You are here

  • Home
  • Archive
  • Volume 100, Issue 11
  • Ventricular rate monitoring as a tool to predict and prevent atrial fibrillation-related inappropriate shocks in heart failure patients treated with cardiac resynchronisation therapy defibrillators

Article Text

Article menu

Download PDFPDF

Arrhythmias and sudden death
Original article
Ventricular rate monitoring as a tool to predict and prevent atrial fibrillation-related inappropriate shocks in heart failure patients treated with cardiac resynchronisation therapy defibrillators
  1. Renato Pietro Ricci1,
  2. Carlo Pignalberi1,
  3. Maurizio Landolina2,
  4. Massimo Santini1,
  5. Maurizio Lunati3,
  6. Giuseppe Boriani4,
  7. Alessandro Proclemer5,
  8. Domenico Facchin5,
  9. Domenico Catanzariti6,
  10. Giovanni Morani7,
  11. Michele Gulizia8,
  12. Lorenza Mangoni9,
  13. Andrea Grammatico9,
  14. Maurizio Gasparini10,
  15. on behalf of the ClinicalService cardiological centres
  1. 1Department of Cardiology, San Filippo Neri Hospital, Rome, Italy
  2. 2Fondazione Policlinico S. Matteo IRCCS, Pavia, Italy
  3. 3Niguarda Ca’ Granda Hospital, Milano, Italy
  4. 4Institute of Cardiology, University of Bologna and Azienda Ospedaliera S.Orsola-Malpighi, Bologna, Italy
  5. 5S. Maria della Misericordia Hospital, Udine, Italy
  6. 6Santa Maria del Carmine Hospital, Rovereto, Italy
  7. 7Borgo Trento Hospital, Verona, Italy
  8. 8Garibaldi-Nesima Hospital, Catania, Italy
  9. 9EMEA Regional Clinical Centre - Medtronic Clinical Research Institute, Rome, Italy
  10. 10IRCCS Istituto Clinico Humanitas, Rozzano, Italy
  1. Correspondence to Dr Renato Pietro Ricci, Department of Cardiology San Filippo Neri Hospital, Via Martinotti, Rome 20 00135, Italy; renatopietroricci{at}tin.it

Abstract

Objective Inappropriate implantable cardioverter defibrillators (ICD) therapies have been associated with multiple adverse effects, including worse quality of life and prognosis. We evaluated the possibility of predicting atrial fibrillation (AF)-related inappropriate ICD shocks through continuous monitoring of device diagnostics.

Methods 1404 ICD patients were prospectively followed in an observational research by 74 Italian cardiology centres. Device diagnostics stored daily information on AF duration and ventricular rate (VR) during AF. Uncontrolled VR was defined as mean VR>80  beats per minute (bpm) and maximum VR>110 bpm. Expert electrophysiologists reviewed the ventricular tachycardia/ventricular fibrillation (VT/VF) episodes electrograms, stored in the device memory, and classified appropriate detections, inappropriate detection mechanisms and ICD therapy outcomes.

Results Over a median follow-up of 31 months, 511 (36%) patients suffered spontaneous VT/VF, which were treated by ICD shocks in a subgroup of 189 (13%) patients. Inappropriate detections occurred in 232 (16%) patients, and inappropriate ICD shocks in 101 (7%) patients. AF was the cause of inappropriate shocks in 60 patients. AF caused 144 inappropriate shocks: 53% of all inappropriate shocks. The likelihood of experiencing AF-related inappropriate shocks was 2.4% at 1 year and 6% at 5 years. Uncontrolled VR during AF proved to be an independent predictor of AF-related inappropriate shocks (OR=3.02, p=0.006); an alarm set at a VR>90 bpm or 100 bpm was associated with prediction of AF-related inappropriate shocks with a sensitivity of 73% or 62%, respectively.

Conclusions AF is the most common cause of inappropriate shocks in ICD patients. Continuous remote monitoring of VR during AF would promptly and efficiently predict AF-related inappropriate shocks.

