OUP user menu

★ Editor's choice ★

The Atrial Fibrillation Ablation Pilot Study: an European Survey on Methodology and results of catheter ablation for atrial fibrillation conducted by the European Heart Rhythm Association

Elena Arbelo, Josep Brugada, Gerhard Hindricks, Aldo P. Maggioni, Luigi Tavazzi, Panos Vardas, Cécile Laroche, Frédéric Anselme, Giuseppe Inama, Pierre Jais, Zbigniew Kalarus, Josef Kautzner, Thorsten Lewalter, Georges H. Mairesse, Julian Perez-Villacastin, Sam Riahi, Milos Taborsky, George Theodorakis, Serge A. Trines
DOI: http://dx.doi.org/10.1093/eurheartj/ehu001 1466-1478 First published online: 31 January 2014


Aims The Atrial Fibrillation Ablation Pilot Study is a prospective registry designed to describe the clinical epidemiology of patients undergoing an atrial fibrillation (AFib) ablation, and the diagnostic/therapeutic processes applied across Europe. The aims of the 1-year follow-up were to analyse how centres assess in routine clinical practice the success of the procedure and to evaluate the success rate and long-term safety/complications.

Methods and results Seventy-two centres in 10 European countries were asked to enrol 20 consecutive patients undergoing a first AFib ablation procedure. A web-based case report form captured information on pre-procedural, procedural, and 1-year follow-up data. Between October 2010 and May 2011, 1410 patients were included and 1391 underwent an AFib ablation (98.7%). A total of 1300 patients (93.5%) completed a follow-up control 367 ± 42 days after the procedure. Arrhythmia documentation was done by an electrocardiogram in 76%, Holter-monitoring in 52%, transtelephonic monitoring in 8%, and/or implanted systems in 4.5%. Over 50% became asymptomatic. Twenty-one per cent were re-admitted due to post-ablation arrhythmias. Success without antiarrhythmic drugs was achieved in 40.7% of patients (43.7% in paroxysmal AF; 30.2% in persistent AF; 36.7% in long-lasting persistent AF). A second ablation was required in 18% of the cases and 43.4% were under antiarrhythmic treatment. Thirty-three patients (2.5%) suffered an adverse event, 272 (21%) experienced a left atrial tachycardia, and 4 patients died (1 haemorrhagic stroke, 1 ventricular fibrillation in a patient with ischaemic heart disease, 1 cancer, and 1 of unknown cause).

Conclusion The AFib Ablation Pilot Study provided crucial information on the epidemiology, management, and outcomes of catheter ablation of AFib in a real-world setting. The methods used to assess the success of the procedure appeared at least suboptimal. Even in this context, the 12-month success rate appears to be somewhat lower to the one reported clinical trials.

  • Atrial fibrillation
  • Catheter ablation
  • Outcomes
  • Survey
  • Registry


Atrial fibrillation (AFib) is the most common cardiac arrhythmia that is associated with a reduced quality of life and an increased number of adverse outcomes such as stroke, heart failure, increased number of hospitalizations, and mortality.13 During the past decade, catheter ablation of AFib has become a common procedure at most major medical centres worldwide. However, there is significant variability in the ablation strategies and the reported mid- and long-term success.49

The European AFib Ablation Pilot registry conducted by the European Heart Rhythm Association (EHRA) of the European Society of Cardiology (ESC) was designed to describe the clinical epidemiology of patients undergoing an AFib ablation procedure, the diagnostic/therapeutic processes (including technical aspects, ways to measure the success of the procedure) and the acute and mid-term outcomes, and complications of the procedure across Europe.10,11

The results of the in-hospital phase of the AFib Ablation Pilot were recently published.10 We present the results of the 1-year follow-up analysis. The main objectives of the follow-up phase were to assess (i) how the participating centres evaluate, in routine clinical practice, the success of the procedure during the follow-up, (ii) the rate of success of AFib ablation procedures in routine clinical practice, and (iii) the long-term safety/complications of the procedure.


Study design

The AFib Ablation Pilot Study is a prospective, multicentre, observational registry of consecutive patients undergoing a first AFib ablation procedure in 72 Cardiology Centres of 10 European countries, selected to represent the different regions of the European continent.

  • Four Western European countries: Belgium, France, Germany, and the Netherlands.

  • Two Eastern European countries: Czech Republic, and Poland.

  • Three Southern European countries: Greece, Italy, and Spain.

  • One Northern European country: Denmark.

The National Cardiology Societies of each country agreed to participate in the registry, contributing in the selection of centres and updating the investigators and the ESC with the ethical and legal requirements. The number of centres in each country varied according to its size and the number of centres available. Site selection targeted hospitals with a medium-to-high expertise (performing ≥50 AFib ablation procedures/year). The National Coordinator was responsible for contacting the investigators and the implementation of the protocol in their country, ensuring performance of the enrolling centres and quality of national data. Centres accepted on a voluntary basis through National Coordinators, and the survey was approved by the national and/or local Institutional Review Board, according to the regulations of each participating country. Participating centres were asked to enrol, between October 2010 and May 2011, 20 consecutive patients undergoing a first AFib ablation procedure and follow them up for 1 year. There were no exclusion criteria. Data were collected after detailed information was given to the patient and a signed informed consent was obtained.

The EURObservational Research Programme (EORP) Department of the ESC operationally coordinated the project, provided support to the Committees, National Coordinators, and participating centres and guaranteed the methodological concepts of the survey. The database was set up at the European Heart House of the ESC (France), according to the requirements defined by the Executive Committee with the support of the EORP Department. Statistical analyses of the database were performed by the EORP team.

