Periprocedural Risk and Survival Associated With Implantable Cardioverter-Defibrillator Placement in Older Patients With Advanced Heart Failure | Cardiology | JAMA Cardiology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Survival From All-Cause Death Following Implantable Cardioverter-Defibrillator Placement Among Patients With Advanced or Nonadvanced Heart Failure (HF)
Survival From All-Cause Death Following Implantable Cardioverter-Defibrillator Placement Among Patients With Advanced or Nonadvanced Heart Failure (HF)

Shaded areas indicate 95% CIs.

Figure 2.  Survival From All-Cause Death Following Device Placement Among Patients With Advanced or Nonadvanced Heart Failure (HF) by Device
Survival From All-Cause Death Following Device Placement Among Patients With Advanced or Nonadvanced Heart Failure (HF) by Device

Given that device-specific data were not reported, 47 patients were removed from the analysis: 44 patients in the nonadvanced HF group and 3 patients in the advanced HF group. Shaded areas indicate 95% CIs. CRT-D indicates cardiac resynchronization therapy-defibrillator; ICD, implantable cardioverter-defibrillator.

Table 1.  Descriptive Statistics for Select Variables by Implanted Device Typea,b
Descriptive Statistics for Select Variables by Implanted Device Typea,b
Table 2.  In-Hospital Event Rates Among Patients With Advanced HF and Patients With Nonadvanced HF
In-Hospital Event Rates Among Patients With Advanced HF and Patients With Nonadvanced HF
Table 3.  Event Rates Among Patients With Advanced HF Stratified by Device Typea
Event Rates Among Patients With Advanced HF Stratified by Device Typea
1.
Abouezzeddine  OF, Redfield  MM.  Who has advanced heart failure?: definition and epidemiology.   Congest Heart Fail. 2011;17(4):160-168. doi:10.1111/j.1751-7133.2011.00246.x PubMedGoogle ScholarCrossref
2.
Al-Khatib  SM, Stevenson  WG, Ackerman  MJ,  et al.  2017 AHA/ACC/HRS guideline 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 on Clinical Practice Guidelines and the Heart Rhythm Society.   Heart Rhythm. 2018;15(10):e190-e252. doi:10.1016/j.hrthm.2017.10.035PubMedGoogle ScholarCrossref
3.
Lindenfeld  J, Feldman  AM, Saxon  L,  et al.  Effects of cardiac resynchronization therapy with or without a defibrillator on survival and hospitalizations in patients with New York Heart Association class IV heart failure.   Circulation. 2007;115(2):204-212. doi:10.1161/CIRCULATIONAHA.106.629261 PubMedGoogle ScholarCrossref
4.
Hammill  SC, Kremers  MS, Kadish  AH,  et al.  Review of the ICD Registry’s third year, expansion to include lead data and pediatric ICD procedures, and role for measuring performance.   Heart Rhythm. 2009;6(9):1397-1401. doi:10.1016/j.hrthm.2009.07.015PubMedGoogle ScholarCrossref
5.
Messenger  JC, Ho  KK, Young  CH,  et al; NCDR Science and Quality Oversight Committee Data Quality Workgroup.  The National Cardiovascular Data Registry (NCDR) data quality brief: the NCDR Data Quality Program in 2012.   J Am Coll Cardiol. 2012;60(16):1484-1488. doi:10.1016/j.jacc.2012.07.020 PubMedGoogle ScholarCrossref
6.
van Buuren  S.  Multiple imputation of discrete and continuous data by fully conditional specification.   Stat Methods Med Res. 2007;16(3):219-242. doi:10.1177/0962280206074463 PubMedGoogle ScholarCrossref
7.
Kadish  A, Dyer  A, Daubert  JP,  et al; Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators.  Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy.   N Engl J Med. 2004;350(21):2151-2158. doi:10.1056/NEJMoa033088 PubMedGoogle ScholarCrossref
8.
Moss  AJ, Zareba  W, Hall  WJ,  et al; Multicenter Automatic Defibrillator Implantation Trial II Investigators.  Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction.   N Engl J Med. 2002;346(12):877-883. doi:10.1056/NEJMoa013474 PubMedGoogle ScholarCrossref
9.
Cleland  JG, Daubert  JC, Erdmann  E,  et al; Cardiac Resynchronization-Heart Failure (CARE-HF) Study Investigators.  The effect of cardiac resynchronization on morbidity and mortality in heart failure.   N Engl J Med. 2005;352(15):1539-1549. doi:10.1056/NEJMoa050496 PubMedGoogle ScholarCrossref
10.
Bristow  MR, Saxon  LA, Boehmer  J,  et al; Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators.  Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure.   N Engl J Med. 2004;350(21):2140-2150. doi:10.1056/NEJMoa032423 PubMedGoogle ScholarCrossref
11.
van Rees  JB, de Bie  MK, Thijssen  J, Borleffs  CJ, Schalij  MJ, van Erven  L.  Implantation-related complications of implantable cardioverter-defibrillators and cardiac resynchronization therapy devices: a systematic review of randomized clinical trials.   J Am Coll Cardiol. 2011;58(10):995-1000. doi:10.1016/j.jacc.2011.06.007 PubMedGoogle ScholarCrossref
12.
Dodson  JA, Reynolds  MR, Bao  H,  et al; NCDR.  Developing a risk model for in-hospital adverse events following implantable cardioverter-defibrillator implantation: a report from the NCDR (National Cardiovascular Data Registry).   J Am Coll Cardiol. 2014;63(8):788-796. doi:10.1016/j.jacc.2013.09.079 PubMedGoogle ScholarCrossref
13.
Poole  JE, Gleva  MJ, Mela  T,  et al; REPLACE Registry Investigators.  Complication rates associated with pacemaker or implantable cardioverter-defibrillator generator replacements and upgrade procedures: results from the REPLACE registry.   Circulation. 2010;122(16):1553-1561. doi:10.1161/CIRCULATIONAHA.110.976076 PubMedGoogle ScholarCrossref
14.
Haines  DE, Wang  Y, Curtis  J.  Implantable Cardioverter-Defibrillator Registry risk score models for acute procedural complications or death after implantable cardioverter-defibrillator implantation.   Circulation. 2011;123(19):2069-2076. doi:10.1161/CIRCULATIONAHA.110.959676 PubMedGoogle ScholarCrossref
15.
Lee  DS, Krahn  AD, Healey  JS,  et al; Investigators of the Ontario ICD Database.  Evaluation of early complications related to de novo cardioverter defibrillator implantation insights from the Ontario ICD database.   J Am Coll Cardiol. 2010;55(8):774-782. doi:10.1016/j.jacc.2009.11.029 PubMedGoogle ScholarCrossref
16.
Al-Khatib  SM, Hellkamp  A, Bardy  GH,  et al.  Survival of patients receiving a primary prevention implantable cardioverter-defibrillator in clinical practice vs clinical trials.   JAMA. 2013;309(1):55-62. doi:10.1001/jama.2012.157182 PubMedGoogle ScholarCrossref
Views 2,668
Citations 0
Original Investigation
March 25, 2020

