Key Points
Question
What are the sex differences in outcomes among patients undergoing percutaneous left atrial appendage occlusion (LAAO)?
Findings
In this cohort study of 49 357 patients undergoing LAAO, women were more likely than men to experience in-hospital major adverse events, prolonged hospital stay, and death.
Meaning
These findings suggest that strategies are needed to reduce the increased risk of LAAO implant–related adverse events for women.
Abstract
Importance
Left atrial appendage occlusion (LAAO) has emerged as an alternative to anticoagulation for select patients with atrial fibrillation; however, women have been underrepresented in clinical trials of LAAO, and sex-specific subanalyses are limited.
Objective
To evaluate the sex differences in the baseline characteristics of patients undergoing LAAO implant and in the in-hospital outcomes after LAAO implant.
Design, Setting, and Participants
A total of 49 357 patients in the National Cardiovascular Data Registry LAAO Registry undergoing LAAO with the Watchman device between January 1, 2016, and June 30, 2019, were included in this study.
Exposure
Female or male sex.
Main Outcomes and Measures
The primary outcomes were aborted or canceled procedure, major adverse event, any adverse event, prolonged hospital stay longer than 1 day, and death. Unadjusted and multivariable adjusted logistic regression analyses were performed to assess sex differences in in-hospital adverse events.
Results
In this cohort study of 49 357 patients (mean [SD] age, 76.1 [8.0] years), 20 388 women (41.3%) and 28 969 (58.7%) men underwent LAAO. Compared with men, women were older and had a higher prevalence of paroxysmal atrial fibrillation, prior stroke, and uncontrolled hypertension but a lower prevalence of congestive heart failure, diabetes, and coronary artery disease. After multivariable adjustment, there were no differences in aborted or canceled procedures between women and men (613 [3.0%] vs 851 [2.9%]; odds ratio [OR] 1.01, 95% CI, 0.90-1.13). Women were more likely than men to experience any adverse event (1284 [6.3%] vs 1144 [3.9%]; P < .001; OR, 1.63; 95% CI, 1.49-1.77; P < .001) or major adverse event (827 [4.1%] vs 567 [2.0%]; P < .001; OR, 2.06; 95% CI, 1.82-2.34; P < .001) owing to pericardial effusion requiring drainage (241 [1.2%] vs 144 [0.5%]) or major bleeding (349 [1.7%] vs 244 [0.8%]). Women were also more likely than men to experience a hospital stay longer than 1 day (3272 [16.0%] vs 3355 [11.6%]; P < .001; adjusted OR, 1.46; 95% CI, 1.38-1.54; P < .001) or death (adjusted OR, 2.01; 95% CI, 1.31-3.09; P = .001), although death was rare and absolute differences were minimal (58 [0.3%] vs 37 [0.1%]; P < .001).
Conclusions and Relevance
This study suggests that, compared with men, women have a significantly higher risk of in-hospital adverse events after LAAO. Further research aimed at risk reduction, particularly strategies to reduce the risk of pericardial effusion and major bleeding, in women undergoing LAAO is warranted.
This cohort study uses data from the National Cardiovascular Data Registry LAAO Registry to evaluate the sex differences in the baseline characteristics of patients undergoing left atrial appendage occlusion implant and in the in-hospital outcomes after left atrial appendage occlusion implant.
Introduction
Atrial fibrillation (AF), the most common sustained arrhythmia in both men and women, is associated with significant mortality and morbidity, including a 5-fold increase in stroke risk.1,2 Long-term anticoagulation therapy has been essential for stroke prevention in those with nonvalvular AF who are at moderate or high risk of stroke.3 Despite the proven effectiveness of anticoagulation for the prevention of AF-related thromboembolism, some patients are not candidates for long-term oral anticoagulation owing to significant adverse effects, issues with adherence, and quality of life concerns. Based on data supporting the left atrial appendage (LAA) as the most important cardiac source of thromboemboli in nonvalvular AF, a nonpharmacologic approach to LAA occlusion (LAAO) using the Watchman device (Boston Scientific) has emerged as a valid alternative to anticoagulation.4 Based on the results of 2 trials, the US Food and Drug Administration (FDA) approved the Watchman device in 2015 for patients with AF who have an appropriate reason to seek a nonpharmacologic alternative to long-term anticoagulation.5,6
Important sex-specific differences in the epidemiology, presentation, and prognosis of AF have been described.7 Women with AF generally experience worse symptoms and poorer quality of life and have a higher risk of stroke and death than men.8 Furthermore, previous studies of invasive cardiac procedures have demonstrated that women are at a higher risk than men for adverse events.9,10 Because only 30.0% (259 of 870) of the enrolled cohort in both LAAO trials were women, sex-specific subanalyses regarding outcomes remain inconclusive.