This study was intended as Pilot because it also aimed to validate Protocol, case report form (CRF), and organizational structure of the study in the perspective of a larger European long-term study on the same matter, including centres with any level of procedure volume.

Data collection

All centres were asked to complete a one-time site questionnaire describing the type and size of the centre, reference area population, facilities, and number of invasive procedures performed.

Data were collected using a web-based system. An electronic CRF was developed by the Executive Committee of the study to capture the following information for each enrolled patient:

  • Enrolment data: demographics, risk factors and co-morbidities, precipitating factors, type of AFib, signs and symptoms, pharmacological and non-pharmacological treatments, invasive/non-invasive diagnostic procedures, echocardiographic characteristics (imaging techniques prior to the procedure).

  • Procedural data: number of personnel, laboratory setting, type of catheters, type of energy, type of imaging techniques, type of anaesthesia, anticoagulation protocol, type of procedure, Rx exposure, outcome parameters used to define success, and complications.

  • Post-procedural data: periodical ECG, 24-h Holter -onitoring, transtelephonic ECG-monitoring, implanted systems of monitoring, other.

  • Follow-up data: information collected for all patients after 12 months from the procedure. Centres planned their follow-up according to their usual clinical practice.

Six countries out of the 10 that participated in the registry were randomly selected for local monitoring of the data (Czech Republic, Denmark, France, Italy, Poland, and Spain). Within these countries, three to four centres per country were also randomly chosen (total of 20 centres). In these centres, consecutiveness of the inclusion of patients and correctness of the data included in the CRF were checked.


One-year success is defined as patient survival free from any atrial arrhythmia, with or without antiarrhythmic drugs, as assessed from the end of the 3-month blanking period to 12 months following the ablation procedure.

Arrhythmia recurrence is defined as an electrocardiographically documented episode of AFib or atrial flutter lasting at least 30s, after a 3-month blanking period.

Early recurrence is defined as a recurrence of any atrial arrhythmia, within 3 months from ablation.

Cavo-tricuspid isthmus-dependent flutter was excluded from all definitions.

A blanking period of 3 months was employed after ablation. Recurrences within the first 3 months were not classified as failure of the procedure.

Statistical analysis

A univariate analysis was done for both the continuous and categorical variables. Continuous variables are reported as means ± standard deviation or as median and inter-quartile range (IQR). Between-group comparisons were made using a non-parametric test (Kruskal–Wallis test). Categorical variables are reported as percentages. Between-group comparisons were made using a Chi-square test or Fisher's exact test if any expected cells count was <5.

A multivariate analysis was used to explore relationship between 12-month arrhythmia-free-survival and baseline covariates for the follow-up population. All the baseline covariates were compared between the two-groups (success and recurrence patients at 12-month follow-up using the Chi-squared or Kruskal–Wallis test for categorical or continuous variables, respectively), and only the significant variables were included in the multiple logistic model. The multiple logistic regression was done using a multiple imputation procedure to correct for the presence of missing data.

The Kaplan–Meier estimator, the product limit estimator, was used for estimating and plotting the survival functions.

A P-value of <0.05 was considered statistically significant. All tests were two-sided. Analyses were performed with the SAS system software (SAS Institute, Inc., Cary, NC, USA). The multiple logistic regression procedure was performed using the programme R (R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL ‘http://www.R-project.org/’) and the package Hmisc (Hmisc: Harrell Miscellaneous; R package version 3.9-0, http://CRAN.R-project.org/package=Hmisc).


The detailed information on the demographic and baseline clinical characteristics as well as the diagnostic interventions and ablation procedure data of the AFib Ablation Pilot registry population have already been published elsewhere.10

Participating centres and total cohort

Seventy-two centres across Europe included a total of 1410 patients. The hospital reference area included a median number of inhabitants of 500 000 (IQR: 200 000–1 500 000) with a median annual number of AFib catheter ablation procedures of 179 (IQR: 80–346). The median annual number of AFib catheter ablation procedures was different across the regions which participated to the study (p 0.0128): Western European countries 250 (IQR: 104–400), Eastern European countries 91 (IQR: 60–346), Southern European countries 90 (IQR: 55–179), Northern European country 235 (IQR: 160–485).

In 19 patients, the ablation was not done (1.3%) due to an intracardiac thrombus (7 patients), tamponade during transeptal puncture (7 patients), non-procedural complications (4 patients), and a cerebrovascular event (1 patient). During the in-hospital phase, there was 1 death out of 1391 AFib ablation procedures (0.07%).

Of the 1390 patients that completed the in-hospital phase of the registry, 1300 patients (93.5%) underwent a follow-up control 1 year after the procedure. Figure 1 shows the number and distribution of centres and patients at the in-hospital and 12-month follow-up phases.

Figure 1

Distribution of centres and patients included in the Atrial Fibrillation Ablation Pilot by country.

There were no significant differences in the baseline and AFib characteristics between the group of patients that completed the 12-month evaluation and the ones lost to follow-up (Supplementary material online, Annex S1). With regard to the ablation procedure, the only difference found was the ablation of the left atrial roof and the cavo-tricuspid isthmus (P < 0.005) (Supplementary material online, Annex SI). However, the significant differences observed for these two parameters could be explained by the center effect (two centres did not perform any follow-up for the enrolled patients, due to centre closure and lack of resources, respectively).

Follow-up status

As defined by the protocol, a follow-up evaluation was done 12 months after the index procedure (median 368 days, IQR: 358–385 days), by an in-person clinical evaluation in 57.6% and telephonic contact in 41.4%. In-person clinical follow-up was more frequent in Southern and Eastern European countries (P < 0.0001). Table 1 summarizes the 12-month follow-up information and the differences across European regions.

