Leadless versus transvenous single‐chamber ventricular pacemakers: 3 year follow‐up of the Micra CED study

The Micra Coverage with Evidence Development (CED) Study is a novel comparative analysis of Micra (leadless VVI) and transvenous single‐chamber ventricular pacemakers (transvenous VVI) using administrative claims data. To compare chronic complications, device reinterventions, heart failure hospitalizations, and all‐cause mortality after 3 years of follow‐up.


| INTRODUCTION
The Micra Coverage with Evidence Development (CED) Study is a novel comparative analysis of Micra (leadless VVI) and transvenous single-chamber ventricular pacemakers (transvenous VVI) using Medicare administrative claims data. This unique study allows the Centers for Medicare and Medicaid Services (CMS) to provide coverage for leadless pacemakers while continuing to assess the performance of the technology as it becomes more widely implemented in real-world practice. 1 The acute (30-day), 6-month, and 2-year outcomes have been previously reported, 2,3 with leadless VVI associated with higher rates of acute pericardial effusion (0.8% vs. 0.4%), but lower rates of chronic complications and reinterventions at both 6 months and 2 years of follow-up (31% lower rate of chronic complications [3.6% vs. 6.5%] and 38% lower rate of device reintervention [3.1% vs. 4.9%] at 2 years). The Micra CED Study results have been consistent with the findings of the Micra transcatheter pacing system investigational device exemption study 4 and Micra leadless pacemaker postapproval registry (Micra PAR), 5 demonstrating the utility of real-world data in complementing traditional clinical and registry studies as a source of evidence for insight into utilization, safety, and outcomes in general practice. The Micra CED Study will continue to follow leadless pacing patients in the Medicare population until CMS determines there is enough evidence to support or negate national coverage, making it a valuable source of evidence for comparative performance of leadless and transvenous pacemakers into the medium-and long-term. The study design also allows for the comparison of relevant health care utilization endpoints. In the present analysis, we compare for the first time rates of heart failure hospitalization between patients implanted with a leadless VVI versus transvenous VVI pacemaker. The objective of this analysis is to compare and report on chronic complications, device reinterventions, heart failure-related hospitalizations, and allcause mortality between leadless VVI and transvenous VVI pacemakers after 3-years of study follow-up.

| METHODS
The overall Micra CED Study design has been described previously. 2,3,6 The purpose of the study is to evaluate complications, utilization, and outcomes of the leadless VVI pacing system in the United States Medicare population. The primary objectives of the Micra CED study were to estimate the acute (30-day) complication rate and the 2-year survival rate associated with the leadless pacing system. The study uses manufacturer device registry information to identify Medicare beneficiaries implanted with a Micra leadless pacemaker (Model MC1VR01, Medtronic, Inc) using a previouslydescribed-linking algorithm. 6 The study also identifies a contemporaneous cohort of patients implanted with a transvenous VVI pacemaker from any manufacturer during the study period directly from their Medicare claims. The study was approved by the Western Institutional Review Board with a waiver of informed consent and is registered on ClinicalTrials.gov (NCT03039712).

| Data and cohort identification
For this analysis, Medicare claims (inpatient, outpatient, and carrier) and enrollment data were used to identify Medicare fee-for-service (FFS) beneficiaries implanted with a leadless VVI or transvenous VVI pacemaker from March 9, 2017 (the first date of Medicare coverage for leadless VVI pacemakers) through December 31, 2018. Pacemaker implants were identified using the International Classification of Diseases, 10th Revision, Procedure Coding System (ICD-10) and the Current Procedural Terminology (CPT) codes for inpatient and outpatient implants respectively (Supporting Information: Appendix Table 1). The index date for outcomes ascertainment was defined as the date of each patient's first observed pacemaker implant procedure during the study period. Patients with evidence of a prior cardiovascular implantable electronic device were excluded to facilitate comparison of de novo pacemaker implants and to reduce the risk of misattribution of outcomes related to a prior device.