Clinical trial registration http://clinicaltrials.gov/ct2/show/NCT01007474

  • QUALITY OF CARE AND OUTCOMES

https://doi.org/10.1136/heartjnl-2013-305259

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    The introduction of implantable cardioverter defibrillators (ICD) was a milestone in the treatment of patients at high risk of ventricular tachycardia/fibrillation and sudden cardiac death.1 Inappropriate ICD therapies, that is, therapies delivered on rhythms that are not ventricular tachyarrhythmias, affect 8–40% of ICD patients,2–4 and have been associated with multiple adverse effects, including impaired quality of life and worse prognosis.2–3 5–7

    In patients with heart failure (HF), systolic dysfunction and cardiac dyssynchrony, cardiac resynchronisation therapy (CRT) via biventricular pacing reduces both the risk of death and HF progression and improves symptoms and quality of life.8 ,9 The number of patients who receive a CRT-ICD is rapidly increasing.10

    Atrial fibrillation (AF) is a frequent comorbidity in HF patients11–13 and is the most common cause of inappropriate shocks in the general ICD population.1–4 14–18

    Modern high-technology ICDs have the capability of continuous monitoring of clinical variables associated with patient status and in particular to store information on AF occurrence and ventricular rate (VR) during AF.

    The objective of our research was to evaluate, in a population of CRT-ICD patients, the occurrence of inappropriate ICD shocks and the possibility of predicting AF-related inappropriate ICD shocks by exploiting information on AF and VR during AF.

    Methods

    Project organisation

    The patients included in the present analysis were followed up in a network of 74 cardiology centres that participate in the Italian ClinicalService project (Clinical Trial Registration Information: http://clinicaltrials.gov/ct2/show/NCT01007474). This is a national cardiovascular data repository and medical care project that aims to describe and improve the use of implantable cardiac devices in Italian clinical practice. The project was approved by each site's Institutional Review Board and conforms with the principles outlined in the Declaration of Helsinki. Each patient provided informed consent for data collection and analysis.

    Patient population and follow-up

    All patients received a Medtronic CRT-ICD in accordance with ICD and CRT guidelines1 ,19; that is, HF in New York Heart Association (NYHA) class II, III or IV, depressed ventricular function (LVEF≤35%) and wide QRS complex (≥120 ms). All patients were in sinus rhythm at the baseline and had no history of permanent AF.

    The first patient considered in this analysis underwent implantation in May 2004 while the last patient underwent implantation in August 2008. The last follow-up examination was held in February 2012.

    All patients were followed up according to each centre's clinical practice through standard in-hospital visits.

    Definitions

    For all arrhythmic episodes detected by the device as ventricular tachycardia/ventricular fibrillation (VT/VF), device diagnostics stored electrogram data 10 s prior to VT/VF detection, 2.5 s after detection and each therapy, and 10 s after therapy or VT/VF termination. A committee of expert electrophysiologists reviewed these data and, after identifying appropriate detections and inappropriate detection mechanisms, classified ICD shocks as appropriate or inappropriate in accordance with the detection classification.

    The device detects an atrial tachyarrhythmia if the median atrial interval is less than 350 ms and if A-V conduction is greater than 1:1 for ≥32 ventricular events. The sensitivity and specificity of the algorithm used to detect atrial tachyarrhythmias have been estimated to be as high as 100% by Swerdlow et al.20 In the present analysis, the term AF refers to any atrial tachyarrhythmia; patients were considered to have experienced AF if, on any day of the observation period, device diagnostics detected a cumulative AF duration >10 min, a cut-off duration recognised as appropriate in order to discard false detections due to premature atrial contraction runs or spurious events. Patients were assigned to five AF profile classes, according to the duration of the longest AF episode: >10 min, >6 h, >24 h, >7 days and >6 months.