View this table:
Table 1

Clinical information at follow-up by European region

Western (n = 681)Eastern (n = 194)Northern (n = 99)Southern (n = 415)P-valueTotal (n = 1300)
Type of follow-up at 12 months (%)<0.0001
 Clinical visit52.665.236.668.558.2
 Telephone contact47.534.863.431.541.8
At least one ECG during follow-up82.794.481.592.2<0.000187.2%
Type of ECG monitoring (%)
 Holter monitoring46.455.640.964.4<0.000152.9
 Transtelephonic monitoring1.020.2015.8<0.00018.4
 Implanted-monitoring systems2.<0.00014.5
 ECG + multiday recording43.865.938.865.6<0.000153.3
Other cardiovascular tests (%)
 Transthoracic echocardiogram27.425.016.138.9<0.000129.9
 Transoesophageal echocardiogram10.28.721.56.1<0.00019.5
 Chest X-ray8.
 Cardiac CT4.
 Cardiac MRI2.
 EP study2.
Documentation of AF/atrial flutter33.137.848.936.00.027935.8
 Transtelephonic monitoringa4.77.42.412.70.03447.5
 Written diagnosisa36.530.465.924.4<0.000134.7
 Hospital discharge recorda22.735.747.611.8<0.000123.8
Repeat ablation procedure
 Percutaneous AF ablationb22.016.531.210.5<0.000118.3
 Surgical ablationb81.596.793.182.90.426885.1
 Surgical ablationb5.203.54.914.3
Clinical visits before 12-month follow-up (%)
 Emergency room11.113.918.315.00.123313.2
Re-admissions (%)
 Secondary to AF/atrial flutter/atrial tachycardia21.519.749.514.7<0.000121.2
 Secondary to other cardiovascular events6.
 Secondary to non-cardiovascular events6.
Cardiovascular events requiring admission (%)
 Acute coronary syndromec5.9009.1>0.9995.7
 Heart failureb38.2009.10.107326.4
 Coronary interventionb22.925.
Baseline rhythm at 12-month follow-up (%)
 Atrial fibrillation5.88.813.08.47.6
 Common atrial flutter2.
 Atypical atrial flutter1.
 Pacemaker rhythm1.31.501.61.3
Anticoagulation treatment (%)
 At discharge97.996.898.993.80.002696.5
 At the 12-month follow-up visit68.371.360.257.10.000464.7
 After the 12-month follow-up visit60.254.857.047.90.001855.4
Antiarrhythmics treatment (%)
 At discharge62.861.243.075.1<0.000164.9
 At the 12-month follow-up visit49.845.715.156.6<0.000148.9
 After the 12-month follow-up visit40.134.012.947.4<0.000139.5
  • CT, computed tomography; MRI, magnetic resonance imaging; EP, electrophysiological.

  • a% of patients with recurrence.

  • b% of repeat ablation procedures.

  • c% of hospital re-admissions due to other cardiovascular events.

Clinical symptoms during follow-up

Over half of the population became asymptomatic after the ablation procedure (54.7%). The most commonly reported symptoms were palpitations (34.3%), fatigue (12.6%), and exertion dyspnoea (12.9%). Palpitations were more frequent in paroxysmal AFib (36.4 vs. 30%, P = 0.025) while dyspnoea was more frequent in non-paroxysmal AFib (15.9 vs. 11.3%, P = 0.023). A higher prevalence of fatigue in non-paroxysmal (15.1 vs. 11.4%) was observed even if the difference did not reach the conventional level of statistical significance (P = 0.065). Other reported symptoms included weakness (6.3%), chest pain (4.8%), dizziness/presyncope (3.2%), and syncope (0.6%).

Diagnostic procedures at follow-up

At least one electrocardiogram during the follow-up was done in 87.2% of patients, showing sinus rhythm in 87.9%, AFib in 7.6%, atypical atrial flutter in 1.3%, common atrial flutter in 1.3%, and a non-defined pacemaker rhythm in 1.7%. Patients in sinus rhythm showed a significantly slower median heart rate (67 b.p.m., IQR: 60–67 b.p.m.) with respect to patients in AFib or atrial flutter (90 b.p.m., IQR: 70–100 b.p.m.) (P < 0.0001).

Holter-monitoring was done in about half of the population. Transtelephonic or implanted-monitoring systems were less frequent. Altogether, patients having repeated continuous monitoring of ≥24 h (Holter- or implanted-monitoring systems) represented only 57.4% of the population. The use of any type of ECG monitoring was more frequent in Eastern and Southern European countries (Table 1).

Other cardiovascular tests included: transthoracic echocardiogram (29.9%), transoesophageal echocardiogram (9.5%), cardiac CT (3.1%), cardiac MRI (2.8%), chest X-ray (5.5%), coronary angiography (1.5%), electrophysiological study (2.7%), and a myocardial scintigraphy (0.6%).

Clinical evaluations and re-admissions

During the period between discharge after the procedure and the 12-month evaluation, 78.4% of patients had a median of 2 (IQR: 2–3) cardiology visits, and 13.2% of patients had a median of 1 (IQR: 1–2) of visits to the emergency room.

Thirty per cent of patients were re-admitted after the procedure: 21.2% due to post-ablation atrial arrhythmias, 4.6% due to other cardiovascular events and 4.6% due to non-cardiovascular events.

Treatment at follow-up

The type of pharmacological treatment at the 1-year follow-up visit compared with the discharge status is shown in Figure 2.

Figure 2

Rate of use of pharmacological treatment at discharge and at the 12-month follow-up.