Using data from the National Cardiovascular Data Registry (NCDR) LAAO Registry, our study aimed to compare sex differences among patients undergoing LAAO implant in terms of patient characteristics, in-hospital adverse events, and prevalence and type of postimplant antithrombotic therapy.
Methods
Data Source
The patients included in this study were enrolled in the NCDR LAAO Registry, which has been previously described.11 In brief, the LAAO Registry was launched in late December 2015 by a multistakeholder team, including the NCDR, FDA, Society for Coronary Angiography and Interventions, Centers for Medicare & Medicaid Services (CMS), and Boston Scientific. The LAAO Registry was designed to function as the formal postmarketing surveillance vehicle required by the FDA and is currently the only registry approved by the CMS to satisfy the coverage decision data submission requirements. Beginning April 1, 2016, US hospitals were required to submit data on all procedures using the Watchman device into the LAAO Registry to qualify for Medicare reimbursement. Hospitals were encouraged to submit data on implants regardless of patients’ insurance status. Demographic information, clinical information, device implant information, adjudicated in-hospital adverse events during the index hospitalization, and discharge medication were collected using standardized definitions. Data were submitted by participating hospitals using a rigorous data quality reporting process to ensure complete and accurate submissions. The NCDR programs use a multistage data quality process, including quality checks on submitted data, outlier analysis, and medical record audits.11,12
Study Population
Between January 1, 2016, and June 30, 2019, a total of 597 hospitals submitted data to the LAAO Registry on 52 560 patients who underwent Watchman procedures. We identified a final cohort of 49 357 patients after exclusion for missing sex (n = 34) and non-Watchman LAAO devices (n = 3169).
Outcomes
First, in-hospital events were compared between women and men. Events that occurred during or after the implant procedure until the time of hospital discharge were collected using the LAAO Registry data collection form, versions 1.1 and 1.2. Periprocedural information, including aborted or canceled procedure, any adverse event, major adverse event, length of hospital stay more than 1 day, and death during the index hospitalization, was collected. Aborted procedures were defined as those in which venous access was performed but a device was not ultimately deployed. A canceled procedure was defined as a procedure in which no venous access was performed. Any in-hospital major adverse event included death, cardiac arrest, ischemic stroke, hemorrhagic stroke, undetermined stroke, transient ischemic attack, intracranial hemorrhage, systemic arterial embolism, major bleeding (access site bleeding, hematoma, gastrointestinal bleeding, retroperitoneal bleeding, other hemorrhage [nonintracranial], or hemothorax requiring hospitalization and/or causing a decrease in hemoglobin level >2 g/dL [to convert to grams per liter, multiply by 10.0] and/or requiring blood transfusion that was not hemorrhagic stroke), major vascular complication (arteriovenous fistula or pseudoaneurysm requiring a fibrin injection, angioplasty, or surgical repair), myocardial infarction, pericardial effusion requiring drainage, and device embolization. Further details regarding specific adverse events have been published previously.11
Second, postimplant antithrombotic therapy was compared between women and men. All medications, including prescriptions and over-the-counter medications, were reconciled prior to hospital discharge. Outcomes included use of direct oral anticoagulant only (including apixaban, dabigatran, rivaroxaban, or edoxaban), warfarin only, single antiplatelet therapy (aspirin, clopidogrel, prasugrel, or ticagrelor), dual antiplatelet therapy (combination of antiplatelet agent and P2Y12 inhibitor), triple therapy (dual antiplatelet therapy plus anticoagulation, including warfarin or direct oral anticoagulant), and oral anticoagulation plus antiplatelet agent.