Periprocedural Risk and Survival Associated With Implantable Cardioverter-Defibrillator Placement in Older Patients With Advanced Heart Failure

Author Affiliations
  • 1Division of Cardiology, Duke University Medical Center, Durham, North Carolina
  • 2Duke Clinical Research Institute, Durham, North Carolina
  • 3Beaumont Health, Dearborn, Michigan
  • 4Center for Outcomes Research and Evaluation, Yale University School of Medicine, New Haven, Connecticut
  • 5Department of Cardiology, The Permanente Medical Group, San Francisco, California.
  • 6Division of Research, Kaiser Permanente Northern California, Oakland, California
  • 7Division of Cardiology, David Geffen School of Medicine at UCLA (University of California, Los Angeles)
  • 8Ahmanson-UCLA Cardiomyopathy Center, Ronald Reagan UCLA Medical Center, University of California, Los Angeles
  • 9Section Editor, JAMA Cardiology
  • 10Division of Cardiology, Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora
  • 11Associate Editor, JAMA Cardiology
JAMA Cardiol. 2020;5(6):643-651. doi:10.1001/jamacardio.2020.0391
Key Points

Question  What are the utilization rates, patient characteristics, and outcomes of implantable cardioverter-defibrillator or cardiac resynchronization therapy-defibrillator placement among patients with advanced heart failure?

Findings  Among 81 492 Medicare patients in this cohort study, a small proportion of initial implantable cardioverter-defibrillator or cardiac resynchronization therapy-defibrillator placements for primary prevention of sudden cardiac death occurred in older patients with advanced heart failure. One-year all-cause mortality was 22%, more than 3-fold higher than that in patients without advanced heart failure, and patients with advanced heart failure had a periprocedural complication rate of 3.7%, primarily driven by in-hospital fatality or cardiac arrest.

Meaning  Among patients undergoing these procedures, only a small proportion had advanced heart failure, and these patients experienced clinically important periprocedural complication rates.

Abstract

Importance  Little is known about the utilization rates and outcomes of implantable cardioverter-defibrillator (ICD) and cardiac resynchronization therapy defibrillator (CRT-D) placement among patients with advanced heart failure (HF).

Objective  To examine utilization rates, patient characteristics, and outcomes of ICD and CRT-D placements among patients with advanced HF.

Design, Setting, and Participants  This cohort study was a post hoc analysis of 81 492 Medicare fee-for-service beneficiaries enrolled in the National Cardiovascular Data Registry ICD Registry between January 2010 and December 2014. Inclusion criteria were patients who had received an HF diagnosis, had a left ventricular ejection fraction of 35% or lower, and showed evidence of advanced HF, which was defined as New York Heart Association (NYHA) class IV symptoms, inotrope use within the last 60 days, left ventricular assist device in situ, or orthotopic heart transplant listing. The comparator group included patients with NYHA class II and no HF hospitalization within the last 12 months, no left ventricular assist device, no orthotopic heart transplant listing, and no current or recent inotrope use. All eligible patients underwent first-time ICD or CRT-D placement for primary prevention of sudden cardiac death. Data were analyzed from January 2010 to December 2014.

