OUP user menu

Safety and efficacy of strategic implantable cardioverter-defibrillator programming to reduce the shock delivery burden in a primary prevention patient population

Jonathan Buber, David Luria, Osnat Gurevitz, David Bar-Lev, Michael Eldar, Michael Glikson
DOI: http://dx.doi.org/10.1093/europace/eut302 227-234 First published online: 9 October 2013


Aims Strategically chosen ventricular tachycardia (VT)/ventricular fibrillation (VF) detection and therapy parameters aimed at reducing shock deliveries were proven effective in studies that utilized single manufacturer devices with a follow-up of up to 1 year. Whether these beneficiary effects can be generalized to additional manufacturers and be maintained for longer periods is to be determined. Our aim was to evaluate the durability and applicability of the programming of strategic implantable cardioverter-defibrillators (ICDs) of various manufacturers, which is aimed at reducing the shock delivery burden in primary prevention ICD recipients.

Methods and results A retrospective analysis of prospectively collected data of 300 ICD recipients of various manufacturers was conducted; 160 devices were strategically programmed to reduce shocks and 140 were not. The primary endpoint was the composite of death and appropriate shocks. Additional outcomes were inappropriate shocks, syncope events, and non-sustained VTs. At a median follow-up of 24 months, 19 patients died, 31 received appropriate shocks, and 41 received inappropriate shocks. Multivariate analysis showed that strategic programming dedicated to shock reduction was associated with a 64% risk reduction in the primary endpoint [hazard ratio (HR): 0.13–0.93; P = 0.03] and a 70% reduction in inappropriate shock deliveries (HR: 0.16–0.72; P = 0.01). Very few syncope events occurred (five patients, 1.6%), and there was no between-group difference in this outcome.

Conclusion Utilization of strategically chosen VT/VF detection and therapy parameters was found to be effective and safe in ICDs of various manufacturers at a median follow-up period of 2 years among primary prevention patients.

  • Implantable cardiac defibrillators
  • Inappropriate shocks
  • Strategic programming

What's new?

  • Strategic programming reduce shocks of any aetiology in implantable cardioverter-defibrillator recipients in real-life series.

  • This was true in devices of all manufacturers.

  • The reduction in shocks was not accompanied by a significant increase in syncopal events.

Implantable cardioverter-defibrillators (ICD's) reduce the risk of sudden cardiac death by terminating malignant ventricular arrhythmias13 and are currently considered a class I recommendation for primary prevention in many patients with a reduced ejection fraction.4 This beneficiary effect, however, may be offset by considerable adverse clinical and psychological events that result from the delivery of shocks.5,6 Moreover, ICD shocks may be pro-arrhythmic7 and thus cause an increase in mortality, regardless of the appropriateness of the delivered shock.810

In 2008, Wilkoff et al. reported that a programming strategy employing an increase in device detection time (i.e. 30–40 beats) of stable, monomorphic ventricular tachycardia (VT) and of preliminary anti-tachycardia pacing (ATP) therapy for VT at pre-specified cycle lengths (CLs) [i.e. threshold of 167 b.p.m. for monitoring only, of 182 b.p.m. of fast VT via ventricular fibrillation (VF), and of 250 b.p.m. of VF] reduced morbidity, spontaneous shock episodes (appropriate and inappropriate), arrhythmic syncope, and untreated sustained symptomatic VT events in primary prevention patients.11 These outcomes were evaluated in patients that received ICD devices of a single manufacturer and the follow-up period was 12 months. Based on these results and those of additional, positive outcome trials,12,13 ICD programming strategies to reduce the shock delivery burden have been increasingly utilized by many practicing groups, yet there are limited data on the intermediate and long-term outcome of such practice. The aims of this study were therefore to evaluate the effectiveness as well as safety outcomes associated with utilization of a programming strategy aimed at reducing the shock delivery burden in a primary prevention population implanted with devices of various manufacturers as compared with a more traditional programming by physician preference.


