ranged from 39 to 95 years, with a mean of 77.6±11.8 years. Among the patients who died, 18 (64.2%) had chronic AF. The deaths were cardiovascular in origin in 18 patients (64.2%), including 7 sudden deaths, 5 nonsudden cardiac deaths, and 6 cerebrovascular accidents (embolic in 5 and cerebral hemorrhage in 1 patient on warfarin). The remaining deaths were noncardiac in 5 (cancer in 4) and of unknown cause in 5. Actuarial probability of dying for the study population is presented in Figure 2⇓.
Among the 728 patients who survived, there were 55 nonlethal major events including congestive heart failure in 30 patients (4.1%). Embolic complications occurred in 13 patients (1.8%), including nonlethal cerebrovascular accidents in 10 and acute limb ischemia in 3. Of these, 5 patients were on warfarin or a similar agent, 6 on aspirin, 1 on ticlopidine, and 1 on no anticoagulant therapy. An additional 2 patients had a cerebrovascular accident for which the embolic origin could not be ascertained. Actuarial probability of embolic event for the study population is shown in Figure 3⇓. Significant hemorrhagic complications occurred in 5 patients and included knee hematoma (n=2), major epistaxis (n=1), hematuria (n=1), or hemoptysis (n=1). An additional 37 events, related in part to pharmacological therapy, were reported including amiodarone-induced hyperthyroidism in 6 and severe bradycardia in 9 (with 2 of the 9 patients requiring a permanent pacemaker).
At the follow-up visit, 260 of the 728 survivors (35.7%) were in sinus rhythm and had remained free of symptomatic recurrences of AF. Underlying heart disease was present in 523 of the 728 patients (71.8%). Among these, 192 patients were in class I of the NYHA classification, 227 in class II, 79 in class III, and 25 in class IV. Comparing the initial examination status to that of the follow-up examination, only 122 patients (22.8%) were in a different class: 71 patients (13.2%) were in a lower class and 51 patients (9.5%) were in a higher class.
Four patients (2.3%) died during the follow-up period of 8.6±3.6 months, including 1 sudden death and 1 death from stroke. Clinical outcome is shown in Figure 4⇓. Among the 51 patients who had recurrence of AF, the precipitating factors could be clearly identified in 19 patients and included exercise or emotional stress (n=10), postprandial period and overeating or occurrence during the nocturnal period only (n=5), caffeine or alcohol (n=2), fever (n=2), or combination(n=5).
Eighteen patients (4.6%) died during the follow-up period (mean:9±4 months), including 5 sudden deaths and 4 deaths from stroke. Six patients (1.5%) experienced a nonlethal embolic stroke. Among the 371 patients who survived, 316 (85.2%) remained in chronic AF, 53 (14.3%) maintained sinus rhythm after a successful cardioversion (electrical in 23 and pharmacological in 30), and 2 (0.5%) had sinus rhythm restored but developed paroxysmal AF (Figure 4⇑).
Recent Onset AF
Cardioversion was considered after the first visit in 177 patients (88.5%): electrical in 31 (15.5%) and pharmacological in 146 patients (73.0%). Sinus rhythm was restored in 148 patients (spontaneously in 2 and after pharmacological cardioversion in 116 or electrical cardioversion in 30). A second cardioversion attempt was successful in 10 of 50 patients who experienced AF recurrence. Failure to restore sinus rhythm after 1 or 2 attempts was observed in 29 patients. A total of 46 patients had 1 or 2 electrical cardioversion sessions over the follow-up period. Six deaths (3.0%), 3 of cardiovascular origin, including 1 sudden death and 1 fatal cerebrovascular accident, were reported. Over a mean follow-up of 8±3 months, 3 patients (1.5%) suffered a nonlethal embolic complication including 1 stroke; 8 patients developed heart failure. The outcome of the 194 patients who survived is shown in Figure 4⇑.
