Clinical Outcomes of Mortality, Readmissions, and Ischemic Stroke Among Medicare Patients Undergoing Left Atrial Appendage Closure via Implanted Device

Key Points Question What are the clinical outcomes among Medicare patients with atrial fibrillation undergoing left atrial appendage closure with an implanted device? Findings In this cohort study of 13 627 patients with atrial fibrillation who underwent left atrial appendage closure, the rate of hospitalization for stroke at 6 months (1.2%) was significantly lower than expected according to their CHA2DS2-VASc scores in the absence of anticoagulation therapy. The 1-year mortality rate (7.5%) was higher than that in previous randomized clinical trials, likely because of the inclusion of older patients with more comorbidities who were ineligible to receive warfarin. Meaning Left atrial appendage closure via an implanted device was associated with a decreased risk of stroke in patients with atrial fibrillation.


Introduction
Atrial fibrillation (AF) has emerged as a cardiovascular epidemic, with prevalence estimated to approach 12.1 million individuals in the United States by 2030. 1 It increases the risk for stroke by 4-to 5-fold and is an independent risk factor for ischemic stroke severity, recurrence, stroke-related disability, and mortality. Anticoagulation therapy is the cornerstone for stroke prevention in patients with AF, according to their CHA 2 DS 2-VASc (congestive heart failure; hypertension; ages Ն74 years [2 points]; diabetes; stroke, transient ischemic attack [TIA], or systemic embolism [2 points]; vascular disease; ages 65-74 years; sex [female]) risk scores. 2 Left atrial appendage closure (LAAC) devices have emerged as nonpharmacological alternatives for stroke prevention in patients with AF. 3 The Watchman device (Boston Scientific) was the first LAAC device that was compared with warfarin in 2 randomized clinical trials: PROTECT AF (Watchman Left Atrial Appendage System for Embolic Protection in Patients With Atrial Fibrillation; ClinicalTrials.gov identifier: NCT00129545) 4 and PREVAIL (Evaluation of the WATCHMAN LAA Closure Device in Patients With Atrial Fibrillation vs Long Term Warfarin Therapy; ClinicalTrials.gov identifier: NCT01182441). 5 The device received US Food and Drug Administration approval in 2015 for stroke prevention in patients with nonvalvular AF who are at high risk of stroke and are deemed suitable for short-term oral anticoagulation therapy but have appropriate rationale to seek nonpharmacologic alternatives to warfarin. To date, the real-world data outside clinical trials regarding the use and outcomes of LAAC devices are limited. To address this gap in knowledge, in this study, we describe the characteristics of patients undergoing LAAC via implanted device and examine the postprocedural outcomes of mortality, hospital readmissions, and ischemic stroke using Medicare data.

Study Population
The study was approved by the University of Iowa institutional review board, and the need for consent was waived because of the use of encrypted patient identifiers. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. 6 The study used Centers for Medicare & Medicaid Beneficiary Summary (enrollment) and We obtained patient dates of death from the Beneficiary Summary File, with mortality information available up to 1 year after the LAAC. We only included patients older than 65 years who had at least 1 year of Medicare enrollment to identify comorbidities from the inpatient claims preceding LAAC.
Patient demographic characteristics (age, sex, and race/ethnicity) were identified on the Beneficiary Summary File. Comorbid conditions included in the CHA 2 DS 2-VASc score were identified using an established algorithm according to International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and ICD-10 diagnoses on the LAAC hospital discharge record, as well as inpatient claims during the year preceding the LAAC admission. 7-9

Primary and Secondary Outcomes
The primary outcome was mortality during admission and within 1 year of LAAC. The secondary outcomes were readmissions within 30 days after LAAC discharge and reasons for readmission among patients discharged alive. To assess reasons for readmission, all first readmissions within 30 days of discharge were categorized according to the primary diagnosis code using the US Agency for Healthcare Research and Quality Clinical Classification System, 9 which collapses individual ICD-9-CM and ICD-10 codes into approximately 200 meaningful clinical categories. In addition, we evaluated the likelihood of admission for ischemic stroke or TIA within 180 days and the rate of admission for ischemic stroke or TIA per patient-year among patients discharged alive for whom at least 6 months of follow-up data were available. Admissions for ischemic stroke were identified by primary diagnosis on inpatient claims using ICD-10 codes starting with I63, I65, and I66 for ischemic stroke and G45 for TIA.