Study patients

All the patients included in this study were implanted with an ICD device for primary prevention of sudden cardiac death due to a severely reduced (i.e. <35%) left ventricular ejection fraction in accordance with the accepted guidelines at the time of implantation. The study cohort comprised of two groups of patients: the first group consisted of patients whose devices were strategically programmed to reduce the incidence of inappropriate shocks (the ‘strategic programming’ group, see below) and the second, of patients whose devices were programmed in a manner considered suitable by the implanting physician and that were yet different from the strategic programming protocols (the ‘control group’). Data of patient's baseline demographics, co-morbidities, device implantation, and programming parameters were prospectively collected for the institution's ICD and pacemaker registry. Thereafter, data from this general registry were obtained and analysed. Strategic device programming was introduced to our electrophysiology department in 2005 and added an additional programming option for the implanting physicians. Thus, data collection for this study was performed for patients implanted between the years 2005 and 2010. Programming selection was based on the discretion of the implanting physician; no attempts were made to affect the physician's choice of programming parameters. Owing to the change in the guidelines for the appropriateness of device implantation during the time period of the data collection, more patients implanted during the earlier years underwent an electrophysiological study prior to the implantation procedure. Informed consent was provided by all the patients prior to the implantation procedure and to the collection of the data. Only recipients of ICDs equipped with intracardiac electrogram storage were included in the current analysis.

Device implantation and programming

All defibrillator systems were implanted in the pectoral region. During the implant procedures, sensing and pacing thresholds were tested. Ventricular fibrillation induction was left to the discretion of the implanting physician. The systems used were manufactured by Biotronik, Medtronic, Boston Scientific, and St Jude Medical. Defibrillators in the strategic programming group were programmed according to the PREPARE settings in Medtronic devices and settings as similar as possible to other manufacturers' devices (Table 1). The main parameters included ventricular arrhythmia faster than 150 b.p.m., monitored by the device without consequent therapy (Zone 1). Ventricular arrhythmias faster than 188 b.p.m. were initially treated with two bursts of ATP and, if unsuccessful, terminated by defibrillator shocks (Zone 2).

View this table:
Table 1

Company-specific programme parameters to minimize shock risk in primary prevention

ManufacturerDetectionRateBeats to detection durationEnhancementTherapy
Medtronic©VT (monitor)167 b.p.m.32 intervalsAF/AFL onOff
 FVT via VF182 b.p.m.30 of 40 intervalsST (1 : 1 VT/ST = 66%) on wavelet (match = 70%) SVT Limit = 300 msBurst ×1, 30–35 J ×5
VF250 b.p.m.30 of 40 intervals30–35 J ×6
Boston Scientific©VT-1 (monitor)165 b.p.m.9.0 sMonitor onlyOff
VT180 b.p.m.7.0 sRhythm ID onBurst ×1, 41 J ×6
VF200 b.p.m.7.0 sNot applicable41 J shocks ×8a
St Jude©VT-1 (monitor)166 b.p.m.30 intervalsOff (monitor)
 VT-2181 b.p.m.30 intervalsV < A if ALLb
Internal stability
V = A if ANY morphology SVT limit = 300 ms
ATP, 30 J, 40 J ×3
VF240 b.p.m.30 intervals30 J, 40 J ×5a
Biotronik©VT1171 b.p.m.32 intervalsSMART ONoff
 VT219030 intervalsSMART ONBurst ×2, 40 J ×8
VF2409 of 13 intervalsNot applicable40 J ×8
  • aATP burst = 8 intervals, at 88% CL.

  • bExtended duration timers turned off.

  • VT, ventricular tachycardia; b.p.m., beats per minute; AF, atrial fibrillation; AFL, atrial flutter; ST, sinus tachycardia; FVT, fast ventricular tachycardia; VF, ventricular fibrillation; J, joules.

In the case of a ventricular arrhythmia faster than 240 b.p.m., high output (i.e −30 to −35 J) device shocks were the initial therapy (Zone 3). Supraventricular tachycardia discriminators were enabled up to rates of 200 b.p.m. In all the devices, stability and sudden onset algorithms were activated to reduce the occurrence of inappropriate shocks. Moreover, additional discriminators were activated in dual-chamber ICDs and cardiac resynchronization therapy defibrillators (CRT-D). The settings were adapted only when clinically indicated (i.e. haemodynamic well-tolerated VT at high rate, VT in the monitor zone). Programming protocols were available to the physicians implanting ICD devices via both written protocols and case simulations provided by all the device manufacturers.