    Guidelines21 define ‘strict’ VR control during AF as a VR≤80 beats per minute (bpm) during resting 12-lead electrocardiography and ≤110 bpm during a 6 min walking test. For each follow-up day with AF, device diagnostics recorded the mean and maximum VR during AF. We assumed that mean VR would be an approximation of VR at rest and that maximum VR would be an approximation of VR during the walking test. Therefore, on the basis of the VR stored in the device diagnostics, we defined patients with uncontrolled VR as those who, during AF days, experienced a mean VR>80 bpm and a maximum VR>110 bpm. The observation period taken into account to estimate VR was the whole observation period for patients without AF-related inappropriate shocks and the period preceding the first AF-related inappropriate shock for patients with this event.

    Project design and objectives

    This observational research had three main objectives. First, we wanted to estimate the incidence of appropriate and inappropriate ICD detections and shocks and to describe the detection mechanisms of inappropriate shocks. Second, we tested the possibility of predicting AF-related inappropriate ICD shocks by correlating these events with the main patient baseline characteristics, with device programming and with patients’ follow-up characteristics, such as AF occurrence and uncontrolled VR during AF. Third, we estimated the accuracy of an alarm, based on uncontrolled VR during AF, in preventing AF-related inappropriate ICD shocks.

    Device programming

    Device programming was left to each centre's clinical practice. At the baseline, the initial number of intervals to detect a VF episode was programmed at 12/16 in 58.0% of patients, at 18/24 in 38.1% of patients and at 24/32 in 3.9% of patients; the minimum value of the VF detection window was set at a rate between 166  and 194 bpm in 78% of patients and at a rate ≥200 bpm in 22% of patients. A VT detection window was programmed in 56% of patients, the minimum VT detection rate being set in an interval between 146 and 188 bpm according to each centre's clinical practice and patients’ characteristics. In 86.2% of patients, a fast VT detection zone was programmed in order to initiate antitachycardia pacing (ATP) as the first therapy in the detection window between 166 and 240 bpm. The ICDs used were endowed with three PR Logic detection criteria designed to withhold inappropriate ventricular detection during episodes of rapidly conducted supraventricular tachycardia (SVT). These criteria are an AFib/AFlutter criterion, designed to identify rapidly conducted AF, atrial flutter or atrial tachycardia; a Sinus Tach criterion, to identify sinus tachycardia; and an Other 1:1 SVTs criterion, designed to identify other 1:1 SVTs in which atrial and ventricular activation are roughly simultaneous. These criteria were programmed according to physicians’ discretion: in 37.0% of patients, all three detection criteria were enabled; in 59.7% of patients, only the AFib/AFlutter and Sinus Tach criteria were programmed, while in 3.3% of patients no criteria were enabled.

    Statistical analysis

    Continuous variables are reported as means and SDs or medians with 25th–75th centile, as appropriate. Absolute and relative frequencies are reported for categorical variables. Comparisons of categorical variables were performed by means of Fisher's exact test.

    Times to the first AF-related shock were described by means of the Kaplan–Meier curve. Predictors of AF-related shock were studied by applying univariate and multivariate logistic analyses to estimate ORs with 95% CIs. After checking for collinearity, we included in the multivariate model any baseline variable with p<0.1 on univariate analysis.

    Stata/SE V.12.1 for Windows (StataCorp LP, Texas, USA) was used for statistical analysis.

    Results

    Our analysis considered 1404 patients with CRT-ICD and atrial, right ventricular and left ventricular transvenous leads. Baseline patient characteristics are shown in table 1. Paroxysmal AF was present in 144 (10%) patients and persistent AF in 218 (16%) patients. In the latter patients, atrio-ventricular (AV) node ablation was performed in 13 patients to optimise CRT delivery; these patients contributed to the analysed population till the time of the AV node ablation and then they were censored.

    View this table:
    Table 1

    Baseline patient characteristics

    All patients were followed up for at least 6 months. Median follow-up was 31 months (25th–75th centile range: 16–46 months). The total follow-up period was 3754 years.

