Electrical cardioversion for persistent atrial fibrillation or atrial flutter in clinical practice: predictors of long-term outcome

Electrical cardioversion for persistent atrial fibrillationor atrial flutter in clinical practice: predictors oflong-term outcome

G. Boriani, I. Diemberger, M. Biffi, G. Domenichini, C. Martignani, C. Valzania, A. Branzi

Introduction

Atrial fibrillation (AF) is the most common

arrhythmia and the economic impact of its manage-

ment is impressive (1). Electrical cardioversion

(which was introduced in clinical practice in 1963)

(2) aims to (i) restore the atrial contribution to

ventricular filling and output, (ii) regularise ventric-

ular rate and (iii) interrupt atrial remodelling. The

results of the Atrial Fibrillation Follow-up Investiga-

tion of Rhythm Management (AFFIRM; 3) and Rate

Control versus Electrical Cardioversion for Persist-

ent Atrial Fibrillation (RACE; 4) trials have tended

to shift attention from rhythm control to (ventric-

ular) rate control. Nevertheless, choice of the best

treatment strategy in individual patients remains a

subject of debate (5). Conversion of AF to sinus

rhythm can itself be beneficial, and therefore

remains an important therapeutic option (5), espe-

cially if new drugs will help in sinus rhythm main-

tenance (6). In this view predictors of maintenance

of sinus rhythm (especially in the long term) are

required to refine indications to electrical cardiover-

sion (for better patient selection). Among the many

proposed predictors of recurrence of AF/atrial flut-

ter (AFL) after external/internal electrical cardiover-

sion, those generally suggested (6) are AF/AFL

duration (7,8), left atrial dimension at echocardiog-

raphy (9) and presence of rheumatic heart disease

(10). However, application of previous knowledge

in current clinical practice should consider the

changes occurred in the last 15–20 years in the type

SUMMARY

Despite the results of Atrial Fibrillation Follow-up Investigation of Rhythm Manage-

ment and Rate Control versus Electrical Cardioversion for Persistent Atrial Fibrilla-

tion trials, which favour a general shift in atrial fibrillation (AF) therapeutic

approach towards control of ventricular rate, a strategy based on restoration of

sinus rhythm could still play a role in selected patients at lower risk of AF recur-

rence. We explored possible predictors of relapses after external electrical cardio-

version among patients with persistent AF or atrial flutter (AFL). We analysed the

clinical characteristics and conventional echocardiographic parameters of patients

with persistent AF/AFL enrolled in an institutional electrical cardioversion pro-

gramme. Among 242 patients (AF/AFL, 195/47; mean age 62 ± 13 years), sinus

rhythm was restored in 215 (89%) and maintained in 73 (34%) at a follow-up of

930 days (median). No baseline clinical/echocardiographic variables predicted acute

efficacy of cardioversion at logistic regression analysis. However, two variables pre-

dicted long-term AF/AFL recurrence among patients with successful cardioversion

at multivariate Cox’s proportional hazards analysis: (i) duration of arrhythmia

‡ 1 year (HR, 2.07; 95% CI, 1.29–3.33) and (ii) presence of previous cardioversion

(HR, 1.67; 95% CI, 1.17–2.38). These variables also presented high-positive pre-

dictive values (72% and 80% respectively).Whereas the high acute efficacy of elec-

trical cardioversion (approximately 90%) does not appear to be predictable, two

simple clinical variables could help identify patients at higher risk of long-term AF/

AFL recurrence after successful electrical cardioversion. We think there could be a

case for initially attempting external electrical cardioversion to patients who have

had AF/AFL for < 1 year. In such patients, the chance of long-term success

appears to be relatively high.

What’s known?• Recent trials have shown that in selected atrial

fibrillation (AF) patients rate control provides

benefits not inferior to those of rhythm control in

terms of quality of life and long-term survival.

However, AF patients requiring care in daily

practice seems to be much more heterogeneous

than the populations enrolled in these trials.

Rhythm control strategy could be more suitable

in subjects who are probably to maintain sinus

rhythm at long-term.

What’s new?• We propose two promising simple clinical

variables which could help in identifying patients

at higher risk of long-term atrial fibrillation/atrial

flutter relapses.

Institute of Cardiology,

University of Bologna, Azienda

Ospedaliera S. Orsola-Malpighi,

Bologna, Italy

Correspondence to:

Prof. Giuseppe Boriani,

Institute of Cardiology,

University of Bologna,

Policlinico S. Orsola-Malpighi,

Via Massarenti no. 9, 40138

Bologna, Italy

Tel.: + 39 051349858

Fax: + 39 051344859

Email:

[email protected]

Disclosures

None of the authors has any

conflict of interest related to

this manuscript, which was

partially supported by a grant

from ‘‘Luisa Fanti Melloni’’

Foundation, Bologna.

doi: 10.1111/j.1742-1241.2007.01298.x

OR IG INAL PAPER

ª 2007 The Authors748 Journal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756

and class of antiarrhythmic agents prescribed for

preventing AF recurrences (11,12).

