Use and Outcomes of Dual Chamber or Cardiac Resynchronization Therapy Defibrillators Among Older Patients Requiring Ventricular Pacing in the National Cardiovascular Data Registry Implantable Cardioverter Defibrillator Registry

Key Points Question Among patients undergoing implantable cardioverter defibrillator (ICD) implantation who are likely to require frequent right ventricular pacing, what are the outcomes of cardiac resynchronization therapy (CRT) compared with a dual chamber device, what is the variability in use of device type, and what are the trends in use of device type over time? Findings In this cohort study of 3100 Medicare patients undergoing first-time ICD implantation with a ventricular bradycardia pacing indication, the use of CRT was associated with a lower risk of heart failure hospitalization and 1-year mortality without an increase in procedural complications compared with a dual chamber device. Variability in use of device type was observed, and the rate of CRT implantation increased over time. Meaning In this study, CRT was associated with better outcomes than dual chamber devices among adult patients.


Introduction
Frequent right ventricular (RV) pacing may be deleterious, given that it results in electrical dyssynchrony, can exacerbate symptoms of heart failure, and is associated with incident atrial fibrillation. 1,2 Cardiac resynchronization therapy (CRT) improves morbidity and mortality among patients with low left ventricular ejection fraction (LVEF) and electrical dyssynchrony. 3 However, among patients without a class I indication for CRT but with an indication for ventricular pacing, it is unclear who might derive benefit from CRT vs a traditional dual chamber (DC) device.
The Biventricular Pacing for Atrioventicular Block and Systolic Dysfunction (BLOCK-HF) randomized clinical trial evaluated the use of CRT in a broader group of patients with reduced left ventricular systolic function with frequent anticipated ventricular pacing and found a lower incidence of a combined end point of time to death from any cause, heart failure visit, or increase in the left ventricular end-systolic volume index among patients randomized to CRT compared with patients receiving a traditional DC system. 4 While the results of this trial were published in 2013, it took until 2018 for an update and change in guidelines. Based on this trial, the 2018 American College of Cardiology, American Heart Association, and the Heart Rhythm Society guidelines include a class IIa recommendation for CRT among patients with an LVEF of 36% to 50% and an indication for permanent pacing who are expected to require frequent ventricular pacing. 5 The extent of practice variation and, importantly, outcomes among real-world patients with anticipated frequent ventricular pacing during this time frame is unknown.
Patients undergoing implantable cardioverter defibrillator (ICD) implantation frequently have conduction disturbances and an indication for ventricular pacing. The National Cardiovascular Data Registry (NCDR) ICD Registry provides a unique opportunity to analyze this group of patients. Accordingly, we sought to examine the trends, variability, and outcomes in the use of CRT defibrillator (CRT-D) and DC-ICD devices among patients undergoing ICD implantation from 2010 to 2016 with a ventricular pacing indication who did not have a class I indication for CRT based on LVEF.

Data Source
Data were obtained from the NCDR ICD Registry, the details of which have been reported previously. 6 The ICD Registry was designed to satisfy the requirements of the 2005 Centers for Medicare & Medicaid Services (CMS) coverage with evidence decision for primary prevention ICD implantation. In addition to expanded coverage, CMS mandated that data on all Medicare primary prevention implantations be entered into the NCDR ICD Registry until the data collection requirement ended on February 15, 2018. 6 The registry collects data from more than 1500 hospitals in the United States and included more than 1.3 million records as of the end of 2014. [7][8][9] Although CMS mandated that only primary prevention devices be entered into the registry, it is estimated that 90% of all ICD implantations are documented. 7 The registry uses a standardized data set and definitions, has requirements in place to ensure uniform data entry and transmission, and is subject to data quality checks. 10 All data submissions are evaluated for errors and completeness. This information is summarized in an automated report that is sent to the participants after each data submission. The NCDR audit program, which includes hospital medical record reviews and blinded data abstractions, serves as an additional mechanism to assess the accuracy of the data and enables participants to identify areas for improved data entry. Outcomes following discharge from the index hospitalization (including mortality, heart failure hospitalization, and complications) were obtained by linking NCDR registry files with Medicare inpatient fee-for-service claims, as previously described. 11 Statistical analysis was approved and completed by the Yale Center for Outcomes Research and

Study Population
Within the NCDR, 10 483 patients aged 65 years or older underwent a primary or secondary prevention ICD with a bradycardia pacing indication (second-or third-degree atrioventricular block or a PR interval of Ն300 ms) and could be linked to Medicare data.