    AF occurrence and AF profile

    AF lasting >10 min occurred in 556/1404 (40%) patients; in 237 (17%) patients, AF was a recurrence of an arrhythmia already documented in the patient's history, while in 319 (23%) patients the detected AF was a new-onset arrhythmia. The longest AF episode lasted between 10 min and 6 h in 144 (10%) patients, between 6 h and 1 day in 88 (6%) patients, between 1 day and 7 days in 72 (5%) patients, and between 7 days and 6 months in 193 (14%) patients; 59 (4%) patients suffered AF episodes longer than 6 months.

    ICD arrhythmia detection

    ICD arrhythmia detection is described in figure 1, which also shows the subclassification of appropriate detections, according to ventricular arrhythmia type, and of inappropriate detections, according to inappropriate detection mechanisms. Out of 1404 patients, 613 (44%) had 5394 episodes: 511 (36%) had 3894 appropriately detected episodes and 232 (16%) had 1500 inappropriate detections; 381 (27%) patients had only appropriate detections, 130 (9%) patients had both appropriate and inappropriate detections, and 102 (7%) patients had only inappropriate detections. On limiting the analysis to the episodes treated by shock, 165 (12%) patients received only appropriate shocks, 24 (2%) received both appropriate and inappropriate shocks, and 77 (5%) received only inappropriate shocks. The annual rate of patients with inappropriate shocks was 2.69 patients per 100 patient-years.

    Figure 1
    Figure 1

    Number of patients with implantable cardioverter defibrillators detected/shocked episodes.

    AF-related inappropriate ICD detections and shocks

    AF was the cause of inappropriate detections in 168 patients (72% of patients with inappropriate detections) and in 962 inappropriately detected episodes (64% of inappropriately detected episodes). It was also the cause of inappropriate shocks in 60 patients (4% of the whole population and 59% of patients with inappropriate shocks). AF caused 144 inappropriate shocks: 53% of all inappropriate shocks. As shown in figure 2, the likelihood of experiencing one or more AF-related inappropriate shocks was 2.4% at 1 year, 4% at 2 years, 5% at 3 years, 5.2% at 4 years and 6% at 5 years.

    Figure 2
    Figure 2

    Kaplan–Meier estimation of the probability of atrial fibrillation-related inappropriate shocks.

    Of 60 patients with AF-related shocks, 38 (63%) experienced only one episode that triggered inappropriate shocks, 9 (15%) experienced two such episodes and 13 (22%) experienced more than three episodes. The distribution of appropriate shocks and AF-related inappropriate shocks is shown in figure 3 as a function of episode ventricular rate. The median (25th–75th centile) ventricular cycle length of episodes that triggered AF-related inappropriate shocks was 300 (280–332) ms, which corresponds to a median rate of 200 (180–214) bpm.

    Figure 3
    Figure 3

    Ventricular rate of appropriate implantable cardioverter defibrillators (ICD) shocks and atrial fibrillation-related inappropriate ICD shocks.

    Predictors of AF-related inappropriate shocks

    The following baseline characteristics of patients were associated with a higher risk of AF-related inappropriate shocks: AF history, NYHA class II, as compared with NYHA class III–IV, and non-ischemic heart disease (table 2). Of the patients with AF history, 8.3% suffered AF-related inappropriate shocks compared with 2.9% of patients without AF history (p<0.001 Fisher's exact test). Among device-detected variables, such as AF profile and uncontrolled VR during AF, and among device programming variable, such as VT and VF detection rates, programming of SVT discrimination algorithms and number of intervals to detect, at univariate analyses, only uncontrolled VR during AF proved to be associated with AF-related inappropriate shocks and this association was confirmed in the multivariate model, as shown in table 2.