Determinants of Clinical Outcome
Analysis of the variables that may have influenced clinical outcome demonstrated that the patients who died were significantly (P<0.0001) older than those who survived (77.6±11.8 versus 68.2±11.3 years). However, using multivariate analysis determined that age, history of embolic event, and hypertension were associated with a higher mortality. Neither age, presence or nature of underlying heart disease, diabetes, hypertension, size of left atrium, type of AF, nor a history of congestive heart failure were associated with a higher risk of embolic complications. Only history of prior embolic event was associated with a higher embolic risk. Seven of the 20 patients who suffered a proven embolic (n=18) or possibly embolic (n=2) complication had a history of prior embolic event and 5 of these 7 were receiving warfarin at the time of recurrent embolic complication.
The results of this study demonstrate, in a French out-of-hospital patient population, a shift toward non-rheumatic AF; in contrast with prior studies, AF has been characterized in a detailed fashion. New data on the relative prevalence of various subsets of AF are provided and significant differences in clinical characteristics between various subsets of AF are outlined on first visit and after 6- to 12-month follow-up. The relative prevalences of paroxysmal, chronic, and recent onset AF were 22.1%, 51.4%, and 26.4%, respectively. Underlying heart disease was present in 534 patients (70.6%). An absence of underlying heart disease referred to as lone AF21 22 23 was observed in 29.4% of the total patient population despite the wide use of echocardiography. Hypertension was the most common underlying condition (39.4%), and hypertensive heart disease was found in 21.4% of the total patient population. Of interest is the fact that the incidence of coronary artery disease (16.6%) and myocardial diseases (15.3%) have reached or exceeded the incidence of valvular heart disease in the total cohort (15.2%). Important differences in the clinical characteristics of patients were found in various clinical subsets of AF.
Comparison With Previous Studies
This prospective study was designed to provide information on an out-of-hospital population of patients with ECG-documented AF. A review of the literature shows that there is only limited information regarding the clinical characteristics of AF. Most reports are older than 20 years and/or deal with hospitalized patients.8 9 10 11 12 13 14 15 A major advantage of the Framingham study2 3 22 24 was that it provided data on the incidence of AF, free of selection bias. However, the patient population in the present report represents the largest AF database available. In the reports of the Framingham study, paroxysmal AF was combined with chronic AF.2 3 22 Our report differs also from that of Godtfredsen,12 who retrospectively analyzed the charts of 1212 patients with AF hospitalized in Slagelse, Denmark, during the years 1940 to 1967. Chronic AF cases represented 65% of his patients. In the interesting study of Kulbertus et al,13 patients with known cardiovascular disease were advised not to participate and AF was found in 3% of subjects >75. The present study represents the first attempt to prospectively characterize in detail the different clinical subsets of AF.
Advantages of Using a Classification System
The classification system used in this study took into account the 2 well known forms of AF, paroxysmal and chronic. A recent onset group allowed classification of patients with first symptomatic and persistent episode of AF of >7 days or recently discovered (asymptomatic) AF. The history of previous AF attacks allows classification of patients with paroxysmal AF, provided through the use of an arbitrary time frame. We realize that there may be an overlap between the recent onset group and the first episode of paroxysmal AF. This emphasizes the necessity to use a time frame to separate various forms of AF. Despite this limitation, the defined subsets of AF allow comparison with other studies. For example, in this study, >50% of patients have chronic AF at the time of the study. This is in contrast to the preliminary results of the Canadian Registry on AF7 which found paroxysmal AF in 64.9% of patients. A possible explanation for the lower percentage of paroxysmal AF in our study is that this group may seek medical attention at a hospital. However, this is unlikely as these patients are referred back to the cardiologist and a large cohort of patients, as in our study, would not miss these patients.
Differences in Various Subsets of AF
Patients with chronic AF were significantly older and less symptomatic than those with paroxysmal AF. A significantly higher percentage of valvular heart disease, dilated cardiomyopathy, and congestive heart failure was found in this group. Another major finding in our study was that the paroxysmal AF group had a significantly lower percentage (53.9%) of detectable heart disease than other groups. As the patients get older and may evolve to chronic AF, there are likely to develop organic heart diseases associated but not necessarily related to AF.