Overall, 64.7% of patients were under some kind of anticoagulant treatment, showing a relationship to the cardioembolic risk: 48% of patients with a CHADS2-Vasc score of 0, 63.3% with a CHADS2-Vasc score of 1, and 76.2 with a CHADS2-Vasc >1 (P < 0.0001). Treatment with oral anticoagulation was more frequent in patients with non-paroxysmal AFib (88.9 vs. 76.9%, P < 0.001).

Forty-nine per cent of patients were under antiarrhythmic treatment at the 12-month evaluation. Patients that had undergone catheter ablation for paroxysmal AFib were also more likely to be off drugs (55%) when compared with non-paroxysmal AFib (46.4%) (P = 0.004). Also, patients referring symptoms during the follow-up were more likely to be under antiarrhythmic treatment (54.8 vs. 41.5%, P < 0.0001).

Regional differences in anticoagulant and antiarrhythmic treatment are summarized in Table 1.

Outcome of atrial fibrillation ablation

Complete data for the assessment of the one-year success were available in 1281 patients.

Globally, at least one episode of atrial arrhythmias (AFib or flutter) was documented in 459 of patients (35.8%) for the duration of the 12 months after the index procedure. Patients with non-paroxysmal AFib showed a higher probability of arrhythmia recurrence (43.6 vs. 32.0% in paroxysmal AFib, P < 0.0001). During the blanking period, 194 patients (15.1%) suffered at least one episode of documented atrial arrhythmia. Two of these patients died within the first 3 months after the procedure (see ‘Complications’ section). On the other hand, 334 (26.1%) had at least one documented recurrence after the 3-month blanking window, 69 of which (20.7%) had also had a recurrence during the blanking period. Asymptomatic recurrences occurred in 25.8% of patients (Figure 3).

Figure 3

Flowchart representing numbers of individuals at each stage of the study.

The procedure was considered successful in 944 patients (73.7%) (i.e. no arrhythmia documentation after the 3-month blanking period and no death). Success without antiarrhythmic drugs was achieved in only 522 patients (40.7%); however, this number should be considered carefully as the protocol did not require the discontinuation of the treatment after the blanking period [332 patients (26.1%) continued on antiarrhythmic drugs despite no documentation of recurrence]. To achieve the reported success rate, 18.3% of patients required a second ablation procedure (percutaneous catheter ablation in 85.0%, surgical in 4.3%, and other cardiac procedures in 10.7%) and 43.4% were still under antiarrhythmic treatment at the 12-month evaluation. There was no statistical difference in the reported success rate across European regions (P = 0.1014).

Success rates were significantly lower in patients with long-lasting AFib than in patients with paroxysmal AFib (Table 2). Figure 4 shows the arrhythmia-free-survival curve by type of AFib. A third of patients (34.4%) in whom the ablation was considered successful, still reported having symptoms, mainly in the form of palpitations, dyspnoea, and fatigue.

View this table:
Table 2

Success rates in relationship with the type of atrial fibrillation

Success without antiarrhythmic drugs (AADs)*Success with antiarrhythmic drugs (AADs)*Overall success*
Type AFNo. countriesNo. centresNo. patientsn (%)Median (IQR) between countriesn (%)Median (IQR) between countriesn (%)Median (IQR) between countries
Paroxysmal1070871381 (43.7)42.0 (39.4–51.9)256 (29.4)29.6 (4.8–42.5)647 (74.3)72.1 (70.7–75.4)
Persistent106126580 (30.2)27.0 (17.9–47.1)99 (37.4)41.2 (17.6–55.6)191 (72.1)70.5 (64.7–76.5)
Long-lasting persistent10409033 (36.7)42.9 (31.6–50.0)18 (20.0)22.2 (14.3–28.9)52 (57.8)56.3 (25.0–64.7)
Overall10701226494 (40.3)41.1 (28.6–50.0)373 (30.4)28.9 (18.5–41.6)890 (72.6)70.5 (58.8–75.4)
  • *No discontinuation of the antiarrhythmic medication was required by the protocol.

Figure 4

Kaplan– Meier arrhythmia-free survival curve by type of atrial fibrillation.

The likelihood of success was not influenced by the number of annual AFib ablation procedures per centre (P > 0.999.). In contrast, body mass index, type of AFib, absence of underlying disorder (i.e. lone AF), and LA atrial diameter were significantly associated with the success (Table 3). With regard to procedure-related characteristics, the type of ablation energy and the attempt and/or achievement of pulmonary vein (PV) isolation were not associated with the 12-month success of the procedure (Table 4). On the contrary, the ablation of fractionated electrograms in the left and right atrium, and the slowing of activation during ablation of fractionated sites were significantly related to the probability of success. Left atrial roof linear ablation and mitral isthmus line and the ablation of ganglionated plexi were of borderline statistical significance (Table 4). Finally, arrhythmia recurrences during the blanking period were also inversely associated with the 1-year success (P = 0.0004).

