OUP user menu

Implantable cardioverter–defibrillators in France: practices and regional variability

Arnaud Lazarus, Nelly Biondi, Jean-François Thébaut, Isabelle Durand-Zaleski, Michel Chauvin
DOI: http://dx.doi.org/10.1093/europace/eur238 1568-1573 First published online: 22 July 2011


Aims The rates of cardiac defibrillator implantation (ICD) in Europe and within countries are heterogeneous. We examined the characteristics of ICD recipients and device implantation rates in France, with the goal of identifying patterns and regional differences in medical practices.

Methods and results We compiled the information available on devices and patients from a manufacturer's database for the years 2008 and 2009 and the Stidefix national registry from January 2008 to April 2010 and reported the descriptive statistics and comparisons of implantation practices among regions. We analysed data from 10 766 patients enrolled in Stidefix (mean age=63 ± 13 years; 83% men) after implantation of single-chamber (24%), dual-chamber (33%), or triple-chamber (43%) ICD. Implantation was for primary prevention in 63% of patients. Between 2008 and 2009, the national implant rate increased from 126 to 147 per million inhabitants, with regional variations related to the number of cardiologists and of implant centres. Patients were significantly older and more likely to undergo implantation for primary prevention or for cardiac resynchronization therapy (CRT) in higher-volume regions. In Ile de France (Paris and its suburbs), patients tended to be younger, as in low-implantation regions, but with a high rate of CRT implants, as in high-implantation regions. A comparison of the Stidefix data with the manufacturer's database revealed that only 57% of all ICD implanted in 2009 were reported to Stidefix.

Conclusion Despite an increasing rate of ICD implantation in France, important regional disparities persist, with a median position occupied by Ile de France.

  • Implantable cardioverter–defibrillator
  • Medical device registry
  • Ventricular tachyarrhythmia
  • Sudden cardiac death
  • Public health


In clinical use for over 30 years, the implantable cardioverter–defibrillator (ICD) is a validated therapy for major ventricular tachyarrhythmias, as well as for prevention in patients at high risk of sudden cardiac death. Several randomized clinical trials14 have shown that the ICD lowers the mortality of patients with a history of sustained ventricular arrhythmias compared with antiarrhythmic drug therapy. Subsequent trials have shown a survival benefit conferred by ICDs preventively implanted in subgroups of patients at risk of sudden cardiac death.1,510

In France, ICDs have been reimbursed since 2005 in a limited number of centres accredited on the basis of an activity target of >50 implants per year. The French health authorities have mandated the organization of a national registry recording all implantation procedures. Although it is mandatory, compliance with this registry is not enforced. The present report presents the French ICD activity, patient characteristics, practices, and factors that might be responsible for regional differences in implantation rates.


The ICD activity in France can be obtained from two sources. Active since March 2007, Stidefix is a national prospective multicentre registry hosted by the French Society of Cardiology, which has three main objectives: (i) to respond to the legal mandate of the French health authorities requiring the enrolment of all new ICD implants in a national registry by the medical centres; (ii) to create a database enabling analysis of the French practices in the area of cardiac pacing and defibrillation; and (iii) to provide a computer-based tool to the implanting centres for managing implantations. A second source of data is the database maintained by the manufacturers, which includes all ICDs implanted from each manufacturer but does not include clinical information.

The Stidefix registry

The clinical data of the enrolled patients and the technical characteristics of the implanted ICDs are registered online in Stidefix by each centre. All ICD recipients should be prospectively enrolled in Stidefix after having obtained their informed consent, without exclusion criterion. The data are validated only by the implanting centre. Our study population exported from the Stidefix database includes all patients who received ICDs between 1 January 2008 and 13 April 2010. We excluded patients who underwent implants outside of mainland France, whose implantation date was uncertain, and/or whose records were not validated by the implanting centre. In case of multiple records, we retained the most recent ones for this analysis. The Stidefix registry includes medical information, indications for ICD implantation, and type of device implanted—single chamber (VR-D), dual chamber (DR-D), or resynchronization therapy (CRT-D, cardiac resynchronization therapy-D)—and distinguishes first implants from device replacements. Stidefix does not systematically collect data regarding survival, clinical outcomes, or complications.

Manufacturer's database

This database lists by centre the type of device implanted and the year of implantation but does not include medical data. The data on defibrillators implanted in years 2008 and 2009 were used as the reference for ICD implant activity in our analysis. A comparison between Stidefix and the total ICD activity supplied by the manufacturer's trade organization was used to assess the completeness of Stidefix for the study period.