Limitations of the Study
The ALFA study was designed to define the clinical characteristics and outcomes of patients with AF but not the incidence of AF in a French patient population. The second limitation is that this study, as others, underestimates the relative frequency of asymptomatic AF. Ambulatory 24-hour ECG recording was not systematically part of the work-up, and the frequency of asympomatic paroxysmal AF is virtually impossible to determine despite the use of ambulatory monitoring or event recorders.25
This study concerns an out-of-hospital population of patients and the results are relevant to the cardiological population in general rather than to a hospital-based population. Because most of the patients were receiving antiarrhythmic therapy, this study provides neither the clinical presentation in an untreated AF population nor the natural history in AF patients. Further studies are needed to characterize AF and outcome of different subsets of AF seen in other countries or in other AF populations.
In conclusion, this large scale study establishes the current demographic profile of outpatients with AF and highlights some of the changes in demographics that have occurred in the past decades. In contrast to prior studies, AF has been characterized in a detailed fashion, providing new data on different subsets of AF. In addition, the outcomes of patients treated by community-based physicians have been characterized, providing a cross-sectional view of contemporary clinical practice.
A. Aidibe (Toulon), K. Allouache (Belley), R. Amoretti (Cergy), L. Amsallem (Saint Laurent du Var), S. Assouline (Cannes), P. Aubry (Pontchateau), C. Avierinos (Marseille), J.C. Avinee (Cagny), B. Ayrault (La Rochelle), R. Bahsoun (Toulouse), C. Baissus (Béziers), F. Baylac-Domengetroy (Cognac), C. Beaudot-Renault (Cosne sur Loire), J.Y. Beigbeder (Pau), D. Benchimol (Saintes), E. Benitah (Paris), J.P. Bernis (Bayonne), P. Berteau (Bois Guillaume), J.P. Bertho (Hasparren), N. Berthonnaud (Tarbes), F.A. Besnainou (Paris), C. Besson (Alençon), F. Bezot (Lyon), F. Biard (Quimper), D. Binet (Valognes), A. Boisson (Lyon), F. Bolte (Montauban), M. Bonavita (Bastia), D. Bontemps (Vannes), P. Bossan (Nice), M. Bougis (Dinan), Y. Bourhis (Nantes), J.P. Bousquet (Marseille), G. Boutonnet (Arnouville les Gonesse), L. Brottier (Biarritz), P. Buisson (Lorient), C. Burggraf (Montbéliard), G. Buzzi (Paris), G. Calvayrac (Clermont l’Hérault), P. Castelnovo (Châlon sur Saône), I. Centa (Le Kremlin Bicêtre), J. Chartron (Belleville sur Saône), B. Chatelain (Tours), R. Chauvel (Montfermeil), M. Chazalette (Bourges), F. Chombard (Blois), S. Cohen (Marseille), J.M. Colin (Périgueux), J.P. Collet (Barbezieux), B. Colombel (Laval), J. Colonna (Port de Bouc), J.P. Couffin (Meulan), G. Couprie (Jonzac), Y. Cuisinier (Limoges), B. Cunin (Gray), R. D’Hotel (Remiremont), H. Dagher (Saint Martin des Champs), J.M. Dahan (Cagnes sur Mer), B. Dassonville (Paris), J. DeLaguerenne (Nantes), G. Decourtye (Saint Leu la Forêt), J.C. Delelis (Lambersart), P. Delelis (Lambersart), P. Dematteo (Valognes), L. Denis (Saint Chamond), J.P. Denizeau (Versailles), T. Denolle (Dinard), F. Desnot (Sceaux), S. Destrac (Pamiers), J. Devys (Montreuil), J.M. Donnadieu (Marseille), B. Doucet (Chambéry), M. Dours (Aubagne), D. Doyen (Andrezieux), M. Dubigeon (Nantes), D. Dubois (Sartrouville), M. Ducoudre (Biarritz), J.M. Duford (Paris), C. Dumars (Beauvais), B. Dupont (Rennes), Y. Eeckhout (Liévin), J.M. Egal (Lyon), J.L. Eyraud (Montmagny), J.C. Eyraud (Istres), S. Faivre D’Acier (Auxerre), J. Farese (Salon de Provence), J. Faure (Aubagne), J.M. Feige (Orange), R. Fouquet (Le Kremlin Bicêtre), M. Fourdilis (Paimpol), J.Y. Fraboulet (Angers), B. Furtin (Saint Fons), D. Garnier (Fontenay le Comte), S. Gaucher (Bagneux), B. Gaudeau (Poitiers), M. Gaudy (Marselle), C. Giraudet (Cholet), J.P. Glaziou (Toulouse), A. Goguey (Dijon), L. Goullard (Wattignies), B. Guenoun (Bordeaux), P. Guenoun (Thionville), F. Guez (Paris), L. Gutton (Meze), B. Hanssens (Amiens), G. Haquin (Juvisy sur Orge), G. Helfre (Saint Etienne), B. Hijazi (Nimes), M.P. Houppe Nousse (Thionville), J.P. Huberman (Houilles), J.P. Hubmann (Malakoff), J.R. Israel (Saint Michel sur Orge), M. Jeune (Autun), J.P. Jourdes (Bron), R. Jubert (Joue les Tours), G. Jullien (Marseille), J. Kerjean (Angers), S. Kerner (Paris), J.P. Labarre (Montauban), G. Lafitte (Pontivy), M. Lagu (Langon), R. Landel (Saint Brieuc), J.M. Laurent (Tourcoing), H. Le Naour (Lesneven), J. Legoux (Dijojn), J.F. Lemaignen (Rueil Malmaison), J.C. Lenoir (Libourne), C. Levy (Aix en Provence), J.P. Maignien (Besançon), P. Maliar (Valenciennes), B. Manne (Le Blanc Mesnil), G. Marthan (Bordeaux), M. Martin (Paray le Monial), J.F. Masson (Albi), D. Mellin (Wattignies), J.P. Michaux (Antibes), J.M. Michel (Strasbourg), Y. Moreau (Chateauroux), M. Nasr (Blanqueford), G. Nava (Pierrelatte), G. Neel (Tours), C. Nell (Amboise), Y. Nolasco (Thorigny sur Marne), A. Nyiri (Bondy), S. Ollitrault (Carentan), D. Ossena (Saint Dizier), L. Ouazana (Paris), J.M. Pages (Montbrison), J. Paulet (Rennes), J. Perez (Villefranche de Rouergue), J.P. Petitalot (Montluçon), R. Peyretou (Cenon), J.C. Picard (Aubenas), B. Pieri (Pertuis), B. Pincon (Decines), A. Pleskof (Chelles), E. Portier (Rouen), C. Portier (Dijon), A. Quentel (Brest), D. Rabot (Angers), M. Raphael (Aubagne), J.P. Reboud (Grenoble), M. Rivier (Saint Etienne), P.M. Robin (Saint Ouentin), F. Roche (Tassin la Demi Lune), C. Roos (Strasbourg), A. Rosey (Saint-Quentin), J.F. Rouge (Pezenas), D. Rousson (Villeurbanne), M. Rouvier (Manosque), P. Rouxel (Rennes), M. Sagot (Boulogne sur Mer), R. Sarda (Carcassonne), J. Sauvaget (Rueil Malmaison), J. Savelli (Saint Raphaël), X. Schaan (Paris), C. Schurtz (Arras), G. Semain (Valenciennes), O. Serrand (Orléans), G. Simon (Paris), J.A. Simon (Clichy), I. Skolka (Thonon), D. Staub (Strasbourg), F. Stecken (Fleury), M. Tailland (Nîmes), A. Terrier de la Chaise (Nancy), B. Thiel (Remiremont), P. Thomas (Sens), E. Tison (Lille), M. Tissot (Pontarlier), P. Touboul (Toulouse), B. Tournoux (Poitiers), Y. Treguer (La Baule), J. Tricoire (Toulouse), M. Vergne (Saint Denis), H. Vergnoux (Bergerac), P. Veron (Epinay), J.J. Viel (Joigny), T. Vo Duy (Noyon), M. Voglimacci (Rouen), J.M. Walch (Aulnay sous Bois), J.M. Weber (Evreux), M. Weingrod (Plaisir), H. Wojewodzki (Montpellier), F. Zettelmaier (Toulon), C. Ziccarelli (Orléans).
This work was supported in part by grants from Knoll France and the Ministry of Health of France (UF 1638).
- Received July 30, 1998.
- Revision received March 16, 1999.
- Accepted March 26, 1999.