    View this table:
    Table 2

    Univariate and multivariate logistic analysis of potential predictors of AF-related inappropriate shocks

    Prediction of AF-related inappropriate shocks on the basis of VR monitoring

    We estimated that an alarm set at a VR of 90 bpm during AF would have triggered an alert in 395/556 (71%) patients, thereby potentially avoiding shocks in 44 of the 60 patients who actually suffered AF-related inappropriate shocks. This alarm would have a sensitivity of 44/60 (73%), a specificity of 145/496 (29%), a positive predictive value of 44/395 (11%) and a negative predictive value of 145/161 (90%). The median alert time, that is, the time between the VR alert and the AF-related inappropriate shock, would have been 138 days: minimum 12 days, 25th–75th centile ranging from 40 to 180 days. A VR alarm set at 100 bpm would have triggered an alert in 294/556 (53%) patients, thus potentially avoiding AF-related inappropriate shocks in 37 of the 60 patients who actually suffered AF-related inappropriate shocks. This alarm would have a sensitivity of 37/60 (62%), a specificity of 239/496 (48%), a positive predictive value of 37/294 (13%) and a negative predictive value of 239/262 (91%). The median alert time, that is, the time between the VR during AF alert and the AF-related inappropriate shock, would have been 120 days: minimum 14 days, 25th–75th centile ranging from 56 to 190 days.

    Discussion

    The main results of our observational research show that (1) AF is frequent in CRT-ICD patients and is the main cause of inappropriate shocks; (2) VR during AF is an independent predictor of AF-related inappropriate shocks; and (3) monitoring uncontrolled VR during AF would promptly and efficiently predict patients prone to AF-related inappropriate shocks.

    AF is the main cause of inappropriate detection and shocks

    AF was a common comorbidity in our CRT-ICD patients, being present in 26% of patients at the baseline and occurring in 40% of patients during the follow-up. AF was the most frequent cause of inappropriate shocks, constituting the mechanism of inappropriate detection in 60% of patients with inappropriate shocks and in 53% of episodes that triggered inappropriate shocks. These data confirm previous findings, which show that rapidly conducted atrial tachyarrhythmias account for 40–60% of inappropriate shocks.2 ,3 ,14 ,15

    Predictors of AF-related inappropriate shocks

    Multivariate analysis showed that AF history and uncontrolled VR during AF were independent predictors of AF-related inappropriate shocks. The relationship between AF history and inappropriate shocks is well documented,2 ,3 ,14 ,15 and it is explained by the fact that AF history is associated with AF recurrence in the follow-up. The association between uncontrolled VR during AF and inappropriate shocks is a new finding of our research and represents the theoretical basis for preventing AF-related inappropriate shocks through VR diagnostics monitoring. Conversely, our data do not show a relationship between AF-related inappropriate shocks and AF profile, probably because, while patients with long-duration AF are exposed to a higher risk of inappropriate shocks, their arrhythmia is clinically known and treated by rate-control agents; by contrast, short-duration AF may be underdetected during standard visits and standard ECG monitoring, and therefore not treated.

    Also, low NYHA class and non-ischaemic heart disease emerged from multivariate analysis associated with AF-related inappropriate shocks. A higher activity level, in low NYHA class patients, and a more lenient rate control, in non-ischaemic patients, may be factors predisposing to inappropriate shocks, but we believe that the predictive value of these variables should be considered with caution.

    A possible way to prevent AF-related inappropriate shocks

    Web-based applications now allow cardiologists to remotely control ICD patients. Both AF and VR during AF can be programmed as clinically relevant conditions that trigger patient alarms or device data transmission in order to alert cardiologists. Our analyses suggest that an alarm set at a mean VR higher than 90 bpm would alert physicians many days before the occurrence of AF-related inappropriate shocks. This alarm would have a sensitivity of about 73% in indicating patients who will suffer inappropriate shocks. Other VR cut-off values may well be used, though their sensitivity would be lower; for example, a 100 bpm cut-off would result in 62% sensitivity. The practical utility of an alarm set at 90 or 100 bpm should be weighed against the fact that, under lenient rate control, many AF patients may have a VR above those cut-off values for sustained periods of time. This was confirmed by our data that show that the possibility to predict 73% of patients with AF-related inappropriate shocks would have been obtained at the cost of reviewing remote control transmission, triggered by the VR alert, in 71% of patients with AF occurrence, which represents a high workload for healthcare personnel. We therefore suggest that VR during AF alerts should be programmed on a patient-by-patient basis, for example, by selecting patients without AF history or without optimal rate control.

