Numerical results are expressed as mean ± standard deviation or median [interquartile range], according to the distribution of the variable. Continuous quantitative variables were compared using analysis of variance or the Kruskal-Wallis test. Qualitative variables were compared using the chi-square test. The relationships between the number of implantations and the devices implanted per million population and the total number of implantations and the number of implantations for primary prevention in each center were studied using linear regression models.

A total of 177 hospitals performing ICD implantations reported their data to the registry (169 in 2015). The data from the 177 hospitals are shown in Table 1; 124 forms were excluded due to errors in the center records; 97 hospitals were public centers. The total number of implanting centers, rate per million population, and total number by autonomous community according to the data sent to the registry are shown in Figure 1. During 2016, only 18 centers implanted ≥ 100 devices; 87, ≤ 10; and 37, only 1.

Figure 1.

Distribution of activity by autonomous community in 2016: number of implanting centers/rate per million population/total number of implantations. Average = 122 implantations/million population.

(0.22MB).

The total number of implantations (first implantations and replacements) in 2016 was 5673, more than in 2015 (total number, 5465). Because the Eucomed data18 showed a total number of devices of 6662, this figure represents 85% of the total. The total number of implantations reported to the registry and those estimated by Eucomed in the last 10 years are shown in Figure 2.

Figure 2.

Total number of implantations recorded and those estimated by the European Medical Technology Industry Association (Eucomed) from 2007 to 2016. ICD, implantable cardioverter-defibrillator.

(0.14MB).

The overall rate of recorded implantations was 122 per million population; according to the Eucomed data, the rate was 143 per million population. The change in the implantation rate per million population during the last 10 years according to the registry and Eucomed data is shown in Figure 3. Implantations reported per implanting center are shown in Table 1.

Figure 3.

Total number of implantations recorded per million population and those estimated by the European Medical Technology Industry Association (Eucomed) from 2007 to 2016. ICD, implantable cardioverter-defibrillator.

(0.12MB).

The implanting hospital was recorded in 96% of cases. Most implantations (5069, 89.3%) were performed in public health care centers.

This information was available in 4557 forms sent to the SEC (80.6%). There were 3044 first implantations, representing 66.8% of the total (71.8% in 2015, 72.6% in 2014, 68.8% in 2013, 69.4% in 2012, 70.2% in 2011, and 73.8% in 2010). The rate of first implantations per million population was 65.5 (75.1 in 2015, 79.0 in 2014, 63.8 in 2013, and 64.0 in 2012).

The mean age ± standard deviation (range) of patients undergoing an ICD implantation or replacement was 62.7 ± 13.4 years (6-90 years) in 2016, compared with 62.8 ± 13.3 (6-98) in 2015 and 61.8 ± 13.7 years (7-94 years) in 2014. The mean age of first implantation patients was 61.2 ± 13.4 years. Most patients were men, who represented 81.9% of all patients and 82.4% of first implantation patients.

The most frequent underlying cardiac condition in first implantation patients was ischemic heart disease (48.6%), followed by dilated cardiomyopathy (29.5%), hypertrophy (7.8%), primary conduction abnormalities (Brugada syndrome and long QT syndrome) (3.1%), valve diseases (2.5%), and arrhythmogenic dysplasias (1.3%) (Figure 4).

Figure 4.

Type of heart disease prompting implantation (first implantations, sole diagnosis).

(0.1MB).

Left ventricular ejection fraction was > 50% in 16.3% of all registry patients, 41% to 50% in 6.5%, 36% to 40% in 7.5%, 31% to 35% in 20.2%, and < 30% in 49.5% (Figure 5). A similar distribution was seen in patients who underwent ICD replacement. These data were recorded in 68.6% of the data collection sheets of the registry.

Figure 5.

Left ventricular ejection fraction of the registry patients (total and first implantations).

(0.07MB).

Most patients were in New York Heart Association (NYHA) functional class II (52.1%), followed by NYHA III (32.1%), NYHA I (14.1%), and IV (1.7%). For this parameter, the distribution was also similar between total implantations and first implantations (Figure 6), and these data were reported in 52.9% of the registry forms.

Figure 6.

New York Heart Association (NYHA) functional class of the patients in the registry (total and first implantation).