Statistical Analysis
Baseline demographic and clinical factors are presented as numbers and percentages for categorical variables and median values and interquartile ranges or mean (SD) values for continuous variables. Categorical variables were compared using the χ2 test, and continuous variables were compared using the Wilcoxon rank sum test or the t test as appropriate. Missing continuous variables were imputed with the overall median value. For missing categorical variables, the most common category of each variable was imputed.
Descriptive, unadjusted outcomes stratified by sex were summarized by the numbers and percentages of events. Then, unadjusted and adjusted multivariable logistic regression was used to obtain odds ratios (ORs) with a 95% CI for women vs men (reference) for in-hospital outcomes (aborted or canceled procedure, hospital length of stay >1 day, death, any adverse event or major adverse event, and postimplant antithrombotic therapy). All tests were 2-sided and P ≤ .05 was considered statistically significant. Variables in the multivariable model were chosen based on both clinical risk-adjusted variable selection and backward elimination. Patient characteristics that differed between sexes in the univariate analysis (at P < .10) were entered into a logistic regression model with backward selection using a P ≤ .05 for removal during the selection process. The final covariates in the multivariable model included age, race/ethnicity, body mass index, congestive heart failure, hypertension, type 1 or 2 diabetes, stroke, vascular disease, glomerular filtration rate, prior P2Y12 inhibitor prescription, prior oral anticoagulant prescription, coronary artery disease, left ventricular ejection fraction, and hospital region. All potential confounding variables were well represented and collected as part of the NCDR LAAO registry. All analyses were performed with the SAS statistical package, version 9.4 (SAS Institute Inc).
Results
Baseline Characteristics
Of the 49 357 patients, 20 388 (41.3%) were women and 28 969 (58.7%) were men. Baseline characteristics are presented in Table 1. Although differences in several baseline characteristics were statistically significant owing to the large sample size, the absolute differences were modest. Compared with men, women were older (mean [SD], 76.5 [7.9] vs 75.8 [8.2] years) with a higher mean (SD) CHA2DS2-VASc (cardiac failure or dysfunction, hypertension, age 65-74 [1 point] or ≥75 years [2 points], diabetes mellitus, and stroke, transient ischemic attack or thromboembolism [2 points]–vascular disease, and sex category [female]) score (5.3 [1.5] vs 4.5 [1.4]) and were more likely to have paroxysmal AF (12 029 [59.0%] vs 14 488 [50.0%]), prior stroke (5288 [26.0%] vs 7104 [24.6%]), and uncontrolled hypertension (systolic blood pressure ≥160 mm Hg; 5953 [29.3%] vs 7722 [26.7%]) and were less likely to have congestive heart failure (7025 [34.5%] vs 11 493 [39.7%]), diabetes (7088 [34.8%] vs 11 468 [39.7%]), and coronary artery disease (7137 [35.0%] vs 16 122 [55.7%]).
Table 1. Baseline Characteristics of Patients.