Devices in the control group were programmed by physicians' preference that would typically include two (45% of the patients belonging to this group) or three zones (55%), with the lower zone starting at 150 b.p.m. with nominal durations of detection, treated with several (6–8, mean: 7 ± 2) bursts of ATP in the lower VT zone, then low-energy shock was followed by high-energy shock. Fast VT or VF zone were programmed to start above 165 b.p.m. in all patients (mean: 170 ± 15 b.p.m.), with nominal detection durations, one or two ATP's in the fast VT zone followed by high (30–35 J) energy shocks.


Follow-up visits were regularly conducted every 6 months at our outpatient clinics. In addition, several patients were followed with remote monitoring systems. During the follow-up visits, the patients were questioned regarding the occurrence of pre-specified symptoms including syncope, presyncope, palpitations, and lightheadedness. In addition, device memory was interrogated for delivered therapy (appropriate or inappropriate). Adjudication of the delivered therapy was performed by a trained electrophysiologist. An inappropriate shock was defined as a shock episode not delivered for VT or VF. Multiple inappropriate shocks in the same episode were counted as a single episode. In addition, the cause of an inappropriate shock was categorized into supraventricular tachycardia (SVT, including atrial fibrillation), sinus tachycardia, or abnormal sensing. Non-sustained VT events were defined as monomorphous tachycardias of ventricular origin ranging from 4 to 30 beats.

Data of interval device interrogations obtained at clinic visits, as well as printouts of interrogations performed at the time of arrhythmic events and shock deliveries were obtained from patients' record files and evaluated retrospectively. In the case that the patients were admitted to institutions other than ours at the time of arrhythmic event occurrence, all the data of hospitalization records were obtained, including hospitalization records and device interrogations.

Patients with data missing for longer than 6 months were considered lost to follow-up.

Study endpoints

The primary endpoint of this study was the occurrence of death from any cause or appropriate shock delivery, whichever came first. Death was included in the primary endpoint to avoid a possible bias of censoring fatal arrhythmic events, which were not interrogated. Secondary endpoints included inappropriate shock delivery, true syncope events, and untreated non-sustained VT events recorded by the devices.

Statistical analysis

The clinical and device associated parameters of study subjects stratified by programming parameters (strategic vs. control) were compared using χ2 tests for categorical variables and t-tests or Mann–Whitney–Wilcoxon tests for continuous variables. The cumulative probabilities for the occurrence of death or appropriate device shocks, of inappropriate device shocks, and of untreated non-sustained VT events during a 2-year follow-up by programming strategy were graphically displayed according to the method of Kaplan and Meier, with comparison of cumulative events by the log-rank test in which patient death and appropriate device shocks were considered censoring events. Multivariate analysis for the endpoint of death or appropriate shock delivery was carried out using Cox proportional hazards regression modelling. Pre-specified covariates in the multivariate models included strategic programming, non-ischaemic aetiology of heart failure, left ventricular ejection fraction (per 5% increment), atrial fibrillation, beta-blocker therapy, and device type.

Causes of inappropriate shocks were examined for the two programming groups with further division for different device types, and compared using the χ2 test. Predictors of inappropriate shocks were determined by the method of Cox proportional hazards regression. First, univariate analysis was performed, containing all the baseline variables. Subsequently, all variables with a P<0.10 were included in the multivariate analysis.

All P values were two-sided, and a P < 0.05 was considered significant.

Analyses were conducted with SAS software (version 9.2, SAS institute).


Complete follow-up data were available for 300 primary prevention patients, who received an ICD between January 2005 and March 2010. Of the 300 patients, 160 (53%) received an ICD device strategically programmed to reduce the shock delivery burden, and 140 patients belonged to the control group. Complete data were not available for 19 additional patients who were lost to follow-up. The median follow-up time was 24 months (interquartile range, 13–48 months). The baseline characteristics of the patients in the two groups are summarized in Table 2. An electrophysiological study was more commonly performed among patients in the control group (43 vs. 21%, P < 0.001) and more patients from this group received CRT-D device (41 vs. 27%). Conversely, more patients in the strategically programmed ICD group were regularly treated with beta-blockers (91 vs. 74%, P = 0.01). Other antiarrhythmic therapy did not differ between the groups, with a borderline difference in utilization of amiodarone, which was more frequently used in the physician tailored programming group (P = 0.08). Even distribution of manufacturers of devices was found between the groups.