Main Outcomes and Measures  All-cause mortality and periprocedural complications.

Results  Of 81 492 Medicare patients, 3343 had advanced HF (4.1%) and 19 424 were in the comparator group (23.8%). Among the advanced HF population, the mean (SD) age of patients was 74 (9) years, and patients were predominantly white individuals (81.5%) and men (71.1%). The all-cause mortality rate at 30 days was 3.1% (95% CI, 2.6%-3.8%) in the advanced HF group vs 0.5% (0.4%-0.6%) in the comparator group (P < .001). In the advanced HF population, the aggregate in-hospital periprocedural complication rate was 3.74% (95% CI, 3.12%-4.44%) vs 1.10% (95% CI, 0.95%-1.25%) in the comparator group (P < .001). Most adverse events in this group were in-hospital fatality (1.82%; 95% CI, 1.40%-2.34%) and resuscitated cardiac arrest (1.05%; 95% CI, 0.73%-1.45%). Patients with NYHA class IV (hazard ratio, 1.40; 95% CI, 1.02-1.93; P = .04), ischemic heart disease (hazard ratio, 1.24; 95% CI, 1.04-1.48; P = .02), or diabetes (hazard ratio, 1.17; 95% CI, 1.04-1.33; P = .01) had a higher risk of death.

Conclusions and Relevance  Among patients undergoing ICD or CRT-D placement for primary prevention of sudden cardiac death, only a small proportion had advanced HF. These patients experienced clinically important periprocedural complication rates associated with in-hospital death and cardiac arrest relative to patients with nonadvanced HF.

Introduction

Heart failure (HF) with reduced ejection fraction is a complex clinical syndrome. Despite major advances in medical therapy, HF with reduced ejection fraction remains a major public health problem with a high burden of morbidity and mortality.1 Advanced HF represents the late stage of HF with reduced ejection fraction and is characterized by a significant reduction in left ventricular ejection fraction with New York Heart Association (NYHA) class IV symptoms that are refractory to optimal medical therapy.1 Professional guidelines recommend against placement of an implantable cardioverter-defibrillator (ICD) without cardiac resynchronization therapy (CRT) in patients for whom meaningful 1-year survival is not expected or in patients who are not candidates for advanced HF therapies, such as cardiac transplant or a left ventricular assist device (LVAD).2

Although the role of the ICD in patients having HF with reduced ejection fraction continues to expand, there is a lack of data on the utilization rates and outcomes of ICD or CRT-defibrillator (CRT-D) use among patients with advanced HF. Ambulatory patients with NYHA class IV HF made up only 4% of the 10 803 patients studied in CRT trials, and the majority of the ambulatory patients with class IV HF were from the COMPANION (Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure) trial.3 With more than 1800 participating institutions in the United States, the National Cardiovascular Data Registry (NCDR) ICD Registry provides a unique opportunity to analyze outcomes associated with ICD or CRT-D placement in patients with advanced HF. Thus, the specific objectives of the present analysis were to systematically describe the clinical characteristics, periprocedural complications, long-term all-cause mortality, and factors associated with all-cause mortality among patients with advanced HF and having an ICD or CRT-D. Furthermore, we aimed to compare outcomes of patients with advanced HF with a comparator group of patients without evidence of advanced disease who underwent ICD or CRT-D placement.

Methods
Data Sources

Patient and device implantation data were obtained from the NCDR ICD Registry. Patient demographic characteristics, clinical characteristics, and procedural data were collected using a standardized data collection form and submitted by participating hospitals to the ICD Registry via a secure website. Data are audited periodically by reviewing records at a random sample of sites.4,5 Data on periprocedural complications are recorded in the NCDR ICD Registry. Annual audits are conducted to assess data validity and reliability.5 Data on all-cause mortality rate, all-cause hospital admission rate, HF hospital admission rate, and cardiovascular (CV) hospital admission rate at 30 days, 6 months, and 1 year were obtained by linking the NCDR ICD Registry with the Master Beneficiary Summary File (MBSF) to obtain date of death and with the Medicare inpatient institutional claims to identify admissions. In addition, enrollment information from the MBSF, inpatient institutional claims, outpatient institutional claims, and noninstitutional claims were used to identify inclusion criteria and clinical outcomes. These databases contain claims for inpatient admissions, outpatient procedures, and enrollment and vital status for Medicare fee-for-service (FFS) beneficiaries. The Human Investigation Committee of the Yale University School of Medicine approved the use of a limited data set from the NCDR for the present research and granted a waiver for obtaining informed patient consent because the study used exclusively retrospective, deidentified administrative records.