We explored possible factors associated with a

higher percentage of long-term maintenance of sinus

rhythm after external electrical cardioversion among

patients who underwent elective electrical cardiover-

sion in the context of our institutional programme.

Methods

Selection criteriaBetween August 1997 and October 2002, all patients

who presented in our Cardiology Institute with the

persistent form of AF/AFL (as defined by a current

classification system) (13) were considered for our

institutional elective DC cardioversion programme.

Specific selection criteria for entry in the programme

were: (i) onset of AF/AFL documented either by

ECG recordings or by an abrupt onset of palpitations

with subsequent ECG evidence of AF/AFL and (ii) a

history of AF/AFL with persistence of arrhythmia for

at least 48 h (as most spontaneous conversions tend

to occur within the first 48 h) (14,15). According to

the programme, patients already receiving (at enrol-

ment) adequate anticoagulant treatment according to

current evidence (16,17), were directly submitted to

electrical cardioversion. The remaining patients were

started on oral warfarin: all those who after at least

4 weeks’ treatment achieved a stable international

normalised ratio (INR) within the therapeutic range

(2.0–3.0) and had a mean ventricular rate of at least

45 beats/min were then submitted to electrical cardi-

oversion (after a washout period of at least five half-

lives of all drugs used to control ventricular response

– verapamil, digoxin, beta blockers, diltiazem, etc.).

All enrolled patients provided prior informed con-

sent for participation in the electrical cardioversion

programme (and anonymous scientific data publica-

tion), which was carried out in accordance with the

guidelines of the current version of the declaration

of Helsinki.

All patients who received electrical cardioversion

by October 2002 were screened for the present analy-

sis. The following exclusion criteria were applied

a priori: (i) age > 80 years; (ii) heart failure, New

York Heart Association (NYHA) class IV; (iii) left

ventricular ejection fraction < 25%; (iv) recent myo-

cardial infarction (< 6 months before cardioversion);

(v) ECG evidence (past or present) of ventricular

pre-excitation; (vi) history of second or third degree

atrioventricular block or bifascicular block; (vii)

known sick sinus syndrome; (viii) hypokalaemia

(potassium values < 3.5 mEq/L) and (ix) severe kid-

ney/liver failure or severe hypoxia (partial pressure of

oxygen < 55 mmHg).

Atrial cardioversionThe institutional procedure for external electrical car-

dioversion, was based on well-established methods

(18). During continuous 12-lead ECG monitoring and

heavy sedation with propofol (2 mg/kg) under the

supervision of an anaesthetist, up to three R wave-syn-

chronised monophasic/biphasic external shocks were

delivered to restore sinus rhythm. Two paddles were

used, one placed on the patient’s anterior chest wall,

the other in axillary position. After an unsuccessful

attempt, at least 1 min was allowed to elapse before

the next shock. Cardioversion was considered success-

ful if sinus rhythm was restored for at least three beats.

Any resumption of AF within 24 h was considered an

early recurrence. Cardiac rhythm was monitored by

telemetry for at least 12 h after electrical cardioversion

and a 12-lead ECG was performed at 4 and 24 h.

Patients were discharged at 30–36 h.

Relapse prophylaxisOur institutional protocol required that patients with

clinical indications for amiodarone for prevention of

relapse (i.e. coexisting heart failure or left ventricular

dysfunction) (19) or considered at high risk of

relapse [i.e. with long-lasting AF/AFL and previous

relapse(s)] started antiarrhythmic prophylaxis with

oral amiodarone in the weeks preceding electrical

cardioversion (leading dose of 600 mg/day for 3 days

and 400 mg/day for at least 4 weeks) (6). Otherwise,

prophylaxis was routinely initiated after cardiover-

sion according to the following treatment criteria:

severe left ventricular dysfunction (amiodarone, with

a loading dose administered intravenously); no cor-

onary artery disease or ventricular dysfunction (fle-

cainide or propafenone) (6). Patients judged at low

risk of relapse (based on age, previous AF/AFL epi-

sodes, left atrial diameter, etc.) (20) did not receive

prophylaxis. Sotalol was considered a second-line

option (but was not actually used in any of the

patients under study).

Follow-upTo identify both symptomatic and asymptomatic

recurrences, the institutional elective DC cardiover-

sion programme required that patients returned for

ECG at the first symptom of recurrence of AF/AFL,

and in any case at 1, 3 and 6 months after cardio-

version. After 6 months, patients were followed

periodically (based on clinical profile and ongoing

therapy) to confirm that new recurrences had not

occurred.

Data collectionAll data were systematically collected in the prospec-

tively assembled database dedicated to the institutional

Outcome predictors of AF ⁄ AFL cardioversion 749

ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756

electrical cardioversion programme [anamnesis data

– including previous ECG-documented episodes of

AF/AFL – were obtained on presentation; baseline

echocardiographic values, ventricular rate (from

12-lead ECG) and body surface area on admission to

the ward]. The only data systematically collected on

rheumatic heart disease regarded mitral valve disease.