| Characteristic | Patients, No. (%) | P value | |
|---|---|---|---|
| Women (n = 20 388) | Men (n = 28 969) | ||
| Demographic information | |||
| Age, mean (SD), y | 76.5 (7.9) | 75.8 (8.2) | <.001 |
| Race/ethnicity | |||
| White | 18 701 (91.7) | 27 068 (93.4) | <.001 |
| Black | 1130 (5.5) | 1187 (4.1) | <.001 |
| Asian | 338 (1.7) | 434 (1.5) | .16 |
| Othera | 108 (0.5) | 135 (0.5) | .32 |
| Hispanic | 789 (3.9) | 1024 (3.6) | .06 |
| Insurance payer | |||
| Private | 12 301 (60.3) | 18 444 (63.7) | <.001 |
| Medicare | 17 681 (86.7) | 24 553 (84.8) | <.001 |
| Medicaid | 1411 (6.9) | 1121 (3.9) | <.001 |
| State-specific plan | 185 (0.9) | 200 (0.7) | .007 |
| Other | 792 (3.9) | 1392 (4.8) | <.001 |
| Admitted for LAAO procedure | 19 617 (96.5) | 27 942 (96.8) | .08 |
| Patient history and risk factors | |||
| CHA2DS2-VASc score, mean (SD) | 5.3 (1.5) | 4.5 (1.4) | <.001 |
| Congestive heart failure | 7025 (34.5) | 11 493 (39.7) | <.001 |
| NYHA class I | 1759 (8.6) | 2903 (10.0) | <.001 |
| NYHA class II | 3247 (15.9) | 5280 (18.2) | <.001 |
| NYHA class III | 1458 (7.2) | 2433 (8.4) | <.001 |
| NYHA class IV | 84 (0.4) | 167 (0.6) | .01 |
| Hypertension | 18 736 (92.0) | 26 719 (92.3) | .28 |
| Type 1 or 2 diabetes | 7088 (34.8) | 11 468 (39.7) | <.001 |
| Prior stroke | 5288 (26.0) | 7104 (24.6) | < 001 |
| Prior transient ischemic attack | 3044 (15.0) | 3900 (13.5) | <.001 |
| Prior thromboembolic event | 3855 (19.0) | 4919 (17.0) | <.001 |
| Vascular disease | 7530 (37.0) | 15 357 (53.1) | <.001 |
| Prior myocardial infarction | 2829 (13.9) | 6619 (22.8) | <.001 |
| Peripheral arterial disease | 2402 (11.8) | 4331 (15.0) | <.001 |
| Known aortic plaque | 646 (3.2) | 1216 (4.2) | <.001 |
| HAS-BLED score, mean (SD) | 2.96 (1.1) | 3.04 (1.1) | <.001 |
| Uncontrolled hypertension | 5953 (29.3) | 7722 (26.7) | <.001 |
| Abnormal renal function | 2431 (12.0) | 4455 (15.4) | <.001 |
| Abnormal liver function | 616 (3.0) | 957 (3.3) | .08 |
| Prior stroke | 5584 (27.5) | 7588 (26.3) | .003 |
| Ischemic | 3203 (57.6) | 4163 (55.1) | .005 |
| Hemorrhagic | 1286 (23.1) | 2131 (28.2) | <.001 |
| Undetermined | 1324 (23.8) | 1738 (23.0) | .28 |
| Prior bleeding | 14 245 (70.0) | 20 287 (70.1) | .76 |
| Labile INR | 2290 (11.3) | 3071 (10.6) | .03 |
| Alcohol use | 579 (2.9) | 2139 (7.4) | <.001 |
| Antiplatelet medication use | 5042 (24.8) | 8225 (28.5) | <.001 |
| Nonsteroidal anti-inflammatory drug use | 5701 (28.1) | 9072 (31.4) | <.001 |
| Indication for occlusion | |||
| Increased thromboembolic risk | 13 188 (64.7) | 18 625 (64.3) | .37 |
| History of major bleeding | 12 908 (63.3) | 18 524 (63.9) | .15 |
| High fall risk | 8111 (39.8) | 9692 (33.5) | <.001 |
| Labile INR | 1850 (9.1) | 2408 (8.3) | .003 |
| Patient preference | 6924 (34.0) | 10 165 (35.1) | .01 |
| Noncompliance with anticoagulation therapy | 686 (3.4) | 931 (3.2) | .35 |
| Other history and risk factors | |||
| Clinically relevant prior bleeding | 13 922 (68.4) | 19 881 (68.8) | .36 |
| Intracranial | 2138 (10.5) | 3555 (12.3) | <.001 |