    Clinical implications

    The findings of our research emphasise the clinical importance of performing continuous monitoring of AF and VR during AF. Indeed, detecting elevated VR during AF and acting accordingly, for example, by adjusting the patient’s medications, could reduce inappropriate ICD shocks. While AF duration and high VR during AF may be used either as automatic audible patient alerts or as remote-care alarms in new-generation devices, we would suggest using ‘silent’ remote alarms in order to achieve appropriate predictive sensitivity without overburdening caregivers with unscheduled hospital visits or causing discomfort or anxiety to patients.

    In MADIT II,3 over 1 year of follow-up, more than 5% of patients experienced inappropriate shocks due to atrial tachyarrhythmias. The probability of AF-related inappropriate shocks in our patient population (figure 2) was lower, suggesting the importance of using SVT discrimination algorithms, which were enabled in 97% of our patients.

    In our research, the median ventricular rate associated with episodes triggering AF-related inappropriate shocks was 200 bpm (figure 3), therefore we can hypothesise that setting the VF detection window at 200 bpm would have prevented about 50% of AF-related inappropriate shocks. Interestingly, Moss et al,14 in the MADIT RIT trial, showed that most inappropriate ICD shocks were due to atrial tachyarrhythmias and that programming ICD therapies only for ventricular tachyarrhythmias of 200 bpm or higher, as compared with conventional programming, reduced inappropriate shocks by 50%. In any case, the VT and VF detection windows should be programmed on the basis of patients’ characteristics; indeed, the benefit of eliminating a slow-VT detection zone in order to reduce the probability of inappropriate therapies should be balanced against the risk of failing to treat VT.

    Finally, the results of one recent randomised trial18 and of two observational studies16 ,17 have shown that prolonging the initial number of intervals for VF detection to 30/40 is safe and reduces the delivery rates of both total and inappropriate shocks. We therefore suggest that such programming should be implemented in patients at high risk of AF-related inappropriate shock.

    Limitations

    The limitations of multicentre observational studies, such as potential bias in patient selection, patient treatment, device programming and lack of a control group, apply to our research. In particular, when AF or VT/VF occurred, treatment was left to the usual clinical practice. Nevertheless, possible biases are mitigated by the fact that data were collected prospectively, the analysis was designed before the data set was opened and research endpoints were prespecified.

    The results of our research have value in that they generate hypotheses on the possibility to prevent inappropriate shocks by leveraging device diagnostics on VR. Nevertheless, we acknowledge that these hypotheses derive from a retrospective analysis and have not been proven prospectively.

    Variability in the programming of CRT-ICD for tachyarrhythmia detection and therapy may be regarded as a limitation. However, we aimed to characterise inappropriate ICD occurrence in a real clinical practice setting.

    As specified in the ‘Methods’ section, since the results of our analyses were mainly derived from device diagnostics, which does not distinguish among atrial fibrillation, atrial flutter and atrial tachycardia, the term AF in this paper must be regarded as representing any atrial tachyarrhythmia rather than atrial fibrillation specifically.

    Device diagnostics data may be used to improve patient care. However, they must be regarded only as added information and must be circumstantiated by a complete evaluation of each patient's baseline characteristics and current clinical conditions.

    Conclusions

    Our data confirm that AF is a frequent occurrence and the most common cause of inappropriate shocks in CRT-ICD patients. Since uncontrolled VR during AF is an independent predictor of AF-related inappropriate shocks, we suggest that alerts based on the remote monitoring of VR may reduce the burden of AF-related inappropriate shocks. Indeed, these alerts would promptly and efficiently predict shock occurrence in patients prone to AF-related inappropriate shocks. Moreover, carefully tailored programming of the VF detection window may also lower the inappropriate shock rate.