(0.07MB).

The baseline rhythm, reported in 45.9% of the patients, was largely sinus rhythm (80.4%), followed by atrial fibrillation (16.7%) and pacemaker rhythm (2.1%); the remaining patients had other rhythms (atrial flutter and other arrhythmias).

These data were contained in 74.9% of the registry forms. For first implantations, most patients had no documented clinical arrhythmias (59.5%), followed by those with sustained monomorphic ventricular tachycardia, nonsustained ventricular tachycardia, and ventricular fibrillation (16.2%, 12.5%, and 10.6%, respectively). In total, patients with no documented clinical arrhythmia comprised 56.3% (Figure 7). The differences in the type of arrhythmia between the first implantation group and the total were not significant, except for those with no arrhythmia. The most frequent clinical presentation in both the total implantation group and the first implantation patients (66.3% and 72.8% of completed responses) was asymptomatic, followed by syncope, cardiac arrest, and “other symptoms” (Figure 8).

Figure 7.

Distribution of arrhythmias prompting implantation (first implantations and total). NSVT, nonsustained ventricular tachycardia; PVT, polymorphic ventricular tachycardia; SMVT, sustained monomorphic ventricular tachycardia; VF, ventricular fibrillation. *P < .02.

(0.07MB).

Figure 8.

Clinical presentation of the arrhythmia in the registry patients (total and first implantations). SCD, sudden cardiac death.

(0.07MB).

Data on electrophysiological studies were available for 2398 first implantation patients (78.8%). Such studies were performed in only 214 patients (8.9%). Sustained monomorphic ventricular tachycardia was the most common induced arrhythmia (44.8%), followed by nonsustained ventricular tachycardia (15.5%), ventricular fibrillation (11.5%) and, to a lesser extent, others (5.2%). No arrhythmia was induced in 23.0% of the electrophysiological studies. Most of these studies were performed in patients with ischemic heart disease or dilated cardiomyopathy.

Data on the clinical history of patients have only been available since 2011 because such data were not recorded in previous years.

Responses to questions on clinical history were obtained for between 66.3% and 82.5% of first implantation patients. The most important findings related to cardiovascular risk factors and history were as follows: hypertension, 56.0%; hypercholesterolemia, 46.2%; smoking, 37.4%; diabetes mellitus, 30.8%; history of atrial fibrillation and family members with sudden cardiac death, 24.8%; kidney failure, 14.0%; and stroke, 9.0%.

The QRS width was recorded in 52.7% of the forms (mean, 125 ± 33ms). In 35.2% of the patients, the recorded width was > 140ms. Of these patients, 60.5% of the first implantation patients and 60.4% of the total had a defibrillator-resynchronization device (ICD-CRT [cardiac resynchronization therapy]).

The changes in the type of heart disease and its presentation in first implantation patients from 2012 to 2016 are shown in Table 2. Data on this parameter were recorded in 53.3% of the registry collection sheets. For ischemic heart disease, the most frequent indication was primary prevention (49.9%), representing an increase from the previous year (48.9%) and from 2014 (45.5%). For dilated cardiomyopathy, the main indication was also prophylactic (63.5% vs 47.6% in 2015 and 47.0% in 2014). For less common heart diseases, the most frequent indications were primary prevention of hypertrophic cardiomyopathy, valve diseases, congenital diseases, and Brugada syndrome. For long QT syndrome, prophylactic implantation was the most common indication (45.5%), similar to the data from 2015 (54.5%) and in contrast to the data from 2014 (18.5%).

The implantation indication was reported in 80.4% of the records. Most first implantations were indicated for primary prevention (62%), an increase from 2015. This variability has been growing and was statistically significant until 2008, between 2009 and 2010, and between 2013 and 2014; differences were not significant between 2014 and 2015 but became significant again between 2015 and 2016 (Table 3).

The response rate to these questions was 86.7%. In 82.1%, the main implantation location was the electrophysiology laboratory (83.2% in 2015, 83.4% in 2014, 79.8% in 2013, and 81.4% in 2012), followed by the operating room (15.2%). Electrophysiologists performed 81.2% of implantations (79.6% in 2015, 81.7% in 2014, 80.7% in 2013, and 81.0% in 2012); surgeons, 5.14% (9.6% in 2015, 11.0% in 2014, 13.8% in 2013, and 14.0% in 2012); and both specialist types, 5.1%. Other specialists and intensivists were involved in 1.6% and 2.1%, respectively.