Statistical analysisContinuous variables were compared using Student’s

t-test for paired data. Comparisons were performed

using the chi-squared or Fisher’s exact test (as appro-

priate) for binary variables, and the Mann–Whitney

test for ordinal variables. Multivariate analysis was

performed using forward stepwise logistic regression

to determine independent predictors of effective DC

cardioversion; variables with p values < 0.20 at uni-

variate analysis were entered. Kaplan–Meier relapse-

free survival plots were prepared (relapse was defined

as ECG-documented recurrence of AF/AFL). Cox

proportional hazards regression analysis was per-

formed to determine characteristics that were related

to the outcome; covariates with p values < 0.20 in the

log-rank test were entered in the multivariate model.

Hazards risks (HR) are reported with 95% confidence

intervals (95% CI). Statistical calculations and survival

plots were prepared with SPSS version 11.0.5 (SPSS Inc,

Chicago, IL, USA), with significance set at p < 0.05.

Mean values are reported as ± SD.

Results

Patient populationTwo hundred and forty-two patients were eligible for

the present analysis. Clinically, there were 56 (23%)

patients in NYHA class I, 138 (57%) in class II and

48 (20%) in class III. Numbers of previous episodes

of AF were as follows: zero in 140 (58%) patients,

one in 58 (24%), two in 24 (10%) and three or more

in 20 (8%). Numbers of previous attempts at cardio-

version were zero in 186 (77%) patients, one in 41

(17%), two in 12 (5%) and three in three (1%).

Table 1 reports baseline clinical and echocardio-

graphic characteristics of the study population

according to the type of arrhythmia (AF or AFL).

Patients with AFL presented higher mean values of

left atrial diameter and ventricular rate (the latter

despite more frequent amiodarone pretreatment and

no significant difference in use of beta-blockers).

Moreover, AFL patients had a shorter period between

onset of arrhythmia and cardioversion, probably

attributable to more disturbing symptoms caused by

higher ventricular rates during AFL [median 90 days

(25–75th percentiles, 10–120) vs. 120 days (25–75th

percentiles, 60–240 days); p ¼ 0.001 at the Mann–

Whitney test]. No significant difference was noted at

the Mann–Whitney test for either NYHA class, num-

ber of previous arrhythmic episodes or number of

previous attempted cardioversions.

Conversion to sinus rhythm and antiarrhythmicdrugsExternal DC cardioversion restored sinus rhythm in

215/242 (89%) patients. No difference in conversion

rates was observable between the AF and AFL sub-

groups (172/195, 88% vs. 43/47, 91%; p ¼ 0.616).

Univariate analysis of all the parameters listed in

Table 1 and use of antiarrhythmic drugs was per-

formed for patients with successful and unsuccessful

cardioversion. Unsurprisingly, those patients who did

not achieve conversion to sinus rhythm (n ¼ 27)

seemed to have higher baseline body weight (82 ±

14 kg vs. 76 ± 16 kg; p ¼ 0.05). They also more

often initially presented an enlarged (‡ 50 mm) left

atrium (19/27, 70% vs. 104/215, 48%; p ¼ 0.03) and

moderate/severe reduction (< 40%) in left ventricu-

lar ejection fraction (8/27, 30% vs. 29/215, 13%; p ¼0.03). Intriguingly, none of the unsuccessful patients

had undergone any previous cardioversion attempts

(0/27, 0% vs. 56/215, 26% in the successful group;

p ¼ 0.05 at Fisher’s exact test). None of the other

variables approached significance (all p > 0.10).

None of the variables analysed turned out to be sig-

nificant predictors of effective cardioversion at logis-

tic regression analysis.

Figure 1 reports usage – according to clinicians’

decisions – of different prophylactic and/or antiar-

rhythmic drugs before and after cardioversion. No

difference was observable between patients with suc-

cessful/unsuccessful cardioversion in terms of drugs

administered before the procedure. Unsurprisingly,

patients who received amiodarone pretreatment pre-

sented a higher mean left atrial diameter

(51 ± 7 mm vs. 49 ± 7 mm, p ¼ 0.02), more often

had longer [‡ 120 days (median value)] AF duration

(71% vs. 65%, p ¼ 0.02), frequently had at least one

previous AF/AFL episode (50% vs. 37%, p ¼ 0.05),

and showed trends towards an almost twofold higher

mean number of previous cardioversion attempts

(32% vs. 17%, p ¼ 0.06), higher age (63 ± 11 years

vs. 61 ± 13 years, p ¼ 0.09) and higher left ventric-

ular end-diastolic diameter at the echocardiography

(53 ± 10 mm vs. 51 ± 8 mm, p ¼ 0.09).