View this table:
Table 2

Baseline characteristics of patient populationa

Physician tailored programming
N = 140
Strategic programming N = 160P value
Male124 (89)144 (90)0.85
Hypertension81 (58)94 (59)0.87
Smoking38 (24)43 (27)0.72
EPS before implantation60 (43)34 (21)<0.001
NYHA class0.77
 NYHA 1 + 269 (49)83 (52)
 NYHA 3 + 471 (51)77 (48)
 Atrial fibrillation12 (9)15 (24)0.06
 Antiarrhythmic treatment24 (18)21 (13)0.1
 Beta-blocker treatment104 (74)145 (91)0.01
 Ischaemic heart failure123 (88)126 (79)0.02
 Ejection fraction2562660.61
Device characteristics0.03
 Single chamber35 (25)56 (35)
 Dual chamber48 (34)61 (38)
 CRT-D41 (57)43 (27)
Years of device implantation0.09
 2005–0660 (43)51 (32)
 2007–0841 (29)54 (33)
 2009–1039 (28)55 (34)
Device manufacturer0.85
 Medtronic51 (36)49 (30)
 Biotronic26 (18)34 (21)
 Boston Scientific40 (28)48 (30)
 St Jude Medical33 (24)29 (18)
  • aValues mean ± SD or n (%).

  • Percents may not total 100%, because of rounding.

  • EPS, electrophysiological study; NYHA, New York Heart Association; CRT-D, cardiac resynchronization therapy with a defibrillator.

Primary endpoint

A total of 19 patients died during follow-up (6.3% of the entire study population). Of these, 12 patients belonged to the control group (8.6% control) and 7 to the strategically programmed group (4%, P = 0.09). Thirty-one patients received an appropriately delivered device shock, 25 of whom belonged to the control group and 6 to the strategically programmed group (P = 0.02). All the shocks were appropriately delivered at heart rates that were within the VF treatment zones in the strategic programming group, while six shocks were delivered in the VT via VF zone in the control group, and the rest were delivered in the VF zone. The tachycardia CL was below 280 ms in all patients belonging to the control group (mean, 264 + 31 ms), and below 310 ms in all patients belonging to the strategic programming group (mean, 294 + 29 ms). Kaplan–Meier estimates of event-free in the composite endpoint in the two groups are shown in Figure 1. At 24-month follow-up, the cumulative probability of survival was 81% in patients in the control group, compared with 95% in the strategically programmed group (log-rank P = 0.03 for the overall difference in event rates during follow-up). In accordance with these findings, multivariate Cox proportional hazards regression analysis showed that ICD strategic programming to reduce the shock burden was independently associated with a 64% reduction in the risk for the combined primary endpoint [95% confidence interval (CI): 0.13–0.93; P value = 0.03, Table 3]. In addition, a non-ischaemic heart failure aetiology was also found to be independently associated with a reduction in the combined endpoint [hazard ratio (HR): 0.46; 95% CI: 0.21–0.96, P = 0.04]. Of note, no difference in the occurrence of the composite endpoint was demonstrated between patients with devices of different manufacturers.

View this table:
Table 3

Multivariate analysis: the effect of strategic ICD programming on death or appropriate shocksa

CovariateEndpoint: ICD shock or death
HR95% CIP value
Strategic ICD programming0.360.13–0.930.03
Non-ischaemic heart failure0.460.21–0.960.04
Ejection fractionb0.990.95–1.030.76
Atrial fibrillation2.400.25–2.800.78
  • aFurther adjusted for beta-blocker treatment and device type.

  • bPer 5% increment.

  • HR, hazard ratio; CI, confidence interval; ICD, implantable cardioverter-defibrillator.

Figure 1

Cumulative probability for the occurrence of death or appropriate device shocks.