Statistical Analysis
Data analyses were conducted from January through August 2019. Summary statistics for patient characteristics were described as proportions for categorical variables (sex, race/ethnicity, and patient comorbidities), or using mean and SD for continuous variables (patient age and CHA 2 DS 2-VASc score). Mortality was expressed as the percentage of patients who died within 30, 90, and 180 days. Readmission rates for ischemic stroke and TIA were reported as the percentage of patients admitted for stroke or TIA within 180 days (among patients discharged alive after LAAC), as well as the rate of stroke or TIA admissions per patient-year of follow-up. In the latter estimate, patients were censored because of death or end of the observation period (December 31, 2017). Confidence intervals around the rate of stroke admissions were generated assuming a Poisson distribution.
Stroke admission rates were reported for a composite representing ischemic stroke or TIA, as well as separately for TIA and ischemic stroke. Finally, we assessed trends in mortality and readmission rates over time and by CHA 2 DS 2 -VASc score using the Cochran-Armitage test for trend. Statistical analyses were performed using SAS statistical software version 9.4 (SAS Institute). All statistical testing was 2-tailed, with P < .05 designating statistical significance. Hispanic patients (3.8%). Table 1 compares the patients undergoing LAAC in our study with those in the randomized clinical trials. 4,5 Compared with the randomized trials, the patients in the current The mortality and stroke readmission rates in the present study compared with those of previous studies 4,5,11-14 are shown in Table 3. The previous studies had ischemic stroke rates of 1.6%, 1.7%, and 1.1% and mortality rates of 3.6%, 5.0%, and 9.8%, respectively. We discuss these comparisons in greater detail later in the article.

Discussion
To our knowledge, this is the largest study outside clinical trial to assess the characteristics of patients with AF undergoing LAACs and the associated outcomes of mortality, readmissions, and stroke. We noted that, compared with previous randomized clinical trials, 4,5 our study had a higher percentage of women, and the patients had higher numbers of comorbidities and higher CHA 2 DS 2 -VASc scores.
The risk of readmission for stroke (1.2% at 6 months) was significantly lower than that expected patients in the control group treated with warfarin. In the LAAC groups of those studies, 4,5 the incidence of ischemic stroke or systemic embolism was 1.6% per 100 patient-years, and the overall mortality rate was 3.6% per 100 patient-years. Left atrial appendage closure was associated with lower risk of all-cause mortality, as well as cardiovascular or unknown death, compared with warfarin.
In the present study, readmission for ischemic stroke occurred in 111 of 9231 patients (1.2%) within 180 days after LAAC implantation. We observed a mortality rate of 7.5% at 1 year, and 28 patients (0.2%) died during the index admission for LAAC. There can be several explanations for the slightly higher rates of stroke and mortality in our study compared with the randomized clinical trials. 4,5 Our study included an older patient population (aged >65 years) with higher numbers of comorbidities and higher CHA 2 DS 2 -VASc scores. Patients in the randomized clinical trials had to be eligible to   Life Utilization), 14 reported 1-year outcome data for 1025 patients undergoing LAAC. The annual rate of ischemic stroke was 1.1% (15 of 1325 patient-years), and the 1-year mortality rate was 9.8%. Many of these patients were not eligible for anticoagulation therapy and underwent LAAC as a last resort option for stroke prevention. In both studies, 12-14 the rate of stroke was lower than that in the present study. However, as observed in the present study, the 1-year mortality rate was significantly higher than those in the randomized studies, 4,5 possibly because of the inclusion of patients with higher numbers of comorbidities.

Limitations
This study is retrospective and based on Medicare claims data, which have inherent limitations. We only included patients older than 65 years; therefore, our results cannot be extrapolated to younger patient populations. Specific details regarding LAAC via device implantation, immediate complications, preimplantation and postimplantation antithrombotic therapy, or postimplantation left atrial appendage imaging results were not available. The exact cause of death and whether it was related to the Watchman implantation procedure were unclear.

Conclusions
In this retrospective real-world study of Medicare patients older than 65 years with AF, LAAC via implanted device was associated with a decreased risk of admission for stroke compared with the expected risk without anticoagulation therapy. The risks of 1-year mortality and stroke at 6 months were higher than those in previous randomized clinical trials, likely because the population in the current study was older and had more comorbidities and higher CHA 2 DS 2 -VASc scores. Although LAAC provides an attractive nonpharmacological option for stroke prevention in patients with AF who are not candidates for long-term anticoagulation therapy, further prospective long-term realworld studies are needed to monitor its safety and efficacy.