| Epistaxis | 1260 (6.2) | 1877 (6.5) | .18 |
| Gastrointestinal | 8874 (43.5) | 11 484 (39.6) | <.001 |
| Other | 2808 (13.8) | 4582 (15.8) | <.001 |
| Fall risk | 9012 (44.4) | 10 709 (37.1) | <.001 |
| Genetic coagulopathy | 197 (1.0) | 232 (0.8) | .05 |
| Cardiomyopathy | 2924 (14.4) | 7307 (25.3) | <.001 |
| Ischemic | 1010 (5.0) | 4138 (14.3) | <.001 |
| Nonischemic | 1267 (6.2) | 2331 (8.0) | <.001 |
| Restrictive | 21 (0.1) | 35 (0.1) | .56 |
| Hypertrophic | 216 (1.1) | 218 (0.8) | < 001 |
| Other | 399 (2.0) | 649 (2.2) | .03 |
| Chronic lung disease | 4524 (22.2) | 6006 (20.8) | < 001 |
| Coronary artery disease | 7137 (35.0) | 16 122 (55.7) | <.001 |
| Sleep apnea | 4194 (20.6) | 8596 (29.7) | <.001 |
| Sleep apnea treatment | 2906 (71.2) | 6144 (73.4) | .01 |
| AF classification | |||
| Paroxysmal | 12 029 (59.0) | 14 488 (50.0) | <.001 |
| Persistent | 3920 (19.2) | 6452 (22.3) | <.001 |
| Longstanding persistent | 1585 (7.8) | 2789 (9.6) | <.001 |
| Permanent | 2773 (13.6) | 5104 (17.6) | <.001 |
| Valvular | 116 (0.6) | 144 (0.5) | .28 |
| Attempt at AF termination | 8089 (39.7) | 11 614 (40.1) | .35 |
| Pharmacologic cardioversion | 4258 (53.1) | 5382 (46.7) | <.001 |
| Direct current cardioversion | 4080 (50.8) | 6820 (59.1) | <.001 |
| Catheter ablation | 3216 (39.9) | 4611 (39.8) | .92 |
| Surgical ablation | 87 (1.1) | 182 (1.6) | .004 |
| Atrial flutter | 2354 (11.6) | 4474 (15.5) | <.001 |
| Typical | 1511 (7.4) | 3047 (10.5) | <.001 |
| Atypical | 644 (3.2) | 1124 (3.9) | <.001 |
| Attempt at atrial flutter termination | 1591 (68.3) | 3108 (69.7) | .21 |
| Pharmacologic cardioversion | 463 (29.3) | 801 (25.9) | .01 |
| Direct current cardioversion | 508 (32.2) | 1103 (35.7) | .02 |
| Ablation | 1242 (78.3) | 2483 (80.1) | .15 |
| Previous cardiac structural intervention | 1316 (6.5) | 2312 (8.0) | <.001 |
| LAA orifice maximal width, mean (SD) | 21.3 (4.2) | 21.7 (4.5) | <.001 |
| Physical examination and laboratory results, mean (SD) | |||
| Body mass indexb | 30.3 (13.8) | 29.9 (8.3) | <.001 |
| Hemoglobin level, g/dL | 12.3 (1.8) | 13.0 (2.2) | <.001 |
| INR | 1.4 (1.0) | 1.4 (0.9) | .01 |
| Creatinine, mg/dL | 1.2 (0.9) | 1.4 (1.2) | <.001 |
| Albumin, g/dL | 3.8 (0.5) | 3.8 (0.5) | .10 |
| Platelet count, ×103/µL | 225 647 (74 004) | 194 924 (67 272) | <.001 |
| Preimplant medications | |||
| Direct oral anticoagulant only | 3264 (11.3) | 2584 (12.7) | <.001 |
| Warfarin only | 2801 (9.7) | 2217 (10.9) | <.001 |
| Single antiplatelet agent | 7051 (24.3) | 4589 (22.5) | <.001 |
| Dual antiplatelet therapy | 1136 (3.9) | 663 (3.3) | <.001 |
| Triple therapy | 461 (1.6) | 182 (0.9) | <.001 |
| Oral anticoagulant agent plus single antiplatelet agent | 6153 (21.2) | 3589 (17.6) | <.001 |
| Other | 8103 (28.0) | 6564 (32.2) | <.001 |
| Procedural characteristics | |||
| Intraprocedural anticoagulation | 20 059 (98.4) | 28 463 (98.3) | .31 |
| Heparin | 19 945 (99.5) | 28 303 (99.5) | .88 |
| Bivalirudin | 92 (0.5) | 127 (0.4) | .84 |