Recurrence of AF/AFL after conversion to sinusrhythm and related factorsRecurrence-free survival in the entire study popula-

tion (n ¼ 242) is shown in Figure 2. Univariate ana-

lysis was performed for the parameters listed in

Table 1, as well as for the use of antiarrhythmic

750 Outcome predictors of AF ⁄ AFL cardioversion


drugs (altogether, and amiodarone considered sepa-

rately). The only variable that reached p < 0.20 was

absence of previous cardioversion, which appeared

more frequent among patients who converted and

maintained sinus rhythm during follow-up (62/73,

85% vs. 124/169, 73%; p ¼ 0.05). However, pre-

sence/absence of previous cardioversion did not turn

Table 1 Baseline clinical and echocardiographic characteristics of the overall study population, and of the atrial

fibrillation (AF) and atrial flutter (AFL) subgroups

Study sample (n ¼ 242) AF (n ¼ 195) AFL (n ¼ 47) p value (AF vs. AFL)

Mean age (years) 62 ± 13 62 ± 12 61 ± 13 0.5

Men 64% 64% 64% 0.98

AF/AFL duration (median, days) 120 (45–210) 120 (60–240) 90 (10–120) 0.001

AF/AFL duration ‡ 120 days 55% 59% 36% 0.04

AF/AFL duration ‡ 365 days 14% 17% 4% 0.03

Mean heart rate (bpm) 85 ± 23 83 ± 21 95 ± 29 0.001

Mean weight (kg) 76 ± 15 77 ± 15 73 ± 17 0.07

NYHA Class I 23% 24% 19% 0.5

Mitral valve disease (any) 51% 52% 49% 0.7

Coronary artery disease 14% 13% 15% 0.7

Hypertensive CM 17% 18% 11% 0.2

Dilated CM 14% 13% 15% 0.7

Hypertrophic CM 5% 5% 6% 0.7

Restrictive CM 2% 1% 4% 0.1

Congenital heart disease 3% 3% 4% 0.7

Isolated hypertension* 21% 22% 19% 0.7

Hyperthyroidism 2% 3% 0% 0.6

Lone AF 9% 10% 6% 0.7

No previous AF/AFL episodes 58% 60% 49% 0.2

Previous CV 23% 21% 32% 0.2

Amiodarone pre-treatment (‡ 1 month) 40% 36% 55% 0.02

Mean left atrium Ø (mm) 49 ± 7 49 ± 7 51 ± 7 0.02

Left atrium Ø ‡ 50 mm 51% 48% 62% 0.1

Mean LVEF (%) 56 ± 15 57 ± 14 54 ± 15 0.3

LVEF £ 40% 15% 15% 17% 0.9

Mean LVEDD (mm) 52 ± 9 52 ± 9 51 ± 9 0.4

Mean LVESD (mm) 37 ± 10 36 ± 10 37 ± 10 0.6

*Isolated hypertension was defined as blood pressure > 140/90 mmHg in the absence of echocardiographic signs of hypertensive

cardiomyopathy. AF, atrial fibrillation; AFL, atrial flutter; CM, cardiomyopathy; CV, electrical cardioversion; LVEDD, left ventricular end

diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end systolic diameter; NYHA, New York Heart

Association; Ø, diameter. Data were number (%), median (interquartile range: 25th and 75th percentiles) or mean ± SD.

Figure 1 Diagram of enrolled patients and procedure

results, considering antiarrhythmic drugs during each step.

AF, atrial fibrillation; AFL, atrial flutter; CV, electrical

cardioversion; SR, sinus rhythm

Figure 2 Kaplan–Meier curve for relapse-free survival in

the entire population. AF, atrial fibrillation; AFL, atrial

flutter



out to be an independent predictor at Cox regression

analysis.

Of the 215 patients who converted to sinus

rhythm, 142 (66%) presented recurrence of AF/AFL

during the follow-up period (median 930 days;

25–75th percentiles 510–1290). Of the parameters

that entered the multivariate Cox’s proportional haz-

ards regression model (male gender, previous

AF/AFL episodes, previous cardioversions, AF/AFL

duration ‡ 1 year, mean left ventricular end-diastolic

diameter, mean left ventricular end-systolic diameter,

left ventricular ejection fraction £ 40%; of note, nei-

ther antiarrhythmic drugs nor amiodarone reached

p < 0.20), the only significant independent long-term

predictors of recurrence were AF duration ‡1 year

(HR, 2.07; 95% CI, 1.29–3.33) and previous cardio-

versions (HR, 1.67; 95% CI, 1.17–2.38). Moreover

these binary variables also presented high specificity

and positive predictive values (Table 2).

Of note, exclusion of patients with AFL did not

substantially affect these findings (AF duration

‡ 1 year, HR 1.82, 95% CI 1.10–3.01; previous cardi-

oversions, HR 1.59, 95% CI 1.06–2.40). Figure 3A, B

depicts relapse-free Kaplan–Meier curves for patients

who responded to cardioversion, according to AF

duration ‡ 1 year and previous cardioversion. Of

note, the differences in the survival curves appear to

attenuate after 2 or 3 years (as electrical cardiover-

sion is not intended to be a curative treatment, pre-

sumably almost all patients will eventually relapse).

Among patients without previous cardioversion, the

median duration of sinus rhythm maintenance was

780 days (interquartile range, 120–1710 days), when

compared with 300 days (interquartile range,

30–1140 days) in the rest of the group. Among

patients with <1 year of AF, the median duration of

sinus rhythm maintenance was 720 days (interquar-

tile range, 120–1440 days), when compared with

150 days (interquartile range, 20–720 days) in the

other patients.