Study Population

All patients enrolled in the NCDR ICD Registry from January 1, 2010, through December 31, 2014, were identified. We included patients who had received a diagnosis of HF, had a left ventricular ejection fraction of 35% or lower, had evidence of advanced HF, and had received a new ICD or CRT-D implant for primary prevention of sudden cardiac death. Advanced HF was defined as one of the following: (1) NYHA class IV, (2) inotropic support within the last 60 days, (3) LVAD in situ, or (4) current orthotopic heart transplant (OHT) listing. Inotropic support and LVAD in situ were identified using historical Centers for Medicare & Medicaid Services claims data 1 year prior to the ICD placement (see eTables 1, 2, and 3 in the Supplement for a list of International Classification of Diseases, Ninth Revision, Clinical Modification codes used to identify these criteria). The comparator group comprised patients with NYHA class II symptoms and no HF hospitalization within the last 12 months, no LVAD, no OHT listing, and no current or recent inotrope use. The rationale for the use of NYHA class II only was because these patients clearly have a milder form of HF symptoms than patients with NYHA class IV symptoms and currently have a class I indication for ICD or CRT-D placement.

All included patients were required to have 1 year of prior Part A and Part B FFS enrollment for us to evaluate the inclusion and exclusion criteria. The designation of primary prevention was determined by the physician who performed the implant. Patients were excluded if they had an acute myocardial infarction within 40 days or revascularization within 90 days, prior ventricular tachycardia or ventricular fibrillation arrest, arrhythmogenic syndromes with an increased risk of sudden cardiac death (eg, long or short QT syndrome, catecholaminergic polymorphic ventricular tachycardia, or Brugada syndrome), a lead-only procedure, or any procedure involving lead extraction. Medicare data were linked to the NCDR ICD Registry using the social security number, date of birth, and sex.

Outcomes

Outcomes of interest were all-cause mortality, all-cause hospital admission, median survival rate, CV hospital readmission, HF hospital readmission, periprocedural complications, and factors associated with all-cause mortality. Mortality was assessed at 30 days, 6 months, and 1 year. Hospital readmissions were assessed at 30 days, 6 months, and 1 year. Medicare data for 1-year outcome were obtained through the end of 2015. Periprocedural complications were ascertained based on the NCDR ICD Registry data collection form completed at the time of ICD or CRT-D placement. A composite periprocedural complication was considered as any occurrence of death, resuscitated cardiac arrest, myocardial infarction, cardiac perforation, coronary venous dissection, tamponade, stroke, transient ischemic attack, hematoma, infection requiring antibiotics, pneumothorax, hemothorax, or urgent cardiac surgery. All outcomes after hospital discharge were evaluated among patients discharged alive. For mortality, patients were followed up until the end of the study period, FFS disenrollment, or death, whichever came first. For hospital admission outcomes, estimates were calculated among patients enrolled in FFS during the entire time frame of interest.

Statistical Analysis

Study participants who met the definition of having advanced HF were grouped based on ICD or CRT-D placement. The unadjusted periprocedural event rates were calculated as an incidence rate (ie, number of events/number of procedures). Unadjusted survival was estimated using Kaplan-Meier product-limit estimates. Median survival and mortality rates at 30 days, 6 months, and 1 year were obtained from the estimated survival curve. Log-rank tests were used to compare the advanced HF group with the comparator group. Hospital admission outcomes were estimated as incidence rates defined by any occurrence of the readmission of interest during the time frame under study among those who were enrolled in FFS Medicare during that time frame (ie, number of all-cause hospital admissions within 30 days divided by the number of patients enrolled in FFS for 30 days after hospital discharge). For all estimates, 95% CIs were calculated.

For the mortality outcome, an adjusted Cox proportional hazards regression model was calculated to obtain hazard ratio (HR) estimates for the following clinical variables of interest: sex, NYHA class IV status, transplant waiting list status, inotrope use, LVAD, ischemic heart disease, nonischemic cardiomyopathy, and diabetes. These models were adjusted for the clinical variables of interest and the following additional potential confounders: age, blood urea nitrogen level, body mass index, digoxin use, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use, and β-blocker use.

Prior to modeling, single imputation was performed for missing continuous and categorical variables using fully conditional specification methods.6 Missingness rates ranged from 0% to 1.9%, with most variables missing less than 1%. All analyses were 2-sided, and P < .05 was set as the threshold for statistical significance. Statistical analyses were performed from January 2010 to December 2014 using SAS software, version 9.4 (SAS Institute Inc).

Results

A total of 81 492 patients with HF and a left ventricular ejection fraction of 35% or lower underwent ICD or CRT-D placement for primary prevention of sudden cardiac death and were successfully linked to Medicare claims (eFigure in the Supplement). Of these patients, 3872 met our criteria for having advanced HF. After excluding 327 cases of acute myocardial infarction or revascularization of 1 year or less, 93 cases of prior ventricular tachycardia or ventricular fibrillation arrest, 95 cases of arrhythmogenic syndromes with an increased risk of sudden cardiac death, and 80 cases of lead-only procedures, the final analytical cohort included 3343 patients with advanced HF (4.1%). A total of 19 424 patients (23.8%) formed the comparator group.

