Upload
g-boriani
View
220
Download
4
Embed Size (px)
Citation preview
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:
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
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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
Outcome predictors of AF ⁄ AFL cardioversion 751
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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
752 Outcome predictors of AF ⁄ AFL cardioversion
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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
Outcome predictors of AF ⁄ AFL cardioversion 753
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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
754 Outcome predictors of AF ⁄ AFL cardioversion
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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.
Outcome predictors of AF ⁄ AFL cardioversion 755
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756
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
756 Outcome predictors of AF ⁄ AFL cardioversion
ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd Int J Clin Pract, May 2007, 61, 5, 748–756