    Key messages

    What is already known on this subject

    • Inappropriate implantable cardioverter defibrillators (ICD) therapies have been associated with multiple adverse effects, including impaired quality of life and worse prognosis.

    What this study adds

    • Our research shows the possibility of predicting atrial fibrillation-related inappropriate ICD shocks through continuous monitoring of ventricular rate during atrial fibrillation.

    How might this impact on clinical practice

    • Web-based applications now allow cardiologists to remotely control ICD patients. Alarms programmed to alert cardiologists about uncontrolled ventricular rate may help to reduce inappropriate ICD shocks by adjusting patient’s medications or device programming.

    References

    1. ↵
      1. Zipes DP,
      2. Camm AJ,
      3. Borggrefe M,
      4. et al
      . ACC/AHA/ESC 2006 Guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death—executive summary: a report of the American College of Cardiology/American Heart Association Task Force and the European Society of Cardiology Committee for Practice Guidelines Developed in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society. Eur Heart J 2006;27:2099–140.
      OpenUrlFREE Full Text
    2. ↵
      1. Sweeney MO,
      2. Wathen MS,
      3. Volosin K,
      4. et al
      . Appropriate and inappropriate ventricular therapies, quality of life, and mortality among primary and secondary prevention implantable cardioverter defibrillator patients: results from the Pacing Fast VT REduces Shock ThErapies (PainFREE Rx II) trial. Circulation 2005;111:2898–905.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Daubert JP,
      2. Zareba W,
      3. Cannom DS,
      4. et al
      . MADIT II Investigators. Inappropriate implantable cardioverter defibrillator shocks in MADIT II: frequency, mechanisms, predictors, and survival impact. J Am Coll Cardiol 2008;51:1357–65.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. van Rees JB,
      2. Borleffs CJ,
      3. de Bie MK,
      4. et al
      . Inappropriate implantable cardioverter defibrillator shocks: incidence, predictors, and impact on mortality. J Am Coll Cardiol 2011;57:556–62.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Schron EB,
      2. Exner DV,
      3. Yao Q,
      4. et al
      . Quality of Life in the Antiarrhythmics Versus Implantable Defibrillators Trial. Impact of therapy and influence of adverse symptoms and defibrillator shocks. Circulation 2002;105:589–94.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Klein RC,
      2. Raitt MH,
      3. Wilkoff BL,
      4. et al
      . AVID Investigators Analysis of implantable cardioverter defibrillator therapy in the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. J Cardiovasc Electrophysiol 2003;14:940–8.
      OpenUrlCrossRefPubMed
    7. ↵
      1. Poole JE,
      2. Johnson GW,
      3. Hellkamp AS,
      4. et al
      . Prognostic importance of defibrillator shocks in patients with heart failure. N Engl J Med 2008;359:1009–17.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Abraham WT,
      2. Fisher WG,
      3. Smith AL,
      4. et al
      . Cardiac resynchronization in chronic heart failure. N Engl J Med 2002;346:1845–53.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Cleland JGF,
      2. Daubert JC,
      3. Erdmann E,
      4. et al
      . The Effect of Cardiac Resynchronization on Morbidity and Mortality in Heart Failure. N Engl J Med 2005;352:1539–49.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. van Veldhuisen DJ,
      2. Maass AH,
      3. Priori SG,
      4. et al
      . Implementation of device therapy (cardiac resynchronization therapy and implantable cardioverter defibrillator) for patients with heart failure in Europe: changes from 2004 to 2008. Eur J Heart Fail 2009;11:1143–51.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Wang TJ,
      2. Larson MG,
      3. Levy D,
      4. et al
      . Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study. Circulation 2003;107:2920–5.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Knight BP,
      2. Desai A,
      3. Coman J,
      4. et al
      . Long-term retention of cardiac resynchronization therapy. J Am Coll Cardiol 2004;44:72–7.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Borleffs CJ,
      2. Ypenburg C,
      3. van Bommel RJ,
      4. et al