Among the advanced HF population, the mean (SD) age of patients was 74 (9) years, and they were predominantly white individuals (81.5%) and men (71.1%). Most patients had hypertension (84.2%), ischemic heart disease (63.1%), or prior HF hospitalization (63.8%). There were 2506 patients (75.0%) who had NYHA class IV symptoms, 679 patients (20.3%) currently or recently receiving inotropes, 129 patients (3.9%) who had an LVAD, and 158 patients (4.7%) who were listed for OHT. The characteristics of the cohort are presented in Table 1. Only 3.5% of the population met 2 or more advanced HF criteria. The majority of patients (2425 [72.5%]) received CRT in addition to an ICD. Among patients with a CRT-D, the prevalent characteristics included atrial fibrillation or flutter (50.7%) and left bundle branch block (51.9%). The prevalent characteristics of patients with ICD alone were prior myocardial infarction (50.4%) and sinus rhythm (68.7%) (Table 1).

Patients in the comparator group had a lower burden of comorbid disease, such as lung disease (20.5% vs 32.8%), diabetes (38.5% vs 47.6%), and atrial fibrillation (35.3% vs 48.4%). Furthermore, the comparator cohort had a higher mean (SD) left ventricular ejection fraction (26.9% [5.7%] vs 22.9% [7.0%]) with a higher use of guideline-directed pharmacological therapy. In contrast to the advanced HF population, the comparator population underwent mostly primary ICD placement as opposed to CRT-D placement (72.5% vs 27.5%).

Clinical Outcomes

All-cause survival rates for both groups are shown in Figure 1. In the advanced HF population, the median survival following device placement was 3.5 years (95% CI, 3.3-3.7 years). The all-cause mortality rate at 30 days was 3.1% (95% CI, 2.6%-3.8%) in the advanced HF group vs 0.5% (95% CI, 0.4%-0.6%) in the comparator group (P < .001) (Table 2). The all-cause mortality rate at 6 months was 13.7% (95% CI, 12.5%-15.0%) in the advanced HF group vs 3.6% (95% CI, 3.3%-3.8%) in the comparator group (P < .001). The all-cause mortality rate at 1 year was 21.9% (95% CI, 20.4%-23.5%) in the advanced HF group vs 7.1% (95% CI, 6.7%-7.5%) in the comparator group (P < .001).

Median survival rates and event rates were further stratified by device type as shown in Table 3, Figure 2, and eTable 4 in the Supplement (comparison with comparator group). A sensitivity analysis for all-cause mortality comparing the contribution of individual, advanced HF criteria with median survival was performed. Patients meeting the NYHA class IV criterion had overall the shortest median survival (3.4 years; 95% CI, 3.1-3.6 years) followed by patients listed for OHT (3.8 years; 95% CI, 3.0-4.6 years) (eTable 5 in the Supplement).

The composite periprocedural complication rate among patients with advanced HF was 3.74% (95% CI, 3.12%-4.44%) in contrast to 1.10% (95% CI, 0.95%-1.25%; P < .001) in the comparator group. For patients with advanced HF, the majority of adverse events were in-hospital fatality (1.82%; 95% CI, 1.40%-2.34%) or resuscitated cardiac arrest (1.05%; 95% CI, 0.73%-1.45%). Other periprocedural and hospital admission outcomes are presented in Table 2 and Table 3. By contrast, in the comparator group, the periprocedural complication profile shifted toward less serious complications, with pneumothorax or hemothorax (0.40%; 95% CI, 0.32%-0.50%) and hematoma (0.27%; 95% CI, 0.20%-0.35%) being the most common. The median time from procedure to first hospital admission was 101 days (interquartile range, 27-295 days) for the advanced HF group and 200 days (interquartile range, 61-485 days) for the comparator group.

In unadjusted Cox proportional hazards regression models of the advanced HF population, NYHA class IV, LVAD, ischemic heart disease, and diabetes were all associated with a higher rate of mortality (eTable 6 in the Supplement). After adjustment, patients with NYHA class IV (HR 1.40; 95% CI, 1.02-1.93; P = .04), ischemic heart disease (HR, 1.24; 95% CI, 1.04-1.48; P = .02), or diabetes (HR, 1.17; 95% CI, 1.04-1.33; P = .01) continued to have a higher risk of death (eTable 6 in the Supplement).

Discussion

In this analysis of NCDR ICD Registry data, we found a small proportion of initial ICD or CRT-D placements for primary prevention of sudden cardiac death in older patients with advanced HF, and most of these patients (72.5%) received a CRT-D device. In the HF population, the all-cause mortality was 22%, and median survival following device placement was 3.5 years (95% CI, 3.3-3.7 years). All-cause mortality was more than 3-fold as high as a comparator group of patients without evidence of advanced HF undergoing ICD or CRT-D placement. Furthermore, patients with advanced HF had a periprocedural complication rate of 3.7%, which was primarily associated with in-hospital fatality or cardiac arrest. Finally, after adjustment, patients with NYHA class IV, ischemic heart disease, or diabetes had a higher rate of mortality.