View this table:
Table 3

Baseline clinical characteristics by 1-year success

Patients with 1-year success (n = 944)Patients with recurrence (n = 337)P-valueTotal (n = 1281)
Age (years) (%)
 Median (IQR)60 (52–66)60 (53–66)0.60660 (52–66)
 >65 years31.030.30.80331.8
 >70 years13.212.80.85013
 >80 years0.60.60.9310.6
Females, %28.926.20.33327.9
Body mass index (kg/m²)
 Median (IQR)27 (24–30)28 (25–31)0.03927 (25–30)
Hypertension (%)50.351.30.73650.6
 Hyperthyroidism (%)
 Chronic kidney disease (%)
 Chronic obstructive pulmonary disease (%)
 Sleep apnoea (%)
Type of AFib (%)
 Not defined11.51.1
Lone atrial fibrillation39.133.20.03937.5
Underlying disorder (%)
 Valvular heart disease11.714.812.6
 Coronary artery disease3.64.53.8
 Dilated cardiomyopathy2.25.03.0
 Hypertrophic cardiomyopathy2.92.42.7
 Chronic heart failure2.73.02.8
 Other cardiac disease2.83.02.9
 Chronic obstructive pulmonary disease0.61.20.8
 Not defined3.82.73.5
Associated symptoms at baseline85.492.50.00187.3
LA diameter (mm)
 Median (IQR)42 (38–46)44 (41–48)0.00142 (39–47)
LV ejection fraction (%)
 Median (IQR)60 (55–65)60 (55–65)0.13760 (55–65)
  • IQR, inter-quartile range; LA, left atrial; LV, left ventricular.

View this table:
Table 4

Procedural data and ablation strategy by 1-year success

Patients with 1-year success (n = 944)Patients with recurrence (n = 337)P-valueTotal (n = 1281)
Energy source (%)
 Non-irrigated radiofrequency3.63.70.3633.6
 Radiofrequency with closed irrigation2.32.82.4
 Radiofrequency with open irrigation77.878.377.9
 Duty-cycled radiofrequency energy5.02.24.2
 Laser balloon (endoscopic ablation system)
Attempt to isolate the pulmonary vein (%)
Achievement of entrance and exit block (%)*
Attempt to isolate all pulmonary veins (%)93.694.10.78493.8
Achievement of entrance and exit block of all pulmonary veins (%)
Left atrial linear lesion (%)
 Roof line17.621.70.09418.7
 Mitral isthmus line12.115.80.08413.1
 Other left atrial linear lesion6.911.30.0108.1
Verification of complete conduction block across linear left atrial lesions (%)
 Roof line17.821.70.26518.7
 Mitral isthmus line13.620.50.03115.2
 Other left atrial line3.18.110.0084.2
Right atrial linear lesion (%)
 Superior vena cava2.91.80.2802.6
 Cavo-tricuspid linear lesion17.016.10.70616.7
 Achievement of bidirectional CTI block32.837.50.27433.9
Ablation at fractionated electrogram sites (%)
 In the left atrium15.322.90.00217.3
 In the right atrium4.07.40.0144.9
 Ablation of fractionated sites slowed activation or terminated AF10.916.40.02512.3
 Ablation of autonomic ganglionated plexi2.94.80.0963.4
  • LSPV, left superior pulmonary vein; LIPV, left inferior pulmonary vein; RSPV, right superior pulmonary vein; RIPV, right inferior pulmonary vein; CTI, cavo-tricuspid isthmus.

  • *No discontinuation of the antiarrhythmic medication was required by the protocol.

When including the significant variables of the univariate analysis into a multivariate model, the only predictor of failure 12 months after the procedure was having at least one arrhythmic recurrence during the blanking period [OR: 3.01 (2.18–4.16), P < 0.0001].

Complications during follow-up

Table 5 summarizes the incidence of complications in relation to AFib ablation. Excluding post-ablation atrial flutter/tachycardia, 32 patients (2.5%) suffered an adverse event during the follow-up: 13 pacemaker implantations, 9 vascular injuries (6 arterial-venous fistulae and 3 pseudoaneurysms), 7 cerebrovascular events, 2 phrenic nerve injuries (one with cryo-balloon ablation and the other with open-irrigation radiofrequency), and 1 significant PV stenosis (≥75%) requiring intervention. Additionally, 272 patients experienced atrial flutter/tachycardia, providing an adverse event rate during the first year after an AFib ablation of 22.5%. No significant differences were found when analysing the complication rate by number of annual procedures per centre (P > 0.999 n.s.).

View this table:
Table 5

Adverse events associated with catheter ablation of atrial fibrillation

Inclusion (n = 1410) (%)In-hospital10 (n = 1391) (%)12-month FU (n = 1300) (%)Overall (n = 1410) (%)
Cardiovascular7 (0.50)46 (3.31)13 (1.00)66 (4.68)
 Bradycardia requiring pacemaker implantation3 (0.22)13 (1.00)16 (1.13)
 Cardiac arrest1 (0.07)1 (0.07)
 Endocarditis1 (0.07)1 (0.07)
 Myocardial infarction1 (0.07)1 (0.07)
 Pericarditis17 (1.22)17 (1.21)
 Cardiac perforation7 (0.50)11 (0.79)18 (1.28)
 Atypical atrial flutter4 (0.29)4 (0.28)
General6 (0.43)6 (0.43)
 Allergic reaction4 (0.29)4 (0.28)
 Sepsis2 (0.14)2 (0.14)
Peripheral/vascular18 (1.29)9 (0.69)25 (1.77)
 AV fistula6 (0.43)6 (0.46)11 (0.78)
 Pseudoaneurysm6 (0.43)3 (0.23)8 (0.57)
 Haematoma or bleeding requiring evacuation or transfusion5 (0.36)5 (0.35)
 Peripheral thromboembolic event1 (0.07)1 (0.07)
PV stenosis (≥75%) requiring intervention1 (0.08)1 (0.07)
 Neuro1 (0.07)9 (0.65)9 (0.69)16 (1.13)
  Cerebrovascular event1 (0.07)8 (0.58)7 (0.54)13 (0.92)
  Phrenic nerve injury2 (0.14)2 (0.15)4 (0.28)
 Pulmonary8 (0.57)8 (0.57)
  Haemothorax3 (0.22)3 (0.21)
  Pleural effusion2 (0.14)2 (0.14)
  Pneumothorax1 (0.07)1 (0.07)
  Unspecified2 (0.14)2 (0.14)
  Gastrointestinal1 (0.07)1 (0.07)
  Oesophageal ulceration1 (0.07)1 (0.07)
  Atrial tachycardia/atrial flutter4 (0.29)272 (20.9)274 (19.43)
 Other30 (2.16)30 (2.13)
 Death1 (0.07)4 (0.31)5 (0.35)
  Cardiovascular1 (0.07)1 (0.08)2 (0.14)
  Non-cardiovascular1 (0.08)1 (0.07)
  Unknown2 (0.15)2 (0.14)
Overall8 (0.006)104 (7.48)293 (22.54)369 (26.74)
  • AV, atrioventricular; PV, pulmonary vein.