Patients with less than 12 months of continuous enrollment in FFS
Medicare were excluded to ensure that patient baseline clinical and demographic characteristics could be adequately captured. Among the patients implanted with a transvenous VVI pacemaker, the cohort was limited to patients implanted in facilities with evidence of leadless implants over the same time period, to increase the likelihood that patients in the study cohort would have had access to and a chance to receive either system. See Supporting Information: Appendix Figure 1 for the study cohort diagram. Index (CCI) score was also calculated for each patient. 7 Characteristics of the implant encounter hospitalization were also identified, including whether the implant occurred in the inpatient or outpatient hospital setting, whether the patient was admitted for the implant procedure hospitalization through the emergency department, whether the patient was implanted during the weekend, the time from admission to the implant procedure, and whether the patient had a concomitant cardiac procedure (transcatheter aortic valve replacement or atrial fibrillation ablation) during the pacemaker implant procedure.

| Outcomes
This analysis focuses on chronic complications, device reinterventions, heart failure-related hospitalizations and all-cause mortality at 3 years as acute complications have been reported previously. 2 Chronic complications have been reported previously and include those prospectively defined using the relevant ICD-10 and CPT codes as complications most likely attributable to the implant procedure or device itself that may continue to occur or persist outside the time period of the implant procedure. These included embolism, thrombosis, device-related complications, including device breakdown, dislodgment, infection, and pocket complications, pericarditis, and hemothorax (Supporting Information: Appendix Table 2).
Device reinterventions were identified using the relevant procedure codes and were defined as system revision, lead revision or replacement, system replacement (e.g., replacing a leadless VVI with another leadless VVI), system removal, switch to the alternative type of system (i.e. switch from leadless VVI to transvenous VVI or transvenous VVI to leadless VVI), upgrade to a dual chamber system, or upgrade to a cardiac resynchronization therapy (CRT) device. A post hoc composite endpoint of any reinterventions requiring a wholly new device (composite of replacement, system switch, removal, upgrade to dual chamber system, upgrade to CRT) was also defined. The rationale for defining this post-hoc endpoint was that these types of reinterventions are particularly burdensome and costly to patients, providers, and payers. Date of death was determined from the Medicare Beneficiary Summary File.
A heart failure-related hospitalization was defined as an inpatient hospitalization with an ICD-10 diagnosis code for heart failure in the primary position on an inpatient claim following discharge from the implant procedure hospitalization or encounter.
A composite endpoint of heart failure hospitalization or death was also defined, as this is a common measure in the heart failure literature. 8 These heart failure hospitalization endpoints were newly-defined for the present analysis and were not previously reported on in the Micra CED Study. 2,3 Similar to reporting on upgrades to CRT, heart failure hospitalizations can provide evidence of and serve as a proxy measure for pacing-induced cardiomyopathy or potentially worsening heart failure among pacemaker patients and may be useful for evaluating long-term device performance. This novel endpoint is designed to be hypothesis generating and is the first health care utilizationrelated endpoint reported from the Micra CED Study. Medicare claims were available through December 31, 2020; patients alive and without an event were censored on that date.

| Statistical analysis
The statistical adjustments used in the analysis were prespecified and have been used for all comparative analyses of the Micra CED Study cohort. Propensity score overlap weights 9,10 were used to account for differences in baseline and encounter characteristics between the leadless-VVI and transvenous-VVI cohorts. Unadjusted and overlapweight adjusted 3-year complication, reintervention, and heart failure hospitalization rates were estimated using the cumulative incidence function. Fine-Gray competing risk models were used to compare the unadjusted and adjusted risk for 3-year chronic complications, device reinterventions, and heart failure hospitalizations between study groups, and Cox proportional hazards models were used to compare all-cause mortality and the composite of heart failure hospitalization and all-cause mortality through 3 years. A sub-analysis of the heart failure hospitalization endpoints was also conducted among patients in the study cohort without a history of heart failure at baseline and a second sub-analysis was conducted among patients in the study cohort with atrial fibrillation at baseline. Because heart failure hospitalizations are not a primary or pre-specified secondary endpoint of the Micra CED Study, the inclusion of this endpoint is designed to be hypothesis-generating. Thus, no statistical correction for the additional endpoint was made and all endpoints were evaluated at a significance level of p < .05 and all p values were 2-tailed. Events occurring in between one and 10 patients were suppressed to protect beneficiary privacy as required by CMS. 11 All statistical analyses were conducted in SAS version 9.4 (SAS Institute).