There are limited data on the outcomes following ICD placement in patients with advanced HF because such patients were excluded from the pivotal trials of primary prevention ICDs (without CRT).7,8 Two trials of CRT allowed for the inclusion of ambulatory patients with NYHA class IV; however, such patients constituted only 50 patients (6%) enrolled in the CARE-HF (Cardiac Resynchronization-Heart Failure) trial9 and 217 patients (14%) enrolled in the COMPANION trial.10 However, because the role of the ICD and the decision for or against it present an important clinical question in this subgroup of patients with HF, it is critical to better understand outcomes, such as perioperative complications and all-cause mortality, in these patients.

A systemic review of major, randomized ICD or CRT clinical trials reported an overall 3.02% rate of periprocedural complications, mainly associated with lead dislodgement and hematoma, and a mean in-hospital mortality of 0.2% for ICD placements and of 0.3% for CRT placements (nonthoracotomy).11 Comparable findings were reported by Dodson et al12 for 240 632 ICD Registry patients undergoing ICD placement from 2010 to 2011. Dodson et al12 found an overall 1.8% periprocedural complication rate and a 0.26% in-hospital mortality rate. Patients with NYHA class IV were twice as likely to experience complications than patients with NYHA class I or II symptoms. In our study among patients with advanced HF, we observed higher periprocedural complication (3.7%) and in-hospital mortality (1.8%) rates than those observed in the comparator group and prior studies. Notably, the periprocedural risk profile differed between patients with advanced HF and patients in the comparator group. The higher in-hospital mortality rate in the advanced HF population in our study was likely associated with the higher morbidity burden in these patients. The increased rate of adverse events was likely associated with our higher-risk patient population and because the majority of the patients in the present study received a CRT device. In prior studies, CRT was independently associated with a higher risk of periprocedural complications compared with other types of devices.12-14

Risk models for in-hospital procedural complications after ICD placement have been developed to identify high-risk patients.12,14 The major risk factors strongly associated with adverse complications are as follows: undergoing initial device placement, dual-chamber device, NYHA class IV, blood urea nitrogen levels higher than 30 mg/dL (to convert to millimoles per liter, multiply by 0.357), and primary hospital admission not for ICD placement.12,14 Early complications have been associated with poor outcomes, including mortality. Lee et al15 reported an association between early major complications and mortality, with a 3- to 4-fold increased risk of death 6 months after device placement. In addition to the risk of procedural complications, our study found a relatively high risk of death (22% at 1 year) and a low median survival rate (3.5 years) in this patient population with advanced HF relative to the comparator group without advanced HF. Consistent with the findings in our comparator group, previous trials with patients having nonadvanced HF described an annual mortality rate of 8.5% (MADIT-II; Multicenter Automatic Defibrillator Implantation Trial II) and of 5% (SCD-HeFT; Sudden Cardiac Death in Heart Failure Trial). Furthermore, a subanalysis of MADIT II indicated a median survival rate of 8 years.15 Risk factors significantly associated with increased risk of all-cause mortality were NYHA class IV symptoms, ischemic heart disease, and diabetes. These factors may further refine risk stratification for all-cause mortality in clinical practice.

Limitations

This study has several limitations. First, the advanced HF population that underwent ICD or CRT-D placement was selected for the procedure despite a high-risk profile. Given the guideline recommendation that device therapy should generally be restricted to those with life expectancy exceeding a year, clinicians may identify candidates for treatment based on factors that are not measured in the ICD Registry. There are potentially unmeasured factors such that patients in this study may be healthier than the “average” patient with advanced HF. The presence of selection bias is reflected by the higher-than-expected life expectancy of our high-risk population. Second, the ICD Registry database is reported by participating sites; therefore, NYHA IV classification may be overreported, especially if the NYHA classification was abstracted during the HF hospital admission or through medical records. Although, NYHA class is intended to represent the degree of symptoms at the time of device placement, it is possible that NYHA class IV symptoms are temporary. Yet, our analysis found that NYHA class IV status was among the strongest factors associated with poor clinical outcomes and associated with the shortest median survival. These results likely indicate that classification as NYHA IV identified an advanced HF population and placements that occurred during nonelective hospitalizations. Third, the comparator group represented patients with a milder form of HF symptoms who currently had a class I indication for ICD or CRT-D placement. Given that patients with an ICD in the NCDR have been shown to have survival rates similar to patients with an ICD in the pivotal randomized clinical trials and that both had better survival rates than patients who received only medical therapy in the randomized clinical trials, our chosen cohorts very likely are representative of populations with or without advanced HF.16 Fourth, our analysis was confined to FFS Medicare beneficiaries and may not apply to a younger advanced HF population. Fifth, we did not censor patients at the time of LVAD placement and OHT receipt; however, the number of such events was too small to warrant this approach. Sixth, periprocedural outcomes were not adjudicated based on objective criteria but based on claims data, which is subject to more inaccuracy. Other limitations include inaccuracy in data collection and underreporting of complications.