Statistical analysis

Analyses of patient and device data were based on the Stidefix registry. Analyses of implantation activity were based on the manufacturer's database. The data were analysed at the national and regional levels. We compared regions with high- and low-implantation rates, as well as two groups of regions with the Ile de France region (Paris and its suburbs). Regions with implantation rates corresponding to the national average minus ≥35 implants or plus ≥35 implants per million inhabitants were defined as implanting, respectively, below or above the national average. The population-based data used to calculate these rates at the national and regional levels were obtained from estimates reported up to 1 January 2007 by the French National Institute for Statistics (INSEE).

Continuous variables are expressed as means ± standard deviation (SD), and categorical variables as counts and percentages. Percentages were based on the known total number of implants pertaining to the variable analysed. Student's t-test and analysis of variance were used to compare continuous variables, after verification of their normal distribution with the Kolmogorov–Smirnov test. Categorical variables were compared using the χ2 test. Potential correlations between continuous variables were tested using Pearson's correlation coefficient. A P value <0.05 was considered significant. The data were analysed using the SAS® version 9.1 for Windows statistical software (SAS Institute, Cary, NC, USA).


Out of 13 531 patients who underwent first ICD implants or replacements, as reported in the Stidefix registry since its creation in 2007, 10 766 from 66 medical centres were retained for analysis (Figure 1).

Figure 1

Flow chart of the selection of defibrillator recipients from the Stidefix registry.

National perspective

The baseline characteristics of the patients are summarized in Table 1. The mean age of the men was 64 ± 13 years and that of women was 61 ± 15 years; 34% of the sample was >70 years of age. The left ventricular ejection fraction (LVEF) ranged from 10 to 80%. The devices implanted were divided by type with 43% CRT-D, 33% DR-D, and 24% VR-D. The implant rate increased from 126 per million inhabitants in 2008 to 147 per million in 2009, with an increase in the percentage of primary prevention ICD indications from 61.8% in 2008 to 63.6% in 2009.

View this table:
Table 1

National baseline patient characteristics, and comparison of the baseline characteristics in Ile de France region vs. regions with <112 × 106 inhabitants (low implant rates) or >182 × 106 inhabitants (high implant rates) from the Stidefix database

Region analysedP
FranceIle de France (1)Low-implant rates (2)High-implant rates (3)(1) vs. (2)(1) vs. (3)(2) vs. (3)
Age, year63 ± 1361 ± 1561 ± 1365 ± 13ns<0.001<0.001
Men8885 (83)664 (82)1996 (82)3610 (83)nsnsns
Disease manifestations
 None502 (5)46 (6)95 (4)281 (8)nsns<0.001
 Syncope1913 (22)115 (16)383 (18)762 (22)ns<0.001<0.001
 Cardiac arrest890 (10)106 (14)225 (11)261 (8)0.005<0.001<0.001
 Others5314 (59)506 (69)1227 (58)1915 (56)<0.001<0.001ns
Underlying heart disease
 Unknown972 (9)33 (4)214 (9)575 (13)<0.001<0.001<0.001
 Ischaemic5605 (58)422 (54)1295 (59)2151 (57)0.022nsns
 Myocardial infarction2698 (28)218 (28)656 (30)849 (23)ns0.001<0.001
 Valvular438 (5)39 (5)69 (3)161 (4)0.016ns0.027
 Hypertension129 (1)18 (2)17 (1)40 (1)0.0010.005ns
 Congenital heart disease71 (1)9 (1)16 (1)16 (0)ns0.012ns
  Dilated3405 (32)344 (42)775 (32)1318 (31)<0.001<0.001ns
  Hypertrophic340 (3)25 (3)89 (4)114 (3)nsns0.018
  Arrhythmogenic right ventricular127 (1)18 (2)28 (1)38 (1)0.0430.003ns
 Isolated electrical abnormality371 (4)32 (4)104 (5)104 (3)ns0.045<0.001
 Others323 (3)23 (3)64 (3)91 (2)nsnsns
Left ventricular ejection fraction, %
 Mean31 ± 1231 ± 1330 ± 1231 ± 11nsns0.003
 ≤356738 (80)564 (80)1591 (81)2521 (82)nsnsns
NYHA functional class I or II4556 (56)376 (51)1262 (65)1566 (48)<0.001ns<0.001
Primary prevention indication6574 (63)512 (64)1431 (61)2767 (66)nsns<0.001
CRT-D4689 (43)412 (51)815 (34)2212 (51)<0.001ns<0.001
  • Values are means ± SD, or numbers (%) of observations; cumulative percentages may exceed 100% because of multiple disease manifestations or underlying diseases in a same patient.

  • CRT-D, cardiac resynchronization therapy–defibrillator; NYHA, New York Heart Association.