| RESULTS
There were 6219 leadless and 10 212 transvenous de novo implant procedures identified during the study period contributing to the analysis cohort (Table 1). Patient baseline characteristics of this cohort have been previously reported. 3 Compared with transvenous, patients implanted with a leadless VVI pacemaker were more likely to have ESRD (12.0% vs. 2.3%, p < .001), renal dysfunction (48.8% vs. 42.1%, p < .001), and a higher mean CCI score (5.1 ± 3.4 vs. 4.6 ± 3.0, p < .001). Mean follow-up time for patients implanted with a leadless VVI pacemaker was 675 days, compared to 727 days for transvenous (p < .001). After applying the propensity score overlap weights, all measured baseline and encounter characteristics were well balanced with all standardized mean differences near zero. There were 2937 patients implanted with a leadless VVI pacemaker and 4821 patients implanted with a transvenous VVI pacemaker in the sub-analysis of heart failure hospitalization endpoint analysis cohort without a prior history of heart failure.   Table 3.

| Reinterventions
Reintervention rates were also significantly lower in the patients implanted

| Heart failure hospitalizations and all-cause mortality
Heart failure hospitalization rates were slightly lower among patients implanted with a leadless VVI pacemaker compared to transvenous in the overall patient cohort (adjusted, 19.9% vs. 22.0%, p = .005) as well as among patients without prior history of heart failure (adjusted, 11.2% vs. 13.6%), p = .003) (Supporting Information: Appendix Table 4). In the time-to-event models, patients with a leadless pacemaker had a slightly, but significantly lower, rate of heart failure hospitalization compared with patients implanted with a transvenous pacemaker through 3 years ( Figure 1C; unadjusted HR 0.90; 95% CI 0.84-0.97, p = .006; adjusted HR 0.90; 95% CI 0.83-0.97, p = .005).
Among patients without history of heart failure, the lower rates among patients implanted with a leadless VVI pacemaker were more  However, our prior analyses suggest that this probability is low, 6 and, if anything, claims-based studies tend to overestimate adverse events. 14 We would also not expect this to have a differential impact between the two study arms. Second, as with any observational study, the possibility of residual confounding following statistical adjustment for measured confounders cannot be completely eliminated. Third, because our study does not include device interrogation data, we are unable to assess variables such as programmed lower rates, pacing thresholds, and battery longevity which may be of particular interest when assessing the need for device reintervention. Fourth, we are unable to capture variables related to AV synchrony, pacing tip location, QRS duration, pacing indication, and reason for heart failure hospitalization, which could help explain the observed relationship between pacemaker type and time to heart failure hospitalization.
Finally, due to data availability, this analysis is limited to the Medicare FFS population and does not capture outcomes beyond December 31, 2020.

| CONCLUSION
In a real-world study of the United States Medicare patients, the leadless VVI pacemaker was associated with a 32% lower rate of chronic complications (4.9% vs. 7.1%) and a 41% lower rate of device reinterventions (3.6% vs. 6.0%) at 3 years. Rates of heart failure hospitalization were slightly lower among leadless VVI patients, and all-cause mortality rates were similar among leadless VVI and transvenous VVI patients at 3 years, suggesting no trade-off between lower rates of device reintervention and chronic right ventricular-only pacing outcomes for patients.
Infections rates were remarkably lower in the leadless group.
The Micra CED Study continues to illustrate the feasibility of utilizing real-world data to generate evidence measuring the safety and effectiveness of new technology and continues to complement existing clinical evidence demonstrating the benefits of leadless pacing.

AUTHOR CONTRIBUTIONS
All authors met the IJCME criteria for authorship. All authors had access to the data according to the terms of the data use agreement with CMS and all authors fully reviewed and vouch for the accuracy of reported results. Academic authors had independent final review and approval of publication.