Inappropriate device shocks

During the follow-up period, a total of 41 patients (13.6% of the entire study population) experienced an inappropriately delivered shock, of whom 33 belonged to the control group and 8 to the strategically programmed group. Kaplan–Meier estimates of event-free from inappropriate shocks in the two groups are shown in Figure 2. At 24-month follow-up, the cumulative probability of survival from inappropriate shock delivery was 67% in patients in the physician-control group, compared with 95% in the strategically programmed group (log-rank P = 0.01 for the overall difference in event rates during follow-up). The most common underlying cause of inappropriate shock delivery in the entire study cohort was failure of discrimination between SVT and VT, which occurred in 34 patients (83% of the inappropriate shocks in the entire cohort).

Figure 2

Cumulative probability for the occurrence of inappropriate device shocks.

As shown in Table 4, no significant differences existed between the study groups regarding the cause of inappropriate shock delivery. Incorrect diagnosis of SVTs was more commonly the aetiology among patients in the control group who had received dual-chamber rather than single-chamber devices. This difference-by-device effect was not observed in the strategically programmed group, possibly due to the very low number of events in this group. Table 5 shows the results of the multivariable model for predictors of inappropriate shock delivery. Strategic programming strategy was associated with a 70% reduction in the risk for inappropriate shock deliveries (95% CI: 0.17–0.62; P = 0.01).

View this table:
Table 4

Inappropriate ICD shocks in the two programming groups by device typea

Physician tailored programming (n = 140)Strategic programming (n = 160)
Single chamber (n = 35)Dual chamber (n = 48)CRT-D (n = 57)Single chamber (n = 56)Dual chamber (n = 61)CRT-D (n = 43)
Total events11 (31)14 (29)8 (14)3 (5)3 (5)2 (5)
SVT5 (14)13 (27)5 (9)1 (2)01 (2)
Sinus tachycardia4 (11)01 (2)2 (4)1 (2)0
Abnormal sensing002 (4)01 (2)1 (2)
Unclassified1 (3)1 (2)001 (2)0
  • aValues are in n (%).

  • Percentage is calculated from the total number of device type.

  • SVT, supraventricular tachycardia, CRT-D, cardiac resynchronization therapy with a defibrillator.

View this table:
Table 5

Multivariate analysis: predictors of inappropriate device shocks in the study population

VariableHR95% CIP value
Strategic programming0.300.17–0.620.01
Atrial fibrillation1.91.05–3.20.01
Dual-chamber device1.20.67–2.100.4
Beta-blocker treatment0.690.33–1.090.07
Antiarrhythmic treatment0.620.27–0.990.05
Ischaemic heart failure1.10.74–1.90.6

Non-sustained ventricular tachycardias, unnecessary shocks, and syncope events

During the follow-up period, a total of 75 non-sustained VT events that were not treated by a shock or ATP occurred in the entire study population, of which 26 occurred within the control group and 49 within the strategic programming group. As shown in Figure 3, strategic programming was associated with an increased probability of the occurrence of non-treated VT episodes (log-rank P = 0.001 for the overall difference in event rates during follow-up). The number of beats and the tachycardia CL in these events were significantly different between the groups: the mean number of beats per event was 22 + 9 in the strategic programming group, while the corresponding number of beats in the control group was 7 + 4 (P < 0.001 for between-group difference). Of the total number of events in the strategic programming group, 17 events were of 20–30 beats in duration, while in the control group none of the events lasted longer than 17 beats, with 20 events lasting 5–10 beats. The mean tachycardia CL was 351 + 78 ms in the strategic programming group as compared with 401 + 112 in the control group (P = 0.02 for between-group difference).

Figure 3

Cumulative probability for the occurrence of untreated non-sustained VT events.

A total of 15 events were treated by ATP in the strategic programming groups, during which the mean CL was 315 + 57 ms and the mean number of beats to detection was 23 + 3. Of these events, 10 were terminated by the ATP and 5 persisted and were terminated by shock delivery. One of the non-sustained VT events was initially treated with shock delivery instead of by ATP, even though the CL was well within the ATP treatment zone: (‘unnecessary shock’); the CL in these events was 320 and the number of beats to detection was 25.