In view of the guidelines for AF management

recently released by NICE (21), we have re-rolled the

analysis of our data considering the criteria proposed

by NICE, to assess the potential predictive value of

these parameters to identify patients at higher risk of

AF/AFL relapses. None of the evaluated criteria (age

> 65 years, presence of coronary artery disease, left

atrial diameter > 5.5 cm, AF duration > 12 months,

history of multiple previous cardioversions, NYHA

class > II, first lone AF) was significantly associated

with AF/AFL recurrence at the univariate analysis

performed on our population of 242 patients. Analy-

sing the subgroup of 215 patients with restoration of

sinus rhythm after electrical cardioversion, three vari-

ables resulted significant predictors of AF relapse

both at univariate and multivariate Cox regression

analysis: age > 65 years (HR, 0.706; 95% CI, 0.500–

0.997), AF duration > 12 months (HR, 2.10; 95%

CI, 1.20–3.66), history of multiple previous cardio-

versions (HR, 2.15; 95% CI, 1.18–3.92). Of note our

Table 2 Specificity, sensitivity, positive and negative predictive values of the two independent long-term predictors of AF/AFL recurrence after

electrical cardioversion at the multivariate analysis

Specificity (%) Sensitivity (%) Positive predictive value (%) Negative predictive value (%)

AF/AFL duration ‡ 1 year 89 15 72 35

Previous cardioversion 85 32 80 39

AF/AFL duration ‡ 1 year and previous cardioversion 100 5 100 35

AF/AFL duration ‡ 1 year and/or previous cardioversion 74 42 76 39

(A)

(B)

Figure 3 Kaplan–Meier curves for relapse-free survival after

effective cardioversion considering (A) presence/absence of

previous cardioversion; (B) duration (<1 year vs. ‡1 year)

of previous arrhythmia. AF, atrial fibrillation; AFL, atrial

flutter; CV, electrical cardioversion



choice of a more conservative criterion, based on

absence of any previous cardioversion (as an alter-

native to multiple previous cardioversions),

improved the sensitivity in predicting AF/AFL long-

term relapses, with an almost negligible decrease in

specificity (Tables 2 and 3).

Discussion

The present study explores the question of which

clinical/echocardiographic parameters may be of rele-

vance in the choice between rhythm and rate control

in clinical practice. Our results provide support for

the possibility of identifying factors for recognition

of subgroups of AF/AFL patients who are more

probably to maintain sinus rhythm in the long term

after electrical cardioversion.

Recent randomised trials (AFFIRM, RACE) (3,4)

have shown that in selected patients rate control pro-

vides benefits not inferior to those of rhythm control

in terms of quality of life and long-term survival.

These studies have encouraged a generalised shift in

interest towards rate control, which also appears

more economically attractive for long-term manage-

ment of AF (22). However, the overall population of

AF patients requiring care in daily practice seems to

be much more heterogeneous than the populations

enrolled in clinical trials. Therefore, the choice

between rate and rhythm control still represents a

debated issue and cardioversion continues to be

adopted in selected subgroups of patients on the

basis of individualised decisions (5,23–26). In partic-

ular, the rhythm control strategy could be more suit-

able in subjects who are probably to maintain sinus

rhythm at long-term (27). Of note, an exploratory

post hoc time-dependent analysis of AFFIRM sugges-

ted that sinus rhythm itself is associated with

improved survival (24). These considerations promp-

ted us to perform the present exploratory analysis.

This study highlights the prognostic influence of a

previous attempt at electrical cardioversion. In our

analysis of the subset of patients who converted, this

baseline characteristic was a predictor of long-term

AF/AFL relapse with a high positive predictive value

(of note, the lack of significance for previous AF/AFL

episodes can be explained by the fact that referral to

electrical cardioversion was based on the severity of

the overall clinical picture). This finding suggests that

DC cardioversion might be worth trying in the

absence of a previous cardioversion, even though in

some cases the procedure may fail to restore sinus

rhythm (this consideration may explain why absence

of previous cardioversion did not predict the main

long-term outcome in the entire study population).

Indeed, no baseline predictor of acute success of DC

shock (i.e. restoration of sinus rhythm) was apparent

in the present analysis. DC cardioversion efficacy is

probably determined by a complex interplay between

clinical, structural and technical factors (28,29). This

consideration, along with the high acute efficacy

(approximately 90%) of the strategy, can probably

explain the difficulty in identifying predictors of car-

dioversion. Those parameters that have been pro-

posed elsewhere require confirmation, especially in

view of the higher efficacy provided by biphasic

shocks (30).