- Copyright © 1999 by American Heart Association
Cabin HS, Clubb KS, Hall C, Perlmutter RA, Feinstein AR. Risk for systemic embolization of atrial fibrillation without mitral stenosis. Am J Cardiol. 1990;61:714–717.
Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983–988.
Petersen P, Godtfredsen J. Embolic complications in paroxysmal atrial fibrillation. Stroke. 1986;17:622–626.
Bialy D, Lehmann MH, Schumacher DN, Steinman RT, Meissmer MD. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation. J Am Coll Cardiol. 1992;19(suppl3):41A. Abstract.
Takahashi N. Clinical features of proxysmal atrial fibrillation. An observation of 94 patients. Stroke. 1986;17:622–626.
Kerr C, Boone J, Connolly S, Greene M, Klein G, Sheldon R, Talajic M. Follow-up of atrial fibrillation: the initial experience of the Canadian Registry of Atrial Fibrillation. Eur Heart J. 1996;17(suppl C):48–51.
Radford MD, Evans DW. Long-term results of DC reversion of atrial fibrillation. Br Heart J. 1968;30:91–96.
Resnekov L, McDonald L. Complications in 220 patients with cardiac dysrhythmias treated by phased direct current shock, and indications for electroconversion. Br Heart J. 1967;29:926–936.
Boissel JP, Wolf E, Gilet J, Soubrane A, Cavallaro A, Mazoyer G, Delahaye JP. Controlled trial of a long acting quinidine for maintenance of sinus rhythm after conversion of sustained atrial fibrillation. Eur Heart J. 1981;2:49–55.
Godtfresen J. Atrial fibrillation: course and prognosis. A follow-up study of 1212 cases. In: Kulbertus HE, Olsson SB, Schlepper M, eds. Atrial Fibrillation. Mölndal, Sweden: AB Hässle; 1982:134–145.
Kulbertus HE, Leval-Rutten F, Bartsh P, Petit JM. Atrial fibrillation in elderly ambulatory patients. In: Kulbertus HE, Olsson SB, Schlepper M, eds. Atrial Fibrillation. Mölndal, Sweden: AB Hässle; 1982:148–154.
Delahaye JP, Milon H, Boissonat P. La fibrillation auriculaire: quelques problémes pratiques actuels. Ann Cardiol Angeiol. 1986;35:597–606.
Bellet S. Atrial fibrillation (clinical manifestations) In: Clinical Disorders of the Heartbeat. Philadelphia, Pa: Lea and Febiger; 1971:223–233.
Devereux RB, Reichek N. Echocardiographic determination of the left ventricular mass. Anatomic variation of the method. Circulation. 1977;55:613–618.
Julian DG, Camm AJ, Frangin G, Janse MJ, Munoz A, Schwartz PJ, Simon P, for the European Myocardial Infarct Amiodarone Trial Investigators. Randomized trial of effect of amiodarone on mortality in patients with left ventricular dysfunction after a recent myocardial infarction. Lancet. 1997;349:667–674.
Royston JP. An extension of Shapiro and Wilk’s test for normality to large samples. Appl Statistics. 1982;31:115–124.
Kruskal WH, Wallis WA. Use of ranks in one-criterion variance analysis. J Am Statistics Assoc. 1952;47:583–621.
Evans W, Swann P. Lone auricular fibrillation. Br Heart J. 1954;16:189–194.
Page RL, Wilkinson WE, Clair WK, McArthy EA, Pritchett EC. Asymptomatic arrhythmias in patients with symptomatic paroxysmal atrial fibrillation and paroxysmal supraventricular tachycardia. Circulation. 1994;89:224–227.This prospective study which included 756 patients with AF, aimed to characterize underlying conditions and different subsets of patients in general practice in France. The relative prevalences of paroxysmal, chronic, and recent onset AF were 22.1%, 51.4%, and 26.4%, respectively. Hypertension (39.4%), coronary artery disease (16.6%), and myocardial diseases (15.3%) were the most common disorders. Detectable heart disease was less common in the paroxysmal AF (53.9%) group. The outcome of this treated population was evaluated (mean follow-up=8.6±3.7 months). This study highlights a shift in cardiac causes toward nonrheumatic AF and significant differences between various subsets of AF.