      . Clinical importance of new-onset atrial fibrillation after cardiac resynchronization therapy. Heart Rhythm 2009;6:305–10.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Moss AJ,
      2. Schuger C,
      3. Beck CA,
      4. et al
      . MADIT-RIT Trial Investigators. Reduction in inappropriate therapy and mortality through ICD programming. N Engl J Med 2012;367:2275–83.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Wilkoff BL,
      2. Hess M,
      3. Young J,
      4. et al
      . Differences in tachyarrhythmia detection and implantable cardioverter defibrillator therapy by primary or secondary prevention indication in cardiac resynchronization therapy patients. J Cardiovasc Electrophysiol 2004;15:1002–9.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Wilkoff BL,
      2. Williamson BD,
      3. Stern RS,
      4. et al
      . PREPARE Study Investigators. Strategic programming of detection and therapy parameters in implantable cardioverter-defibrillators reduces shocks in primary prevention patients: results from the PREPARE (Primary Prevention Parameters Evaluation) study. J Am Coll Cardiol 2008;52:541–50.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Gasparini M,
      2. Menozzi C,
      3. Proclemer A,
      4. et al
      . A simplified biventricular defibrillator with fixed long detection intervals reduces implantable cardioverter defibrillator (ICD) interventions and heart failure hospitalizations in patients with non-ischaemic cardiomyopathy implanted for primary prevention: the RELEVANT [Role of long dEtection window programming in patients with LEft VentriculAr dysfunction, Non-ischemic eTiology in primary prevention treated with a biventricular ICD] study. Eur Heart J 2009;30:2758–67.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Gasparini M,
      2. Proclemer A,
      3. Klersy C,
      4. et al
      . Effect of long-detection interval vs standard-detection interval for implantable cardioverter-defibrillators on antitachycardia pacing and shock delivery: the ADVANCE III randomized clinical trial. JAMA 2013;309:1903–11.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Dickstein K,
      2. Vardas PE,
      3. Auricchio A,
      4. et al
      . Committee for Practice Guidelines of the European Society of Cardiology. 2010 focused update of ESC Guidelines on device therapy in heart failure: an update of the 2008 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure and the 2007 ESC Guidelines for cardiac and resynchronization therapy. Developed with the special contribution of the Heart Failure Association and the European Heart Rhythm Association. Eur J Heart Fail 2010;12:1143–53.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Swerdlow CD,
      2. Schsls W,
      3. Dijkman B,
      4. et al
      . Detection of atrial fibrillation and flutter by a dual-chamber implantable cardioverter-defibrillator. For the Worldwide Jewel AF Investigators. Circulation 2000;101:878–85.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Camm AJ,
      2. Kirchhof P,
      3. Lip GY,
      4. et al
      . Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery. Eur Heart J 2010;31:2369–429.
      OpenUrlFREE Full Text

    Footnotes

    • Contributors Each author has contributed significantly to the submitted work. In particular, RPR, MS, CP and AG contributed to the conception and design of the research project. RPR and CP interpreted data. LM and AG performed data analysis. RPR, MS, GB, MG ML, ML, CP, AP, DF, DC, GM and MG contributed to data acquisition and drafting of the manuscript or revising it critically for important intellectual content. RPR is responsible and guarantees for the overall content.

    • Funding This research was performed within the framework of the Italian ClinicalService, a project funded by Medtronic Italia, an affiliate of Medtronic Inc. No other funding sources were involved in the research.

    • Competing interests RPR has minor consultancy fees from Biotronik and Medtronic. GB received modest speaker fees from Boston Scientific and Medtronic. MS received modest research grants and/or speaker fees from Biotronik, Medtronic and St. Jude Medical. LM and AG are employees of Medtronic Italia, an affiliate of Medtronic Inc. The other authors have no relationship to disclose with regard to the present research.

    • Patient consent Obtained.

    • Ethics approval Ethics Committees or Medical Directors according to each of the involved hospitals.

    • Provenance and peer review Not commissioned; externally peer reviewed.