Randomised trials of rate vs. rhythm control have

highlighted that long-term maintenance of sinus

rhythm is currently possible only in a minority of

patients (3,4,25). Our patients with no previous elec-

trical cardioversion showed a much better outcome

in terms of sinus rhythm maintenance in comparison

with patients who underwent previous cardioversion

attempts (62% vs. 45% at 1 year; Figure 3A). The

median duration of sinus rhythm maintenance was

over twice as long in patients without previous cardi-

oversions (780 vs. 300 days). These considerations

lead us to hypothesise that AF/AFL patients who

have not previously undergone cardioversion (77%

of our population) might be more probably to main-

tain sinus rhythm in the long term, and in the

absence of specific contraindications could be prom-

ising candidates for electrical cardioversion. This

issue has important implications for daily clinical

Table 3 Specificity, sensitivity, positive and negative predictive values of multiple previous cardioversions, alone or in combination with

arrhythmia duration, in predicting AF/AFL recurrences after electrical cardioversion, according to criteria proposed by NICE (21).

Specificity

(%)

Sensitivity

(%)

Positive predictive

value (%)

Negative predictive

value (%)

Multiple previous cardioversions 96 9 80 35

AF/AFL duration > 1 year and multiple previous cardioversions 100 0* NA* 34

AF/AFL duration > 1 year and/or multiple previous cardioversions 88 19 75 36

*No patients with combined characteristics in our population. NA, not available



practice. Unfortunately, no randomised comparative

trial of electrical cardioversion vs. rate control in the

subset of patients with persistent AF/AFL with no

previous history of cardioversion is currently avail-

able. However, this line of research could be attract-

ive as the subset of persistent AF patients candidate

to a first attempt at cardioversion has so far received

limited attention in the literature. For example,

RACE only included patients with at least two previ-

ous cardioversions (4). With regard to AFFIRM, the

heterogeneity of the population (most of the patients

had self-terminating AF) makes it difficult to derive

specific information on persistent AF, and only 30%

of overall enrolled population patients had no his-

tory of electrical cardioversion.

In our study those patients who had AF or AFL

lasting <1 year prior to electrical cardioversion also

showed better long-term sinus rhythm maintenance

(66% vs. 48% at 1 year), with an almost fivefold

higher median duration of remission (720 vs.

150 days). As this observation is in line with findings

from previous observational studies (20,31), we think

it might be worth also considering the potential rele-

vance of the duration of AF/AFL in selection of can-

didates for electrical cardioversion. The importance

of this characteristic is reinforced by a higher inci-

dence of rhythm control abandonment in patients

with long durations of AF on presentation, as evi-

denced in a post hoc analysis of the AFFIRM trial

(32). There is great expectation that newer antiar-

rhythmic drugs will help in improving sinus rhythm

maintenance at long term (25,33).

As in other studies (34,35), we decided to include

patients with AFL in the analysis. Despite the well-

known efficacy of cavo-tricuspid isthmus ablation in

AFL treatment, atrial cardioversion still plays a role

for patients with the first AFL episode or for subjects

who refuse catheter ablation (or are not selected for

it) (36). Interestingly, our main finding regarding the

possible role of absence of previous cardioversion

and duration of arrhythmia for prediction of recur-

rence was substantially reproduced in the AF sub-

group (i.e. excluding AFL patients).

A recent registry study based at the Mayo Clinic

also focused on (elderly) patients for whom electrical

cardioversion had been effective (35). In terms of

long-term maintenance of sinus rhythm (in a study

population with a median follow-up of 3.5 months),

AFL patients with no previous DC cardioversion

fared better than the other three subgroups, compri-

sing AF patients either with or without previous car-

dioversion and AFL patients who had already

experienced cardioversion (interestingly, in this study

too, use of antiarrhythmic drugs did not appear to

affect recurrence). Although we originally preferred

to explore the same main variables singly, we feel it

might be interesting also to attempt to look at the

variables in combination. The results of this analysis

broadly support the results of the Mayo clinic study.

In the context of our median follow-up of about

30 months, Kaplan–Meier survival analysis showed a

statistical trend towards better long-term mainten-

ance of sinus rhythm in the subgroup of patients

with AFL and no previous cardioversion (when com-

pared with the other three subgroups; data not

shown). Taken together, the results from the Mayo

Clinic study and the present investigation both seem

to support the validity of absence of previous cardio-

versions as a clinically relevant predictor of medium-

to-long term maintenance of sinus rhythm after

electrical cardioversion in AF/AFL patients. We think

that prospective studies might be warranted to

investigate the relative weight of type of arrhythmia

(AFL vs. AF) and of presence/absence of previous

cardioversion as potentially useful markers for pre-

diction of sinus rhythm maintenance.

Finally we think that the results of the present

analysis, in light of AFFIRM and RACE multicenter

trials on rate vs. rhythm control (3,4), highlight the

potential of two simple clinical variables, absence of

previous cardioversions at patient history and

AF/AFL duration shorter than 1 year, as criteria for

selecting candidates to a strategy based on rhythm

control. Our data broadly support current NICE

guidelines regarding management of persistent AF,

but the choice of the more conservative criterion for

preferring a rate control strategy (absence of previous

cardioversion rather than multiple previous cardio-

versions) seems to improve the predictive value of

this variable with regard to the risk of AF/AFL relap-

ses.

Study limitationsThis is an exploratory, retrospective analysis (albeit

of prospectively collected data). Nevertheless, we

think that the clinical relevance of the issues and the

lack of available randomised trials justify this initial

exploratory approach. Furthermore, the relatively

unselected character of the study population (derived

from daily clinical practice) may enhance the interest

of the observations.