Information on the placement of first implants was provided in 2446 registry forms (80.3%). Placement was subcutaneous in 95.8% of patients and subpectoral in the remaining 4.2%. The figures were 95.1% and 4.9% for all devices implanted, respectively.

This information was available in 88.3% of the records and is summarized in Table 4. Subcutaneous defibrillator devices represented 6.4% of first implants (2.4% in 2015).

In patients with ischemic heart disease, 68.3% of implants were single- or dual-chamber devices (71.5% in 2015, 76.8% in 2014, 74.8% in 2013, and 72.3% in 2012) and 26.9% were ICD-CRTs (27.7% in 2015, 23.1% in 2014, 25.5% in 2013, and 27.7% in 2012). In patients with dilated cardiomyopathy, ICD-CRT devices comprised 44.3% (55.6% in 2015, 53.7% in 2014, 51.7% in 2013, and 56.5% in 2012).

Of the 1513 replacements, information was available on 1335 (89.6%). The most frequent reason for replacement was battery depletion (78.7%); complications motivated 8.8% (8.6% in 2015 and 7.9% in 2013) and a change of indication prompted 12.5%. Of the 93 replacements due to a change of indication that provided this information, 9.7% were required before 6 months (10.2% in 2015, 9.62% in 2014, and 11.6% in 2013).

Information was available on the status of the leads in 62.5% of the replacements; 8.1% were malfunctioning (54 records) and they were extracted in 14.8% of the patients reporting this problem.

Information on this parameter was provided in 73.9% of records. The most widely used programming was VVI (51.8%), followed by DDD (32.0%), VVIR (8.4%), DDDR (5.1%), and other modes (2.7%), largely algorithms to prevent ventricular pacing.

Induction of ventricular fibrillation was tested in 232 patients, with 4.1% of the 472 records providing this information (2.7% in 2015, 2.9% in 2014, 5.1% in 2013, and 6.7% in 2012). The mean threshold was 22.1± 7.6 J (23.6 ± 8.9 in 2015, 19.7 ± 6.8 in 2014, 20.4 ± 6.5 in 2013, and 20.5 ± 7.1 in 2012) and the mean number of shocks was 1.1.

With a response rate of 85.3%, 33 complications were reported: 10 coronary sinus dissections, 6 tamponades, 3 pneumothoraces, 1 death, and 13 unspecified complications. The mortality rate was 0.02%, which, with only 1 death, was lower than that of the previous 2 years (0.07% and 0.05%).

The Spanish Implantable Cardioverter-defibrillator Registry was first published in 2005 with the results of the 2002 to 2004 period.4 The number of implanted ICDs increased each year until 2010,5–10 with 2011 and 2012 then showing a lower total number of implantations in both the registry11,12 and Eucomed data. From 2013,13–15 there has been slight uptick in the number of implantations and, this year, they exceeded the values obtained in 2010 (Figure 2 and Figure 3). In 2016, the number of implanted devices increased again. A continual increase was also seen in Europe, in both the number of ICDs and the number of ICD-CRTs.18

The current registry data show an increase in the indication for primary prevention, which has exceeded 60% for the first time (Table 3). The increase has been steady, except for a marked decrease in 2013. In the last 10 years, the indication has increased by 10%.

For the first time, there was a slight decrease in the percentage of first implantations of ICD-CRTs (32.1% vs 33.9% in 2015). In 2016, due to the increased indication for subcutaneous ICDs (6.4% in 2015 vs 2.4% in 2015), the percentage of single-chamber ICDs increased, which reversed the recent trend for a reduction (54.8% vs 52.8% in 2015 and 53.4% in 2014). The use of dual-chamber ICDs decreased (13% in 2016 vs 13.2% in 2015 and 16.3% in 2014). The resynchronization rate for total implantations has slowly increased in recent years, due this year to replacements, and no major changes are expected in the coming years.