Conclusions

In this analysis of the Medicare-linked NCDR ICD Registry, we found that 4.1% of recipients of an ICD for primary prevention had advanced HF and that these patients experience a nontrivial risk (22%) of all-cause mortality at 1 year, although the median survival following device placement was 3.5 years. Periprocedural complications occurred in approximately 1 in 25 patients and was primarily associated with in-hospital death and cardiac arrest. After adjustment, patients with NYHA class IV symptoms, ischemic heart disease, and diabetes had a higher risk of death. Finally, given the observed safety concerns, the poor long-term prognosis, and the high competing risk of death from progressive pump failure in patients hospitalized for HF, future randomized clinical trials should be conducted to test the safety and efficacy of the ICD and the CRT-D in patients with advanced HF.

Back to top
Article Information

Accepted for Publication: February 5, 2020.

Corresponding Author: Sana M. Al-Khatib, MD, MHS, Duke Clinical Research Institute, 200 Morris St, Durham, NC 27710 (alkha001@mc.duke.edu).

Published Online: March 25, 2020. doi:10.1001/jamacardio.2020.0391

Author Contributions: Dr Fudim and Mr Parzynski had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Fudim, Ambrosy, Friedman, Al-Khatib.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Fudim, Ali-Ahmed, Parzynski.

Critical revision of the manuscript for important intellectual content: Fudim, Parzynski, Ambrosy, Friedman, Pokorney, Curtis, Fonarow, Masoudi, Hernandez, Al-Khatib.

Statistical analysis: Fudim, Ali-Ahmed, Parzynski.

Administrative, technical, or material support: Ambrosy, Masoudi.

Supervision: Ambrosy, Friedman, Al-Khatib.

Conflict of Interest Disclosures: Dr Fudim reported being supported by grant 17MCPRP33460225 from the American Heart Association and receiving personal fees from Axon Therapies, Coridea, and Galvani Bioelectronics outside the submitted work. Mr Parzynski reported receiving salary support from a contract with the Yale Center for Outcomes Research and Evaluation via the American College of Cardiology during the conduct of the study. Dr Ambrosy reports significant research funding from Abbott Vascular Inc, Novartis, Amarin Corporation, the National Heart, Lung, and Blood Institute, National Institute on Aging, National Institute of Diabetes and Digestive and Kidney, and the Kaiser Permanente Northern California Community Benefit Program and modest reimbursement for travel from Novartis. Dr Friedman reported receiving educational grants from Abbott, Boston Scientific, and Medtronic; research grants from Abbott, Biosense Webster, Boston Scientific, Medtronic, and the National Cardiovascular Data Registry; consulting fees from Abbott and AtriCure outside the submitted work; and filing a provisional patent application pertaining to the use of electrogram analysis for cardiac resynchronization therapy optimization. Dr Pokorney reported receiving research grants from Bristol-Myers Squibb, Boston Scientific, Gilead Sciences Inc, Janssen Pharmaceuticals, Pfizer, and the US Food and Drug Administration and personal fees from Boston Scientific, Bristol-Myers Squibb, Janssen Pharmaceuticals, Medtronic, Pfizer, and Portola Pharmaceuticals. Dr Curtis has a contract with the American College of Cardiology for his role as Senior Medical Officer, NCDR; receives salary support from the American College of Cardiology, NCDR; receives funding from the Centers for Medicare & Medicaid Services to develop and maintain performance measures that are used for public reporting; and holds equity interest in Medtronic. Dr Fonarow reported receiving research grants from the National Institutes of Health and receiving personal fees from Abbott, Amgen, Bayer, Janssen Pharmaceuticals, Medtronic, and Novartis. Dr Masoudi reported receiving financial support from the American College of Cardiology outside the submitted work. Dr Hernandez reported receiving grants and personal fees from AstraZeneca, Merck, and Novartis; personal fees from Bayer and Boston Scientific; and grants from American Regent and Verily outside the submitted work. No other disclosures were reported.

Funding/Support: Funding for the statistical analysis was provided by the National Cardiovascular Data Registry.

Role of the Funder/Sponsor: The funder provided initial input on design and conduct of the study; however, the funder had no influence on the analysis and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Dr Fonarow is a section editor of JAMA Cardiology, and Dr Hernandez is an associate editor of JAMA Cardiology. They were not involved in any of the decisions regarding review of the manuscript or its acceptance.