In addition to 1 death during the in-hospital phase, 4 deaths occurred during the 1-year follow-up period: 2 during the blanking period and 2 after the first 3-month window. The cause of death was vascular in two patients. The first was due to a haemorrhagic stroke 3 days after discharge in a woman that had undergone catheter ablation for paroxysmal AF with bridging (two doses of enoxaparin 1 mg/kg) from intraprocedural unfractionated heparin to acenocumarol during admission with a normal INR (2.1) upon re-admission. The second vascular death was secondary to an out-of-hospital ventricular fibrillation probably due to an acute coronary syndrome in a man with prior history of coronary artery disease (surgical revascularization 30 years earlier) that had undergone an ablation for persistent AFib 10 months earlier. There was one non-vascular death secondary to metastatic pancreatic cancer 8 months after paroxysmal AFib ablation. Finally, in one patient the cause of death remained unknown (1 out-of-hospital sudden death 38 days after persistent AFib ablation with no autopsy).


The EORP of the ESC started in August 2009 with the aim to provide a better understanding of cardiology practiced in Europe. The Atrial Fibrillation Ablation Pilot Study is the first, systematic, prospective international study specifically designed for collecting information reflecting the current clinical practice in patients undergoing an AFib catheter ablation. Importantly, according to the observational nature of the study, no recommendation was done in terms of drug therapy, arrhythmia documentation methods, etc. Thus, the information obtained by the registry reflects the everyday practice across Europe.

Registries are an important source of information on how the evidence acquired through controlled trials, usually on highly selected patients managed according to detailed protocols, are adopted in clinical practice. The consecutiveness of patients warrants the representativeness of the population and allows reflecting the actual clinical practice. Barely 6.5% of patients were lost to follow-up, ensuring the quality of the information retrieved during the 12 months after the procedure. To guarantee a clear presentation of our observational research and its results, we have followed the STROBE recommendations.12

Follow-up evaluation

This survey provides intriguing information on how patients are being followed up and how arrhythmia recurrences are sought. Despite the recommendations of 2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of AFib13 over a third of the 12-month follow-up evaluations were done by telephonic contact. For arrhythmia-recurrence monitoring, 77% of patients had a periodical electrocardiogram and only 57.4% of the population underwent repeated long-duration (≥24 h) electrocardiographic monitoring. This is clearly another gap between 2012 HRS/EHRA/ECAS expert consensus13 and the actual clinical practice that should be considered in the future. Of note, non-compliance with the current recommendations appeared to be higher in Northern and Western European countries with lower rates of electrocardiogram or long-duration ECG monitoring.

Over half of the population became asymptomatic after the ablation procedure (symptoms were present in 90% of patients at baseline). Of note is that 26% of recurrences were asymptomatic. In contrast, patients with no documented recurrence reported symptoms in a third of cases. It is impossible to confirm whether these were related or not to post-ablation arrhythmias. Moreover, the low rates of long-duration ECG monitoring suggest that arrhythmias could occur in a relevant proportion of both symptomatic and asymptomatic patients.14,15

One year after the procedure, 65% of patients were taking anticoagulants. Surprisingly, about half of the population with a low cardioembolic risk (CHASD2-Vasc 0) where still anticoagulated 1 year after the procedure. Conversely, up to 25% of patients with high-cardioembolic risk (CHASD2-Vasc >1) were not taking any anticoagulant drug. These figures document how difficult it is to implement simple guidelines in everyday clinical practice, even in expert centres performing catheter ablation of AFib.

Outcomes of atrial fibrillation ablation

In this relatively large sample of patients undergoing a first AFib ablation in centres with medium-to-high expertise, catheter ablation was able to maintain sinus rhythm in 40.7% in the absence of any antiarrhythmic drug, and in an additional 31.1% with the use of antiarrhythmic treatment formerly reported as ineffective. These results required a repeat ablation procedure in almost a fifth of the population. This success rate without antiarrhythmic drugs appears lower than the previously reported.1618

As expected, patients with non-paroxysmal AFib had more AFib recurrences, confirming results from other clinical trials.13,1922 Furthermore, this AFib Ablation registry corroborates that the factors impacting procedural success are the type of AFib and the diameter of the left atrium. However, the impact of these variables as predictors of success is minimized by the occurrence of atrial arrhythmias during the blanking period. This survey, in line with some previously published data,2326 suggests that the occurrence of arrhythmias within the first 3 months after the procedure may be a relevant predictor of the likelihood of 1-year success. However, these data should be interpreted with caution, as the type of follow-up and arrhythmia documentation varied widely among centres. Also, this finding should not prompt early reablation as 63% of patients with early recurrence did not have any further arrhythmias, as it has also been observed by other authors.2427

Complications of atrial fibrillation ablation

A total of five deaths were reported, three of which occurred perioperatively (one during the in-hospital phase due to infective endocarditis in a pacemaker patient10 and two within the first 2 months after the ablation due to a haemorrhagic stroke and of unknown origin). This represents a 0.18% death rate per AFib ablation procedure (3 in 1629 ablations). The other two deaths were delayed and not procedure related.