The most frequent indication in 2016 continued to be ischemic heart disease (48.6%), followed by dilated cardiomyopathy (29.5%). In contrast to previous years,13–15 when more than half of devices implanted in patients with dilated cardiomyopathy were ICD-CRTs, this year has seen a marked decrease in the percentage of such implantations (44.3% in 2016 vs 55.6% in 2015). In patients with dilated cardiomyopathy, the incidence of ICD-CRT devices was similar (26.9% in 2016 and 27.7% in 2015). These data explain the decrease in the percentage of ICD-CRT devices as first implantations observed in this registry.

The progressive increase in the number of ICD implantations stopped in 2011 and 2012. The 2013 results showed a slight recovery, with the total number of implantations slightly surpassing the rate per million population of 2010 (102 vs 100).13 This increase was confirmed by the data from 2013 onward, with an implantation rate of 122 in our registry and of 143 according to the Eucomed data.18 In 2010, the implantation rate in Spain was about half of the European rate (116 vs 248); in subsequent years, the gap has increased and the values are now 144 vs 320 per million.18

No recent studies have modified the ICD implantation indications. In 2002, the Multicenter Automatic Defibrillator Implantation Trial II19 study was published, followed by the Comparison of Medical Therapy, Pacing and Defibrillation in Heart Failure20 in 2005 and the Sudden Cardiac Death in Heart Failure Trial21 in 2006. These studies established the current indications in primary prevention and CRT and triggered a progressive increase in the number of implantations during that decade. The indications for ICD and CRT implantation are well supported in clinical practice guidelines.22–28 However, the implantation rate per million population does not correspond with that expected from the clinical evidence, both in Spain and in other European countries,29 a tendency that has become consolidated with time in these countries. According to the results of the current registry, it appears that the findings of the Defibrillator Implantation in Patients with Nonischemic Systolic Heart Failure study30 have still not affected the indications for ICDs in nonischemic patients. As in previous registries, the 2016 registry represents 85% of the implantations reported to the Eucomed, exactly the same as in 2015. The number of implanting centers participating in the registry has yet to reach 100%, and there were certain differences in the data with respect to those published in Eucomed that could be explained by sending and processing errors.

The number of implanting centers slightly increased from 2015. Two hospitals reported more than 200 implantations, 15 hospitals (11 in 2015 and 11 in 2014) reported more than 100 implantations, and 86 centers, mainly private, reported less than 10 implantations. The data show a tendency for an increase in implanting centers with low activity. Some studies have shown an inverse relationship between the implantation volume and the number of complications.31

There were no changes from previous registries in the epidemiological characteristics of the patients. Patients with severe ventricular dysfunction and in NYHA II and III continue to predominate. There were no differences from 2015 in the implantation setting: 82.1% were performed in the electrophysiology laboratory (83.2% in 2015 and 83.4% in 2014) and 81.2% were performed by electrophysiologists (vs 79.6% in 2015 and 81.7% in 2014).

Differences remain among autonomous communities. The implantation rate was 122 per million population according to the registry and 143 according to the Eucomed data; both databases showed an increase from 2015 (118 and 138, respectively). Several autonomous communities showed higher rates than the average: Principality of Asturias (173 implantations per million), Extremadura (166), Aragon (145), Castile and León (142), Cantabria (139), Castile-La Mancha (137), the Valencian Community (133), Community of Madrid (127), Andalusia (124), and Galicia (123). The following were below the average: Chartered Community of Navarre (80), La Rioja (80), the Region of Murcia (81), the Canary Islands (88), the Balearic Islands (86), the Basque Country (95), and Catalonia (102). The data per community were similar to those of previous years, except in the case of Andalusia, which, for the first time, was slightly higher than the average. The difference between the communities with the highest and lowest implantation rates is currently more than double (173 vs 80) and is similar to that seen in the previous registry (167 vs 67). In general, all autonomous communities have increased their implantation rate, except Madrid, the Valencian Community, Navarre, Catalonia, and the Canary Islands, which showed slight decreases, particularly the Canary Islands (88 in 2016 and 105 in 2015) and Navarre (80 in 2016 and 93 in 2015).