The use of various antiarrhythmic medications

might be considered a confounding factor in the

analysis (although this does reflect what occurs in

current clinical practice). However, neither use of

antiarrhythmic drugs nor use of amiodarone (consid-

ered separately) reached p < 0.20 at univariate

analysis (the threshold for multivariate Cox’s propor-

tional hazards analysis). Furthermore, propensity

score analysis for use of antiarrhythmic drugs with



respect to long-term recurrences of AF/AFL did not

affect the results (data not shown).

The ability to detect asymptomatic AF/AFL recur-

rences is dependent on the type and frequency of fol-

low-up visits, and various methods have been

applied in some trials (transtelephonic monitoring,

Holter recordings) (37). However, these methods are

difficult to apply in the long term and do not mirror

current clinical practice. Although most of the

patients received monophasic shocks, the overall

efficacy of cardioversion (89%) was almost as high as

in series where only biphasic shocks were used (6),

and recurrence of AF/AFL is not influenced by the

specific procedure.

It should also be noted that we could gain only

scanty information regarding ‘very long-term’ main-

tenance of sinus rhythm (after about 3 years). How-

ever, a 2- to 3-year period of sinus rhythm

maintenance appears to be a clinically relevant target.

ConclusionsWhereas the high acute efficacy of electrical cardiover-

sion (approximately 90%) does not appear to be pre-

dictable, two simple clinical variables related to

patients’ clinical history – namely, absence of previous

cardioversion and <1 year duration of arrhythmia –

might allow selection of candidates for a rhythm con-

trol strategy who are more likely to maintain sinus

rhythm after successful electrical cardioversion. We

think that there could be a case for initially attempting

external electrical cardioversion to patients who have

had AF/AFL for <1 year. In such patients, the chance

of long-term success appears to be relatively high.

Acknowledgements

We thank Robin M.T. Cooke for writing assistance

and advice (provided under contract with the Uni-

versity of Bologna).

References

1 Waktare JE, Camm AJ. Acute treatment of atrial fibrillation: why

and when to maintain sinus rhythm. Am J Cardiol 1998; 81:

3C–15C.

2 Silverman ME. History and personal observations of electrical car-

dioversion of atrial fibrillation. Am J Cardiol 2004; 94: 751–2.

3 Wyse DG, Waldo AL, DiMarco JP et al. A comparison of rate con-

trol and rhythm control in patients with atrial fibrillation. N Engl J

Med 2002; 347: 1825–33.

4 Van Gelder IC, Hagens VE, Bosker HA et al. A comparison of rate

control and rhythm control in patients with recurrent persistent

atrial fibrillation. N Engl J Med 2002; 347: 1834–40.

5 Zimetbaum P, Josephson ME. Is there a role for maintaining sinus

rhythm in patients with atrial fibrillation? Ann Intern Med 2004;

141: 720–6.

6 Fuster V, Ryden LE, Asinger RW et al. ACC/AHA/ESC guidelines

for the management of patients with atrial fibrillation. 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 Confer-

ences (Committee to develop guidelines for the management of

patients with atrial fibrillation) developed in collaboration with the

North American Society of Pacing and Electrophysiology. Eur

Heart J 2001; 22: 1852–923.

7 Frick M, Frykman V, Jensen-Urstad M et al. Factors predicting

success rate and recurrence of atrial fibrillation after first electrical

cardioversion in patients with persistent atrial fibrillation. Clin

Cardiol 2001; 24: 238–44.

8 Lehto M, Kala R. Persistent atrial fibrillation: a population based

study of patients with their first cardioversion. Int J Cardiol 2003;

92: 145–50.

9 Okcun B, Yigit Z, Kucukoglu MS et al. Predictors for maintenance

of sinus rhythm after cardioversion in patients with nonvalvular

atrial fibrillation. Echocardiography 2002; 19: 351–7.

10 Van Gelder IC, Crijns HJ, Van Gilst WH et al. Prediction of

uneventful cardioversion and maintenance of sinus rhythm from

direct-current electrical cardioversion of chronic atrial fibrillation

and flutter. Am J Cardiol 1991; 68: 41–6.

11 Lundstrom T, Ryden L. Chronic atrial fibrillation. Long-term results

of direct current conversion. Acta Med Scand 1988; 223: 53–9.

12 Nieuwlaat R, Capucci A, Camm A et al. Atrial fibrillation manage-

ment: a prospective survey in ESC Member Countries The Euro

Heart Survey on Atrial Fibrillation. Eur Heart J 2005; 26: 2422–34.

13 Gallagher MM, Camm AJ. Classification of atrial fibrillation.

Pacing Clin Electrophysiol 1997; 20: 1603–5.

14 Boriani G, Capucci A, Lenzi T et al. Propafenone for conversion of

recent-onset atrial fibrillation. A controlled comparison between

oral loading dose and intravenous administration. Chest 1995; 108:

355–8.