In the physician tailored treatment group, five events were treated by ATP, during which the mean CL was 351 ± 93 ms and the mean number of beats to detection was 9 + 2. Two of these events were terminated by ATP, while the remaining three necessitated shock delivery. In these three events, the mean CL was 340 ± 89 and the mean number of beats to detection was 9 ± 3. Five additional events were treated initially by unnecessary shock delivery, in which the mean CL and the numbers of beats to detection were similar to the corresponding numbers during the events that were initially treated by ATP (means of 343 ± 90 and 8 ± 3, respectively).

The overall number of syncope events was very low in the entire study cohort during follow-up: five patients described the syncope events, of these three were in the strategic programming group (2%) and two were in the control group (1.5%). Four of these events were attributed to sustained VT events that were below the detection thresholds (two in each group), and one event (that occurred in a patient in the strategic programming group) of non-tachycardia-related aetiology.

Finally, the devices of eight patients from the physician tailored group that had VT events were reprogrammed in a manner that was in agreement with the strategic programming parameters for their corresponding device types; five of these patients were treated with ‘unnecessary’ shocks (i.e. the CL at the time of the events was in the ATP treatment zone) and the other three were initially treated with ATP. None of the strategic programming group devices were reprogrammed during follow-up.


Our study was performed to evaluate the efficacy and safety of ICD programming strategy aimed at reducing the overall shock delivery burden by using pre-specified VT/VF detection and treatment parameters in primary prevention patients. The results indicate that such strategic programming can be considered both effective and safe, as it was found to be independently associated with a 64% reduction in the risk in occurrence of the combined outcome of overall mortality and appropriate shock deliveries, as well as with a 70% reduction in the risk of inappropriate shock deliveries over a median follow-up period of 24 months.

One of the important reasons for the relative under-utilization of primary prevention ICD devices is the candid concern regarding the potential associated complications, the most dreaded being inappropriate shock deliveries. Shock delivery, whether appropriate or not, might be a devastating event for patients and lead to significant psychological issues, decrease in the quality of life,5,6 and to increase the risk of mortality possibly via arrhythmogenic mechanisms.710 In prior attempts to overcome this issue, discrimination algorithms were developed to eliminate shock delivery for supraventricular rhythms, yet utilization of these strategies was found to confer only modest effect on inappropriate therapy reduction.14,15 These studies were followed by developments of novel approaches involving long detection times and liberal use of ATP, which were found to be highly effective as well as safe,1113 as shown by our group, even when compared with the previous discrimination algorithms.16,17 Our study shows that the results of the originally published PREPARE trial, in which devices of a single manufacturer were used and the follow-up period was 1 year,11 can be generalized to the various device manufacturers and be considered as efficacious and safe for longer follow-up periods.

These findings were more recently confirmed by two large, randomized trials: the MADIT-RIT trial and the ADVANCE III trial. MADIT-RIT showed significant reduction in inappropriate therapy and all-cause mortality during long-term follow-up with strategic ICD programing.18 Although the reduction in the mortality endpoint was only numerically lower in the strategic programming group of this study, it is conceivable that reductions in the number of inappropriate shocks resulted in a diminished myocardial damage and lower mortality in this group as compared with the control group. It is likely that our study was underpowered to demonstrate this beneficiary effect of strategic programming, which was shown to exist in the larger MADIT-RIT study.18 The ADVANCE III trial demonstrated that a delay in the arrhythmia detection (30 out of 40 vs. 18 out of 24 intervals, and standard detection) resulted in a lower rate of ATP and shocks, a reduction of inappropriate shocks and hospitalizations, regardless of aetiology or prevention criteria, thus ameliorating the acceptance of these devices.19