The trends observed in previous registries have been consolidated in relation to the slow and progressive increase in the number of implantations per million individuals, but also the differences among communities, given that, as seen previously, there is no correlation with gross domestic product because the communities above the average implantation rate are the least populated, except in the case of the Community of Madrid, the Valencian Community, and Andalusia. Of course, there are other possible explanations for these differences, such as the health care organization of each community (the number of arrhythmia units and the distribution of referral hospitals). There was also no relationship between the incidence of ischemic heart disease and heart failure in the various communities. Other possible explanations include the health care organization of each community, the number of arrhythmia units, and the distribution of the referral hospitals.

When ICDs and ICD-CRTs were included, the implantation rate in the countries participating in the Eucomed was 320 per million population (315 in 2015). Germany, with 550 devices per million population, remained the country with the highest number of implantations. Spain (144 implantations per million) was the country with the lowest number of implantations. Several countries showed higher than average rates: Italy (414), the Netherlands (372), the Czech Republic (355), and Poland (322). Below the average were Denmark (302), Austria (291), Ireland (280), Belgium (242), Sweden (238), Norway (244), France (222), Switzerland (227), Finland (261), the United Kingdom (245), Portugal (183), Greece (186), and Spain (144). The difference in the implantation rate in Spain from the European mean was maintained in 2016 (144 vs 320 and 138 vs 315 in 2015). The difference between Spain and the second-last country increased (144 vs 183).

The ICD implantation rate was 191 per million population in 2016 (189 in 2015). Germany (333 implantations per million population) had the highest number of implantations, whereas Spain (90) had the lowest.

The ICD-CRT implantation rate was 129 per million population (126 in 2015). Germany (217 implantations per million population) continued to be first, whereas Spain (54) had the lowest implantation rate.

The proportion of ICD-CRTs with respect to the total varies from 29% in Ireland and Poland to 46% in the United Kingdom. The European average is 40%. Above the average are France, Austria, the United Kingdom, Italy, and the Czech Republic. Ireland and Poland are below 30%. Spain has a proportion of 37%.

These countries have the same regional differences32,33 seen in the Spanish registry, for unknown reasons. One possible explanation is the number of available arrhythmia units, but that does not explain the relationship, at least in Spain, because communities with the highest number of available units had lower implantation rates. Other explanations, such as per capita income, also fail to show a correlation, with countries such as Ireland, the Czech Republic, and Poland showing much higher implantation rates than Spain. The prevalence of cardiovascular diseases, access to the health care system and its organization, and degree of acceptance of and adherence to the clinical practice guidelines could be related to the rate of implantations in Spain and its variability.

The registry included 85% of implantations performed in Spain according to the Eucomed data, the same percentage as in 2015. The percentage has slightly decreased from 2007, when the representativeness was 90%. The number of recorded implantations continues to accurately reflect the situation in Spain. The number of participating centers has remained practically unchanged from previous years.

The true number of implantations in some hospitals differs from that reported to the registry, given that the registry only includes received data collection sheets. Because data can now be sent in various ways, some sheets were not received or correctly registered. A transition was expected in 2016 because the registry data were to be collected in 2 ways, on paper and via the Internet. Unfortunately, the project allowing data to be collected via the Internet remains to be implemented, despite already having been developed. We hope that in 2017 all data will be collected via the website, which should improve the results and minimize the differences between the data obtained and those provided by the Eucomed.

There is excessive variability in the percentage of responses to the various questions in the ICD registry sheet, ranging from 96% for the implanting hospital to 45.9% for baseline rhythm. Finally, the percentage of complications reported to the registry fails to reflect reality because these data are provided during or immediately after the implantation, meaning that most subacute complications are not recorded. This difference could be accentuated in the case of outpatient defibrillator implantation, which is becoming more and more common in Spain.34

This registry is the 13th official report. The durability of this registry is a credit to all of the participating members of the Section of Electrophysiology and Arrhythmias of the SEC. The continued modernization of the registry will allow more and better information to be obtained with less effort by the staff involved in its maintenance. The quality of the information will improve with further computerization of the registry, and the completion of certain fields will be obligatory. In the future, the system may permit more ambitious clinical objectives and include parameters such as death, shocks, and complications that provide relevant clinical information.