| Other | 737 (3.7) | 1113 (3.9) | .18 |
| Anticoagulation reversal | 12 168 (61.0) | 16 940 (59.8) | .01 |
| Hospital characteristics | |||
| Hospital location | |||
| Rural | 1601 (7.9) | 2283 (7.9) | .91 |
| Suburban | 5500 (27.0) | 8009 (27.6) | .10 |
| Urban | 13 287 (65.2) | 18 677 (64.5) | .11 |
| US Region | |||
| Northeast | 3282 (16.1) | 5259 (18.2) | <.001 |
| West | 4510 (22.1) | 6381 (22.0) | .80 |
| Midwest | 4617 (22.6) | 6291 (21.7) | .01 |
| South | 7968 (39.1) | 11 030 (38.1) | .02 |
| Type of facility | |||
| Government | 231 (1.1) | 328 (1.1) | .10 |
| Private | 16 168 (79.3) | 22 381 (77.3) | <.001 |
| University | 3989 (19.6) | 6260 (21.6) | <.001 |
| Teaching hospital | 12 927 (63.4) | 18 922 (65.3) | <.001 |
| Hospital No. of certified beds, median (range) | 500 (341-683) | 500 (341-691) | .25 |
| Annual LAAO procedure volume, median (range) | 26 (1-60) | 26.0 (1-60) | .45 |
Abbreviations: AF, atrial fibrillation; CHA2DS2-VASc, (cardiac failure or dysfunction, hypertension, age 65-74 [1 point] or ≥75 years [2 points], diabetes mellitus, and stroke, transient ischemic attack or thromboembolism [2 points]–vascular disease, and sex category [female]); HAS-BLED, hypertension, abnormal renal/liver function, stroke, bleeding history or predisposition, labile INR, elderly, drugs/alcohol concomitantly; INR, international normalized ratio; LAAO, left atrial appendage occlusion; NYHA, New York Heart Association.
SI conversion factors: To convert hemoglobin to grams per liter, multiply by 10.0; creatinine to micromoles per liter, multiply by 88.4; albumin to grams per liter, multiply by 10.0; and platelets to ×109/L, multiply by 1.0.
Other races include American Indian or Alaskan Native, Pacific Islander, or multiple origins.
Calculated as weight in kilograms divided by height in meters squared.
The most common indications for LAAO in both women and men were increased thromboembolic risk (13 188 [64.7%] vs 18 625 [64.3%]) and history of major bleeding (12 908 [63.3%] vs 18 524 [63.9%]), with no difference between groups (Table 1). Women were more likely than men to have a high fall risk as an indication for LAAO (8111 [39.8%] vs 9692 [33.5%]; P < .001).
Outcomes
The unadjusted rates of in-hospital adverse events are presented in Table 2. An aborted or canceled procedure occurred for 613 women (3.0%) and 851 men (2.9%). The rates of any adverse event were higher for women than men (1284 [6.3%] vs 1144 [3.9%]; P < .001). The rates of major in-hospital adverse events were more than 2-fold higher for women than men (827 [4.1%] vs 567 [2.0%]; P < .001). Specific complications, including major bleeding (349 [1.7%] vs 244 [0.8%]; P < .001) and pericardial effusion requiring drainage (241 [1.2%] vs 144 [0.5%]; P < .001), were significantly more common for women than men. Length of hospitalization more than 1 day was more frequent for women than men (3273 [16.0%] vs 3355 [11.6%]; P < .001). In-hospital death was rare, although statistically higher for women than men (58 [0.3%] vs 37 [0.1%]; P < .001).