Several of our findings also provide an insight into the underlying mechanisms of these effects: first, we found that patients with strategically programmed devices clinically tolerated non-sustained VT events that were believed to be of potential haemodynamic or even life-threatening effects in the past, without an evident increase in the risk for overall mortality or syncope (although the latter occurred very infrequently in this study). Low-to-medium rate self-terminating VT events have been shown to be comprised of a relatively large portion of the entire VT episode burden as recorded by the ICD's in previous studies: the RELEVANT investigators reported that 91% of the VT episodes with a CL between 360 and 240 ms were self-terminating13 and did not pose an additional risk for the study patients. Secondly, we showed that ATP was an efficient therapeutic option for terminating VT events that in previous programming strategies necessitated the delivery of a shock. This finding concurs with past studies' results, which demonstrated that ATP therapy can be safely employed as a first therapeutic option for fast VT, and is associated with a considerable reduction in shock deliveries.12,13 Our finding that patients with dual-chamber devices belonging to the control group experienced more inappropriate shock due to misinterpretation of SVT is of considerable interest, although it should be interpreted with caution. It is possible that the post hoc interpretations by the reviewers of events in the dual-chamber device group were more accurate, given the additional information available from the atrial channel, as previously described by Diemberger et al.20 Nevertheless, as noted in Table 4, this effect was not present among patients in the strategic programming group, and so may prove to be a true benefit associated with the strategic programming parameters. Finally, it is important to note that a history of atrial fibrillation remained an important cause of inappropriate shock delivery in this study.

In a recent editorial discussing the numerous benefits of ICD's as they emerged in recent years, Gasparini and Nisam21 state that given the much improved technological aspects of ICD therapy, a large body of evidence exists that, rather than ‘harm’, the ICD/CRT-D provides substantial protection for appropriately selected patients. Indeed, based on the results of this and of the previous studies, it seems that strategic programming parameters aimed at enhancing detection and modifying the therapeutic parameters for the primary prevention population should be practiced whenever possible. The associated reduction in shock deliveries without compromising patient's safety is likely to lead to improved quality of life and less fear and anxiety from the patient's side, and to improved implementation of the current recommendations regarding ICD implantation from the physician's side. Improvement in the patient–physician relationships may also be an anticipated welcomed effect as a result of such practice.

Study limitations

The outcome data for this study were retrospectively analysed and were drawn from a single-centre ICD registry. In addition, this was not a randomized trial, and programming was performed at the discretion of the implanting physician, and thus open to selection bias. We attempted to overcome this limitation by performing carefully designed multivariable models that included the statistically significant baseline between-group differences. Ventricular fibrillation induction was left to the discretion of the implanting physician even in the early years of data collection (2005–07), yet this was practiced in both groups. Year of enrolment, ischaemic heart disease as aetiology for heart failure, and beta-blockers utilization were not balanced between the two study groups, which may have contributed to difference in the study's outcomes. To overcome this, a multivariate model was performed for the effect of strategic ICD programming on death or appropriate shocks, yet even this model cannot be counted on for eliminating all the between-group differences due to the retrospective fashion in which the outcome data were analysed. The devices of different manufacturers each utilize a slightly different protocol for strategic programming, and hence uniform programming was not possible for the entire study cohort. The occurrence of syncope was very low in our study cohort compared with previous studies that evaluated this outcome among the primary prevention patient population. In some of the study patients, the specific type of SVTs at the time of inappropriate therapies (i.e. atrial fibrillation, flutter, etc.) were not available for analysis, and the heart rate at the time of the tachycardia was not collected. We therefore decided to identify these events uniformly as ‘SVT’ rather than using specific aetiologies. The term ‘unnecessary shock’ was used in this manuscript, yet it should be acknowledged that there is currently no standard definition for this term. In this manuscript, VT events that were initially treated by shock, even though by device programming ATP treatment should have served as the initial treatment, were defined as unnecessary shocks. Data regarding the underlying rhythm at the time of death among the study patients were not available in most cases, and hence were not incorporated into this analysis. The underlying mechanism for the numerically lower mortality events rate in this analysis remains thus theoretical. Finally, only primary prevention patients were included in this study, and so its results should not be extrapolated to patients with previously known arrhythmias.


Carefully conducted strategic ICD programming that includes modified detection, treatment, and discrimination parameters in a primary prevention population is effective and safe in terms of overall mortality, appropriate and inappropriate shock deliveries. These effects were maintained at a median of 2-year follow-up and in devices of various manufacturers. Systematic programming for primary prevention patients should be warranted and will probably lead to a reduction in the various adverse effects associated with shock deliveries, as well as to enhanced implementations of current ICD implantation guidelines. Further randomized multicentre studies, with long-term follow-up periods are necessitated to validate these results.


We are grateful to Elaine Finkelstein for editing the manuscript

Conflict of interest: None declared.


View Abstract