Effects of underlying heart disease

Patients who presented with underlying heart disease were older (64 ± 12 years) than patients with isolated electrical abnormalities (55 ± 17 years; P < 0.001), were more likely to have undergone ICD implant for prophylactic indications (65 vs. 45%, P < 0.001), and had a lower mean LVEF (30 ± 10 vs. 54 ± 15%; P < 0.001). Patients with isolated electrical abnormalities were more likely to present with syncope (41 vs. 20%; P < 0.001) or cardiac arrest (36 vs. 8%; P < 0.001), and to be in New York Heart Association (NYHA) functional class I or II (93 vs. 53%; P < 0.001).

Types of devices implanted

The differences in patients receiving VR-D or DR-D, compared with CRT-D were as follows: recipients of VR-D or DR-D were younger (61 ± 14 years) than recipients of CRT-D (66 ± 11 years; P < 0.001), were more often asymptomatic (6 vs. 4%; P < 0.001), more often presented with syncope (28 vs. 13%; P < 0.001) or cardiac arrest (15 vs. 3%; P < 0.001), and were more likely to present in NYHA functional classes I or II (85 vs. 21%; P < 0.001). Their mean LVEF was higher (35 ± 14% vs. 26 ± 6%; P < 0.001) and were less likely to undergo ICD implant for prophylactic indications (52%) than recipients of CRT-D (79%; P < 0.001).

Regional perspective

Figure 2 shows the regional and national rates of implants per million inhabitants in 2009, based on the manufacturer's data. Pearson's correlation showed a significant relationship between the number of ICDs implanted in 2009 and the number of cardiologists (63%; P = 0.002) and number of implanting centres per million inhabitants (51%; P = 0.022). Regions with implantation rates <112 per million inhabitants included Alsace, Basse-Normandie, Bourgogne, Bretagne, Champagne-Ardennes, Franche-Comté, and Picardie, while regions at or above this threshold included Aquitaine, Languedoc-Roussillon, Provence-Alpes-Cote-d'Azur, Corse, and Rhône-Alpes. The baseline characteristics of patients from the Ile de France region are compared with those of patients from regions with low and high implantation rates in Table 1.

Figure 2

Regional number of defibrillator implants per million inhabitants in 2009, based on data from the manufacturers.

Number of implants and completeness of the Stidefix registry

Table 2 shows the numbers of implanting centres and numbers of ICDs implanted per million inhabitants in years 2008 and 2009 in France and in Ile de France, according to the manufacturer's database and the percentages of implants entered in the Stidefix registry. In 2009, 32% of the 96 French accredited medical centres implanted <50 ICDs per year (low-volume centres), and 40% of centres implanted >100 ICDs per year (high-volume centres). In the Ile de France region, 39% of the 18 accredited centres remained below the 50 ICD implants per year threshold, while 39% were high-volume implanting centres. Between 2008 and 2009, the percentage of ICD implants recorded in Stidefix increased from 51 to 57%, and ICD data were entered in the registry by 61 of 88 centres (69%) in 2008 and 65 of 96 centres (68%) in 2009. Compared with the manufacturer's database, only 25% of all ICD implants from Ile de France were entered in the Stidefix registry in 2008 and 2009. This was not related to difficulties in obtaining the patient consent for the data to be entered but mainly due to the lack of participation by some teams or centres, despite the legal requirement to participate in the Stidefix registry.

View this table:
Table 2

Number of defibrillator implants and implant centres in 2008 and 2009, in Ile de France and in France, and completeness of Stidefix data compared with the manufacturer's database

Ile de FranceFrance
Manufacturer's database
 Medical centres18188896
 Medical centres/106 inhabitants1.
 ICD implants1292149079529309
 ICD/106 inhabitants110127126147
Stidefix registry
 % Completeness20295157
  • Values are numbers (manufacturers’ database) or percentages (Stidefix registry) of observations.

  • ICD, implantable cardioverter–defibrillator.


This survey, based on a national registry and on the manufacturer's sales database, provides a global overview of the consumption of ICDs and implantation activities in France and its various regions. The median age of French ICD recipients was 63 years, the majority were men, and one-third of patients were >70 years of age. LVEF was <35% (due to ischemic or dilated cardiomyopathy) in 80% of ICD recipients, and 44% were in NYHA functional classes III or IV. The main indication was primary prevention, which accounted for most of the growth in implant rates between 2008 and 2009. The patients who presented with underlying heart disease were more likely to undergo CRT-D and prophylactic ICD implantations. VR-Ds and DR-Ds tended to be implanted in younger patients in lower NYHA functional classes and with higher LVEFs, and these implantations were more often for purposes of secondary prevention.