Table 2. Crude Comparisons of Prevalence of In-Hospital Adverse Events Stratified by Sex.
| Adverse event type | Patients, No. (%) | P value | |
|---|---|---|---|
| Women (n = 20 388) | Men (n = 28 969) | ||
| In-hospital outcomes | |||
| Aborted or canceled procedure | 613 (3.0) | 851 (2.9) | .66 |
| Any adverse event | 1284 (6.3) | 1144 (3.9) | <.001 |
| Any major adverse eventa | 827 (4.1) | 567 (2.0) | <.001 |
| Hospital stay >1 d | 3272 (16.0) | 3355 (11.6) | <.001 |
| Death | 58 (0.3) | 37 (0.1) | <.001 |
| Individual adverse events | |||
| Cardiovascular | |||
| Cardiac arrest | 62 (0.3) | 40 (0.1) | <.001 |
| Myocardial infarction | 8 (0.04) | 7 (0.02) | .34 |
| Pericardial effusion (no intervention required) | 245 (1.2) | 256 (0.9) | <.001 |
| Pericardial effusion with tamponade (requiring percutaneous drainage) | 241 (1.2) | 144 (0.5) | <.001 |
| Left atrial thrombus | 10 (0.05) | 15 (0.05) | .89 |
| Pericarditis | 25 (0.1) | 24 (0.1) | .17 |
| Air embolism | 20 (0.1) | 13 (0.04) | .02 |
| Neurologic | |||
| Ischemic stroke | 30 (0.1) | 27 (0.1) | .08 |
| Hemorrhagic stroke | 4 (0.02) | 2 (0.007) | .21 |
| Undetermined stroke | 1 (0.005) | 4 (0.01) | .17 |
| Transient ischemic attack | 10 (0.05) | 8 (0.1) | .22 |
| Intracranial hemorrhage (other than hemorrhagic stroke) | 2 (0.01) | 1 (0.003) | .37 |
| Systemic | |||
| Arterial thromboembolism (other than stroke) | 0 | 1 (0.003) | .40 |
| Deep vein thrombosis | 9 (0.04) | 5 (0.02) | .08 |
| Bleeding | |||
| Major bleeding | 349 (1.7) | 244 (0.8) | <.001 |
| Access site bleeding | 93 (0.5) | 91 (0.3) | .01 |
| Hematoma | 159 (0.8) | 111 (0.4) | <.001 |
| Gastrointestinal bleeding | 54 (0.3) | 38 (0.1) | .001 |
| Retroperitoneal bleeding | 14 (0.1) | 9 (0.03) | .06 |
| Other hemorrhage (nonintracranial) | 121 (0.6) | 98 (0.3) | <.001 |
| Hemothorax (requiring drainage) | 8 (0.04) | 8 (0.03) | .48 |
| Vascular | |||
| Major vascular complication | 43 (0.1) | 33 (0.1) | .007 |
| Arteriovenous fistula | 10 (0.05) | 16 (0.06) | .77 |
| Pseudoaneurysm | 53 (0.3) | 40 (0.1) | <.001 |
| Device | |||
| Device migration | 7 (0.03) | 10 (0.03) | .99 |
| Device systemic embolization (catheter retrieval) | 5 (0.02) | 13 (0.04) | .24 |
| Device systemic embolization (surgical retrieval) | 12 (0.1) | 9 (0.03) | .14 |
| Other | |||
| Esophageal injury | 31 (0.2) | 20 (0.1) | .005 |
| New dialysis requirement | 10 (0.05) | 9 (0.03) | .32 |
| Pulmonary embolism | 5 (0.02) | 4 (0.01) | .39 |
| Pneumonia | 26 (0.1) | 33 (0.1) | .67 |
Any in-hospital major adverse event included death, cardiac arrest, ischemic stroke, hemorrhagic stroke, undetermined stroke, transient ischemic attack, intracranial hemorrhage, systemic thromboembolism, major bleeding, major vascular complication, myocardial infarction, pericardial effusion requiring intervention, and device embolization.