Importantly, in this real-world population, complication rates were similar to the rates as observed in other multicentre surveys28,29 but slightly higher than single-centre experiences.16,17,3035 The incidence of complications during the year after the index AFib ablation procedure was low (2.5% of which, 0.8% were major), a third which (37.5%) were related to the AFib ablation procedure (one significant PV stenosis, two phrenic nerve injuries, and nine vascular injuries following a second procedure). However, the peri-operative complication rate is not negligible (7.7% of which, 1.7% was major) and must be weighed against the potential benefits when considering AFib ablation in patients with mild symptoms. Furthermore, if we include post-ablation atrial tachycardia/flutter as a complication of the procedure, the adverse event rate at 12 months raises to 26.5%.


There are several limitations to this study that should be taken into account. This registry was based on voluntary participation of centres. The centres were selected proportionately to the size of the population of the participating countries in order to favour representativeness of the cohort. However, not all contacted centres finally contributed to the registry in its Pilot phase. All the same, the high rate of response (73% of the contacted centres) minimizes the risk of an inclusion bias, and offers a good picture of the real situation of AFib ablation across Europe. On the other hand, it has to be emphasized that only medium to high expertise centres were approached and selected. Therefore, it is likely that everyday practice could be even worse. Consecutiveness of enrolment was reinforced and a local audit was performed in 20 randomly selected centres to confirm the correctness of data (which was verified in over 95% of cases, with no significant differences between the participating countries). In addition, the EORP Department of the ESC monitored the study data closely and all data underwent extensive automatic edit and plausibility checks to detect inaccuracies and inconsistencies. Ninety-one patients were lost to follow-up (6.5%). Being a prospective observational study, this could introduce a selection bias. However, there were no major clinical differences between patients lost to follow-up and the ones reported (Supplementary material online, Annex SI) and a particular effort was done by the investigators to exclude the occurrence of major adverse events (particularly, death). Also, the follow-up period to test the efficacy is relatively short and does not assess long-term results of the procedure, which has recently been shown to decrease considerably overtime.79,36,37 Finally, no recommendation was made with regard to post-ablation arrhythmia monitoring. The variability in the monitoring methodology may limit the interpretation of recurrence data as asymptomatic episodes of arrhythmia may have been overlooked.38

Clinical implications

The AFib Ablation Pilot represents a reasonably large sample reflecting the current practice in the use of catheter ablation in the management of patients with AFib in Europe. Moreover, it offers a very detailed profile of the AFib ablation population, analysing demographic characteristics, cardiovascular risk factors as well as the clinical history of the patient, and evaluating how these factors influence outcomes in a real-world setting. These data may prove useful when designing management strategies within the different European countries.

The results demonstrate the degree to which physicians implement the results of clinical trials and guideline recommendations. Also, it raises relevant clinical questions, such as the optimal post-ablation monitoring or the need of anticoagulant treatment, which should be addressed in future strategic programmes.


The Survey was funded by the ESC. Each participating National Cardiology Society was granted €10.000 to help for the organizational needs regarding the national network implementation. At present, the following companies are supporting the EURObservational Research programme: GOLD: Abott Vascular, Bayer Pharma, Bristol Myers Squibb (BMS), Pfizer, Boehringer Ingelheim, Daiichi Sankyo Europe, Menarini international Operations, Novartis Pharma, Sanofi-Aventis, Servier International. SILVER: Amgen. BRONZE: Boston Scientific International, Merck & Co. (MSD).

Conflict of interest: none declared.


Executive Committee (appointed by the ESC Heart Rhythm Association)

Josep Brugada (Chairperson), Elena Arbelo, Gerhard Hindriks, Aldo P. Maggioni (non-voting member), John Morgan, Luigi Tavazzi, Panos Vardas.

Oversight Committee

Angeles Alonso, Roberto Ferrari, Michel Komajda, Luigi Tavazzi (Chairman), David Wood, Panos Vardas, Presidents of the ESC Associations.

Steering Committee

Josep Brugada (Chairman), Georges Mairesse (Belgium), Milos Taborsky (Czech Republic), Josef Kautzner (Czech Republic), Thorsten Lewalter (Germany), Sam Riahi (Denmark), Pierre Jais (France), Frédéric Anselme (France), George Theodorakis (Greece), Giuseppe Inama (Italy), Serge Trines (The Netherlands), Zbigniew Kalarus (Poland), Julian Perez Villacastin (Spain).

EURObservational Research Program Department

Aldo P. Maggioni, Malika Manini, Gérard Gracia, Cécile Laroche, Viviane Missiamenou, Charles Taylor, Marème Konte, Emanuela Fiorucci, Elin Folkesson Lefrancq, Myriam Glémot, Patti-Ann McNeill, Timothée Bois.