Based on the manufacturer's data, major regional disparities were observed in ICD implantation rates in 2009, ranging from 72 to 211 per million inhabitants. The patients were significantly older in higher-rate regions and were significantly more likely to undergo CRT-D implants for primary prevention indications. Our analysis suggests that the higher ICD implant rate observed in some regions may be attributable to a more active management of elderly patients, as well as to a wider acceptance or better understanding of the role of ICDs in primary prevention and CRT.

The Ile de France region had the highest crude number of device implantations in France and a number of centres per million inhabitants identical to the national average. While the population of Ile de France is generally a high consumer of health care, the implant rate was uncharacteristically lower than the national average. The patterns of ICD implantation in Ile de France are paradoxical: patients are younger than the national average, as in other regions with low implantation rates, but CRT-D devices were implanted more frequently, as in high-implantation regions. These two findings may explain the median regional position of Ile de France, reflecting a reluctance to refer or implant elderly patients, especially in primary prevention where no immediate clinical benefit is usually obtained from VR-D/DR-D ICD therapy, and a possible preference for CRT-P (resynchronization pacemaker) instead of CRT-D in older patients. Resynchronization therapy is well accepted in Ile de France, given that patients can perceive a clinical improvement soon after implantation of a CRT-D, and that some Ile de France centres were among the pioneers of this therapy.

Regional differences have also been observed in other European and non-European countries.1116 The reasons behind variable clinical practices are widely debated and are not all well understood. A common explanation is a higher number of implanting centres found in regions with higher ICD implantation rates.11,12,16 We also found significant relationships between the numbers per million inhabitants of ICDs implanted, implanting centres, and cardiologists, highlighting the influence of the health care supply on the implantation of ICDs.

Other explanations for the variability in ICD implantation rates include variable awareness or adherence to practice guidelines,1719 under-utilization of ICDs in primary prevention due to insufficient proactive screening, competition for the time of electrophysiologists with complex ablation procedures reducing the time available for ICD implantation,17 and uncertainty regarding which patients are the best candidates for implantation of an ICD. These reasons are among the factors identified by the Heart Rhythm Association Group of the European Society of Cardiology to explain national variations.20 Differences in patient, physician and hospital characteristics, distribution of the cardiovascular diseases, or economic limitations placed on ICD implants19,21 may also account for these regional or national disparities.

Based on the manufacturer's database, the number of ICD implants in France increased from 126 to 147 per million inhabitants, consistent with the increase observed in recent years in other European and non-European countries.1116 Technological progress, broadening of the indications for ICD implantation to primary prevention, and a wider communication of these indications seem to be the main explanations for this trend. However, despite the increasing rate of ICD implants in France, it remains far below the European average of 234 per million inhabitants in 2009.22

Finally, the French ICD implantation practice did not fulfil the ministerial mandates on two grounds: one-third of centres had implantation rates below the lower legal limit for accreditation of 50 ICD implants per year; and just over half of all ICDs implanted in France were included in the national Stidefix registry in 2008–10. As a result, this national registry is not fully representative of the nationwide ICD population, a significant loss for the French medical community, since registries provide important, comparative, independent information on the consumption of devices, their indications, and the appropriateness of their implantation.17,23

Conflict of interest: The authors have no conflicts of interest to disclose.


This work was supported by a grant of the Union Régionale des Professionnels de Santé d'Ile de France and by the following members of the Syndicat National de l'Industrie des Technologies Médicales (SNITEM): Biotronik France, Boston Scientific France, Medtronic France, Saint Jude Medical France, and Sorin Group France.


We thank Michèle Salvador-Mazenq, MD and Michel Chauvin, MD, the co-founders of the Stidefix registry, Antoine Leenhardt, MD, Serge Cazeau, MD and Philippe Mabo, MD, the 2007–2011 Presidents of the Working Group on Rhythmology and Cardiac Pacing of the French Society of Cardiology, and Jean Yves Le Heuzey, MD, Patrick Assyag, MD, Elodie Deberne and Edwige Zouaghi-Packowski (URPS Ile de France), Geneviève Mulak (French Cardiology Society), Benjamin Cadier and Biljana Parapid, MD (URC Eco) for their assistance in data access and analysis. The authors are indebted to Karen Berg Brigham (URC Eco) for her edits and comments on the manuscript.


Medical centres implanting ICD in Ile de France, participants in the Stidefix registry (by decreasing number of records included in Stidefix): Clinique Bizet, Paris; Hôpital Pitié-Salpétrière, Paris; Hôpital Européen G. Pompidou, Paris; Hôpital Le Raincy-Montfermeil, Montfermeil; Hôpital Antoine Béclère, Clamart; Centre Cardiologique du Nord, Saint Denis; Hôpital Lariboisière, Paris; Hôpital Percy, Clamart; Clinique Ambroise Paré, Neuilly-sur-Seine.


View Abstract