15 Capucci A, Boriani G, Rubino I et al. A controlled study on oral

propafenone versus digoxin plus quinidine in converting recent

onset atrial fibrillation to sinus rhythm. Int J Cardiol 1994; 43:

305–13.

16 Gustafsson C, Asplund K, Britton M et al. Cost effectiveness of

primary stroke prevention in atrial fibrillation: Swedish national

perspective. BMJ 1992; 305: 1445–6.

17 Prystowsky EN, Benson DW Jr, Fuster V et al. Management of

patients with atrial fibrillation. A statement for healthcare profes-

sionals. From the Subcommittee on Electrocardiography and Elec-

trophysiology, American Heart Association. Circulation 1996; 93:

1262–77.

18 Ewy GA. Optimal technique for electrical cardioversion of atrial

fibrillation. Circulation 1992; 86: 1645–7.

19 Singh SN, Fletcher RD, Fisher SG et al. Amiodarone in patients

with congestive heart failure and asymptomatic ventricular

arrhythmia. N Engl J Med 1995; 333: 77–82.

20 Van Gelder IC, Crijns HJ, Tieleman RG et al. Chronic atrial fibril-

lation. Success of serial cardioversion therapy and safety of oral

anticoagulation. Arch Intern Med 1996; 156: 2585–92.

21 National Collaborating Centre for Chronic Conditions. Atrial

Fibrillation: National Clinical Guideline for Management in Primary

and Secondary Care. London: Royal College of Physicians, 2006.

22 Boriani G, Biffi M, Diemberger I et al. Rate control in atrial fibril-

lation: choice of treatment and assessment of efficacy. Drugs 2003;

63: 1489–509.

23 Chung MK. Randomized trials of rate vs. rhythm control for atrial

fibrillation. J Interv Card Electrophysiol 2004; 10: 45–53.

24 Corley SD, Epstein AE, DiMarco JP et al. Relationships between

sinus rhythm, treatment, and survival in the Atrial Fibrillation Fol-

low-Up Investigation of Rhythm Management (AFFIRM) Study.

Circulation 2004; 109: 1509–13.

25 Nattel S, Opie LH. Controversies in atrial fibrillation. Lancet 2006;

367: 262–72.

26 Vikman S, TH M, Yli-Mayry S et al. Heart rate variability and

recurrence of atrial fibrillation after electrical cardioversion. Ann

Med 2003; 35: 36–42.



27 Omran H, Jung W, Schimpf R et al. Echocardiographic parameters

for predicting maintenance of sinus rhythm after internal atrial de-

fibrillation. Am J Cardiol 1998; 81: 1446–9.

28 Barbaro V, Bartolini P, Calcagnini G et al. Effect of atrial fibrilla-

tion organization on internal defibrillation threshold. Ann Ist Super

Sanita 2001; 37: 377–81.

29 Kirchhof P, Borggrefe M, Breithardt G. Effect of electrode position

on the outcome of cardioversion. Card Electrophysiol Rev 2003; 7:

292–6.

30 Joglar JA, Kowal RC. Electrical cardioversion of atrial fibrillation.

Cardiol Clin 2004; 22: 101–11.

31 Duytschaever M, Haerynck F, Tavernier R et al. Factors influen-

cing long term persistence of sinus rhythm after a first electrical

cardioversion for atrial fibrillation. Pacing Clin Electrophysiol 1998;

21: 284–7.

32 Curtis AB, Seals AA, Safford RE et al. Clinical factors associated

with abandonment of a rate-control or a rhythm-control strategy

for the management of atrial fibrillation in the AFFIRM study. Am

Heart J 2005; 1492: 304–8.

33 Naccarelli GV, Wolbrette DL, Bhatta L et al. A review of clinical tri-

als assessing the efficacy and safety of newer antiarrhythmic drugs

in atrial fibrillation. J Interv Card Electrophysiol 2003; 9: 215–22.

34 VanGelder IC, Tuinenburg AE, Schoonderwoerd BS et al. Pharma-

cologic versus direct-current electrical cardioversion of atrial flutter

and fibrillation. Am J Cardiol 1999; 84: 147R–51R.

35 Elesber A, Rosales A, Herges R et al. Relapse and mortality follow-

ing cardioversion of new-onset vs. recurrent atrial fibrillation and

atrial flutter in the elderly. EHJ 2006; 27: 854–60.

36 Blomstrom-Lundqvist C, Scheinman MM, Aliot EM et al.

ACC/AHA/ESC guidelines for the management of patients with

supraventricular arrhythmias–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 (writing committee

to develop guidelines for the management of patients with su-

praventricular arrhythmias) developed in collaboration with NA-

SPE-Heart Rhythm Society. J Am Coll Cardiol 2003; 42: 1493–

531.

37 Fetsch T, Bauer P, Engberding R et al. Prevention of atrial fibrilla-

tion after cardioversion: results of the PAFAC trial. Eur Heart J

2004; 25: 1385–94.

Paper received November 2006, accepted December 2006



Documents

Electrical cardioversion for persistent atrial fibrillation or atrial flutter in clinical practice: predictors of long-term outcome