Participating centres

BELGIUM Leuven: H. Heidbüchel, D. Nuyens, Liege: J. Boland, V. Dinraths, J.-M. Herzet, E. Hoffer, D. Malmendier, M. Massoz, S. Pourbaix, Yvoir: E. Ballant, D. Blommaert, O. Deceuninck, F. Dormal, O. Xhaet, Aalst: T. De Potter, P. Geelen, Brugge: K. Derycker, M. Duytschaever, R. Tavernier, Y. Vandekerckhove, D. Vankats; CZECH REPUBLIC Ceske Budejovice: A. Bulava, J. Hanis, D. Sitek, Prague 4: M. Blahova, R. Cihak, L. Hanyasova, H. Jansova, P. Peichl, M. Tanzerova, D. Wichterle, Hradec Kralove: J. Duda, L. Haman, P. Parizek, Prague: L. Coling, P. Neuzil, J. Petru, L. Sediva, J. Skoda, Trinec: J. Chovancik, M. Fiala, R. Neuwirth; DENMARK Copenhagen: A. Karlsdottir, S. Pehrson, Aarhus N: C. Gerdes, H.K. Jensen, P. Lukac, J. C. Nielsen, Hellerup: J. Hansen, A. Johannessen, Varde: P. S. Hansen, A.K. Pedersen, Aalborg: F.P. Heath, S. Hjortshoj, A.M. Thogersen; FRANCE Saint-Etienne: A. Da Costa, I. Martel, C. Romeyer-Bouchard, N. Sadki, A. Schmid, Pessac: M. Haissaguerre, M. Hocini, S. Knecht, F. Sacher, Neuilly sur Seine: M. Ait Said, B. Cauchemez, F. Ledoux, O. Thomas, Nantes: J.-P. Cebron, N. Decarsin, D. Gras, S. Hervouet, Villeurbanne: C. Durand, A. Durand-Dubief, H. Poty, Tours: D. Babuty, B. Pierre, Toulouse: J.-P. Albenque, S. Boveda, N. Combes, R. Mas, Amiens: J-S. Hermida, M. Kubala, Rouen: B. Godin, A. Savouré, Y. Soublin, Grenoble: P. Defaye, P. Jacon, Lille: F. Brigadeau, S. Corbut, F. Flament-Balzola, S. Kacet, D. Klug, D. Lacroix, Saint Denis: X. Copie, L. Gilles, Z. Hocine, O. Paziaud, O. Piot, Rennes: C. Crocq, G. Kaballu, V. Le Moal, P. Lotton, P. Mabo, D. Pavin, Vandoeuvre-les-Nancy: M. Andronache, C. De Chillou, I. Magnin-Poull, Marseille: J.-C. Deharo, C. Durand, F. Franceschi, E. Peyrouse, S. Prevot, Paris: M. Etchegoin, F. Extramiana, A. Leenhardt, A. Messali; GERMANY Leipzig: T. Heine, A. Schneider, N. Winter, Coburg: J. Brachmann, G. Ritscher, B. Schertel-Gruenler, H. Simon, A.-M. Sinha, O. Turschner, A. Wystrach, Bonn Venusberg: M. Stemberg, Hamburg: K.-H. Kuck, A. Metzner, R. Tilz, E. Wissner, Hamburg: K. Heitmann, S. Willems, Berlin: D. Andresen, S. Mueller, Bremen: M. Volkmer, Frankfurt: B. Schmidt; GREECE Athens: A. Kostopoulou, E. Livanis, V. Voudris, Athens: M. Efremidis, K. Letsas, S. Tsikrikas, Athens: E. Christoforatou, P. Ioannidis, A. Katsivas, S. Kourouklis, Athens: G. Andrikopoulos, I. Rassias, S. Tzeis, Thessaloniki: G. Dakos, S. Paraskevaidis, G. Stavropoulos, E. Theofilogiannakos, V.P. Vassilikos; ITALY Pisa (PI): M.G. Bongiorni, G. Zucchelli, Mestre (VE): A. Raviele, S. Themistoclakis, Ferrara (FE): C. Pratola, Castellanza (VA): M. Tritto, Milano (MI): P. Della Bella, P. Mazzone, Milan (MI): M. Moltrasio, C. Tondo, Roma (RM): L. Calo, L. De Luca, F. Guarracini, E. Lioy, Cotignola (RA): L. Dozza, E. Frigoli, L. Giannelli, C. Pappone, M. Saviano, G. Schiavina, G.G. Vicedomini, Varese (VA): R. De Ponti, L. A. Doni, R. Marazzi, J.A. Salerno-Uriarte, C. Tamborini, Torino (TO): M. Anselmino, F. Ferraris, F. Gaita, Mirano (VE): E. Bertaglia, G. Brandolino, F. Zoppo; THE NETHERLANDS Rotterdam: N. De Groot, P. Janse, L. Jordaens, Maastricht: L. Pison, Groningen: C. Roos, I. Van Gelder, Eindhoven: R. Manusama, A. Meijer, P. Van der Voort, Leiden: S. Trines, Marieke G. Compier; POLAND Szczecin: J. Kazmierczak, Z. Kornacewicz-Jach, M. Wielusinski, Warszawa: J. Baran, P. Kulakowski, Wroclaw: M. Dzidowski, A. Fuglewicz, K. Nowak, Zabrze: P. Pruszkowska-Skrzep, A. Wozniak, Katowice: S. Nowak, M. Trusz-Gluza; SPAIN Madrid: J. Almendral, F. Atienza, E. Castellanos, C. De Diego, M. Ortiz, Madrid: J. Moreno Planas, N. Perez Castellano, Madrid: J. Benezet, J. Farre Muncharaz, J.M. Rubio Campal, Madrid: A. Hernandez Madrid, R. Matia, Sevilla: E. Arana, A. Pedrote, Madrid: R. Cozar, R. Peinado, I. Valverde, Barcelona: E. Arbelo, A. Berruezo, N. Calvo, E. Guiu, S. Husseini, L. Mont Girbau.


  • Names of Committee Members and Investigators are reported in the Appendix.


View Abstract