References
1.
Abouezzeddine  OF, Redfield  MM.  Who has advanced heart failure?: definition and epidemiology.   Congest Heart Fail. 2011;17(4):160-168. doi:10.1111/j.1751-7133.2011.00246.x PubMedGoogle ScholarCrossref
2.
Al-Khatib  SM, Stevenson  WG, Ackerman  MJ,  et al.  2017 AHA/ACC/HRS guideline 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 on Clinical Practice Guidelines and the Heart Rhythm Society.   Heart Rhythm. 2018;15(10):e190-e252. doi:10.1016/j.hrthm.2017.10.035PubMedGoogle ScholarCrossref
3.
Lindenfeld  J, Feldman  AM, Saxon  L,  et al.  Effects of cardiac resynchronization therapy with or without a defibrillator on survival and hospitalizations in patients with New York Heart Association class IV heart failure.   Circulation. 2007;115(2):204-212. doi:10.1161/CIRCULATIONAHA.106.629261 PubMedGoogle ScholarCrossref
4.
Hammill  SC, Kremers  MS, Kadish  AH,  et al.  Review of the ICD Registry’s third year, expansion to include lead data and pediatric ICD procedures, and role for measuring performance.   Heart Rhythm. 2009;6(9):1397-1401. doi:10.1016/j.hrthm.2009.07.015PubMedGoogle ScholarCrossref
5.
Messenger  JC, Ho  KK, Young  CH,  et al; NCDR Science and Quality Oversight Committee Data Quality Workgroup.  The National Cardiovascular Data Registry (NCDR) data quality brief: the NCDR Data Quality Program in 2012.   J Am Coll Cardiol. 2012;60(16):1484-1488. doi:10.1016/j.jacc.2012.07.020 PubMedGoogle ScholarCrossref
6.
van Buuren  S.  Multiple imputation of discrete and continuous data by fully conditional specification.   Stat Methods Med Res. 2007;16(3):219-242. doi:10.1177/0962280206074463 PubMedGoogle ScholarCrossref
7.
Kadish  A, Dyer  A, Daubert  JP,  et al; Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators.  Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy.   N Engl J Med. 2004;350(21):2151-2158. doi:10.1056/NEJMoa033088 PubMedGoogle ScholarCrossref
8.
Moss  AJ, Zareba  W, Hall  WJ,  et al; Multicenter Automatic Defibrillator Implantation Trial II Investigators.  Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction.   N Engl J Med. 2002;346(12):877-883. doi:10.1056/NEJMoa013474 PubMedGoogle ScholarCrossref
9.
Cleland  JG, Daubert  JC, Erdmann  E,  et al; Cardiac Resynchronization-Heart Failure (CARE-HF) Study Investigators.  The effect of cardiac resynchronization on morbidity and mortality in heart failure.   N Engl J Med. 2005;352(15):1539-1549. doi:10.1056/NEJMoa050496 PubMedGoogle ScholarCrossref
10.
Bristow  MR, Saxon  LA, Boehmer  J,  et al; Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators.  Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure.   N Engl J Med. 2004;350(21):2140-2150. doi:10.1056/NEJMoa032423 PubMedGoogle ScholarCrossref
11.
van Rees  JB, de Bie  MK, Thijssen  J, Borleffs  CJ, Schalij  MJ, van Erven  L.  Implantation-related complications of implantable cardioverter-defibrillators and cardiac resynchronization therapy devices: a systematic review of randomized clinical trials.   J Am Coll Cardiol. 2011;58(10):995-1000. doi:10.1016/j.jacc.2011.06.007 PubMedGoogle ScholarCrossref
12.
Dodson  JA, Reynolds  MR, Bao  H,  et al; NCDR.  Developing a risk model for in-hospital adverse events following implantable cardioverter-defibrillator implantation: a report from the NCDR (National Cardiovascular Data Registry).   J Am Coll Cardiol. 2014;63(8):788-796. doi:10.1016/j.jacc.2013.09.079 PubMedGoogle ScholarCrossref
13.
Poole  JE, Gleva  MJ, Mela  T,  et al; REPLACE Registry Investigators.  Complication rates associated with pacemaker or implantable cardioverter-defibrillator generator replacements and upgrade procedures: results from the REPLACE registry.   Circulation. 2010;122(16):1553-1561. doi:10.1161/CIRCULATIONAHA.110.976076 PubMedGoogle ScholarCrossref
14.
Haines  DE, Wang  Y, Curtis  J.  Implantable Cardioverter-Defibrillator Registry risk score models for acute procedural complications or death after implantable cardioverter-defibrillator implantation.   Circulation. 2011;123(19):2069-2076. doi:10.1161/CIRCULATIONAHA.110.959676 PubMedGoogle ScholarCrossref
15.
Lee  DS, Krahn  AD, Healey  JS,  et al; Investigators of the Ontario ICD Database.  Evaluation of early complications related to de novo cardioverter defibrillator implantation insights from the Ontario ICD database.   J Am Coll Cardiol. 2010;55(8):774-782. doi:10.1016/j.jacc.2009.11.029 PubMedGoogle ScholarCrossref
16.
Al-Khatib  SM, Hellkamp  A, Bardy  GH,  et al.  Survival of patients receiving a primary prevention implantable cardioverter-defibrillator in clinical practice vs clinical trials.   JAMA. 2013;309(1):55-62. doi:10.1001/jama.2012.157182 PubMedGoogle ScholarCrossref
×