Patient Demographic Characteristics
The patient population is described in

Outcomes
There was no evidence of difference between CRT-D and DC-ICD in the unadjusted incidence of

Institutional Variation and Temporal Trends
Hospital level variation in the use of DC-ICD or CRT-D was present among this patient cohort. The   To our knowledge, this is the first real-world study to evaluate the association of device therapy Consistent with prior clinical trials, in an analysis of contemporary real-world patients, we observed a lower incidence of adverse outcomes among the patients undergoing CRT-D compared with those with RV pacing (DC-ICD), which continued beyond a year of follow up. While the outcome of death was not detectable in BLOCK-HF, we observed an association between CRT-D and death despite the fact that patients in the current study tended to be older, with lower LVEF and greater burden of comorbidities, including prior myocardial infarction, hypertension, and diabetes. However, the findings of the present study are observational and thus the observed associations cannot be interpreted as causal.
We observed variation in the use of CRT or DC-ICD implantations among this cohort. For example, patients undergoing CRT-D implantation were more likely to have advanced atrioventricular block and NYHA class III or IV, prior heart failure hospitalization, and wider QRS duration. This is in concordance with prior literature showing improvement in heart failure symptoms with CRT and the need for chronic pacing. 3,4 In contrast, patients with a very long PR interval (Ն300ms) were more likely to undergo DC-ICD implantation. While efforts to minimize ventricular pacing in patients with a long PR interval and DC devices is possible with ventricular pacing avoidance algorithms, this has limitations regarding upper-rate behavior and without causing pacemaker syndrome. As such, these patients were included in the BLOCK-HF trial as likely needing chronic RV pacing and should be considered for CRT at the time of implantation. 4 We also observed hospital-level variation in the type of device implanted, which was independent of patient characteristics. Plausible explanations for this include familiarity with the  The diamond represents the mean, the box represents the 25th and 75th percentile, and the median is the line between 25th and 75th. Whisker bars indicate 95% CI, and circles, outliers. CRT-D DC-ICD more technically challenging LV lead placement or regional culture regarding the value of CRT.
Interestingly, nonelectrophysiology implanting clinicians were more likely to implant a DC-ICD, which can be seen to provide a basis for lack of specific training or familiarity to implanting an LV lead.
Describing this type of variation in care is an important first step in understanding the extent to which patient factors, compared with institutional factors, are associated with variation. Clustering of device therapy (ie, DC-ICD use more often at some hospitals) could provide opportunities for quality improvement programs to address hospital-level differences associated with variation.
A higher proportion of CRT-D implantations was noted among this cohort across the study years. The BLOCK-HF trial was published in 2013, and following this publication, there appears to have been an increase in the selection of CRT devices. 4 This study provides a perspective on the pace of adoption of CRT among this patient population based on the publication of an important trial.
The updated 2018 guidelines on the evaluation and treatment of patients with bradycardia and cardiac conduction delay were the first to provide recommendations on the use of CRT among patients with mildly reduced LVEF. The document provides a class IIa indication for physiologic ventricular activation, including CRT or His bundle pacing, among patients with LVEF between 36% and 50% who are expected to require ventricular pacing. 5 This is supported by level of evidence B-R (moderate-quality evidence from randomized trials) which is based on data from BLOCK-HF. While the current study has limitations inherent in observational research, including inability to determine causality, it adds to the limited data that support the more frequent use of CRT among patients with mildly reduced LVEF and anticipated high RV pacing requirement.
Further randomized and real-world investigations are warranted to both confirm the findings seen in BLOCK-HF, which would strengthen the recommendation to a class I indication among those with reduced LVEF, and to potentially expand the use of CRT among patients with normal LVEF and the need for frequent RV pacing, particularly those with evidence of electrical dyssynchrony or wide QRS intervals at baseline. Additionally, His or left bundle pacing may develop into an alternative or replacement to the traditional CRT system, and well-designed randomized trials are needed. Hospital variation in use of different modalities of cardiac resynchronization may be associated with procedural volumes and outcomes, and additional analysis evaluating these associations is warranted. Ultimately, the risk of implanting a device with higher procedural complexity and reduction in battery longevity requiring more frequent generator changes needs to be weighed against the potential benefits.

Limitations
This study has limitations that should be considered in the interpretation of this study. First, because this study is observational, we cannot exclude the possibility that unmeasured confounding variables influenced the association between device type and outcomes. Second, the amount of RV pacing in follow-up is not available in the ICD Registry; an understanding of this parameter would provide additional mechanistic insights into the association between CRT-D and outcomes. However, this is commonly the case in clinical practice: clinicians need to use their best judgement on what the expected RV pacing burden will be with the limited information available at the time of device implantation. Additionally, specific information about device interventions, including ICD shocks or antitachycardia pacing, is not available in the ICD Registry. Third, the position of the RV lead has been implicated in the development of cardiomyopathy, and this location is not available in the registry.
Fourth, while the NCDR is a voluntary reporting system that encompasses most ICD implantations, it is not all inclusive, and it is possible that the difference in device implantation varies significantly among centers that do not report to the NCDR.

Conclusions
In this cohort study of older patients in contemporary practice who underwent ICD implantation with a bradycardia pacing indication but without a class I indication for CRT, CRT use was associated with