As shown in the Figure, in both unadjusted and adjusted analyses, there was no difference in aborted or canceled procedures between women and men (OR, 1.01; 95% CI, 0.90-1.13; P = .87). However, after multivariable adjustment, women had a significantly higher odds than men of any adverse event (OR, 1.63; 95% CI, 1.49-1.77; P < .001), any major adverse event (OR, 2.06; 95% CI, 1.82-2.34; P < .001), length of hospitalization more than 1 day (OR, 1.46; 95% CI, 1.38-1.54; P < .001), and death (OR, 2.01; 95% CI, 1.31-3.09; P = .001).
Figure. Unadjusted and Adjusted Outcomes of In-Hospital Adverse Events for Women and Men After Receiving Watchman Implant.
Any in-hospital major adverse event included death, cardiac arrest, ischemic stroke, hemorrhagic stroke, undetermined stroke, transient ischemic attack, intracranial hemorrhage, systemic arterial embolism, major bleeding, major vascular complication, myocardial infarction, pericardial effusion requiring intervention, and device embolization. Multivariable models adjusted for age, race/ethnicity, body mass index, congestive heart failure, hypertension, diabetes, stroke, vascular disease, glomerular filtration rate, P2Y12 inhibitor prescription, oral anticoagulant prescription, coronary artery disease, left ventricular ejection fraction, and hospital region.
Postimplant Antithrombotic Therapy
The differences in postimplant antithrombotic therapy between women and men are presented in Table 3. The most common therapy for both women and men was combined anticoagulant and single antiplatelet therapy (12 658 women [62.1%] and 18 998 men [65.6%]). After multivariable adjustment, women were more likely than men to receive a direct oral anticoagulant only (OR, 1.07; 95% CI, 1.01-1.13; P = .02) and warfarin only (OR, 1.12; 95% CI, 1.05-1.19; P < .001) and were less likely to receive clinical trial–recommended combined oral anticoagulant plus single antiplatelet therapy (OR, 0.91; 95% CI, 0.87-0.95; P < .001).
Table 3. Unadjusted and Multivariable Adjusted Rates and ORs of Antithrombotic Discharge Prescription Comparing Women With Men.
| Discharge prescription | Patients, No. (%) | Unadjusted OR (95% CI) | P value | Adjusted OR (95% CI) | P value | |
|---|---|---|---|---|---|---|
| Women (n = 20 388) | Men (n = 28 969) | |||||
| Direct oral anticoagulant only | 2781 (13.6) | 3354 (11.6) | 1.21 (1.14-1.27) | <.001 | 1.07 (1.01-1.13) | .02 |
| Warfarin only | 2300 (11.3) | 2758 (9.5) | 1.21 (1.14-1.28) | <.001 | 1.12 (1.05-1.19) | <.001 |
| Single antiplatelet agent | 584 (2.8) | 793 (2.7) | 1.05 (0.94-1.17) | .42 | 1.03 (0.92-1.16) | .58 |
| Dual antiplatelet therapy | 1156 (5.7) | 1619 (5.6) | 1.02 (0.94-1.10) | .70 | 1.02 (0.93-1.10) | .73 |
| Triple therapy | 503 (2.5) | 1002 (3.5) | 0.71 (0.63-0.79) | <.001 | 0.93 (0.83-1.05) | .26 |
| Oral anticoagulant agent plus single antiplatelet agent | 12 658 (62.1) | 18 998 (65.6) | 0.86 (0.83-0.89) | <.001 | 0.91 (0.87-0.95) | <.001 |
| Other | 406 (2.0) | 445 (1.5) | 1.30 (1.37-1.49) | <.001 | 1.18 (1.02-1.36) | .02 |
Abbreviation: OR, odds ratio.
Discussion
In this analysis of the largest procedural registry of LAAO implants worldwide, including 49 357 patients between 2016 and 2019, we demonstrated several significant sex-based differences. First, women undergoing LAAO implant had different risk profiles than men, including older age and prior stroke, and were more likely to have a high fall risk as a primary indication for an implant. Second, while there was no difference in the rate of aborted or canceled procedures, women had a higher rate of any in-hospital adverse event and double the risk of major adverse events, owing to higher rates of pericardial effusion requiring drainage and major bleeding, compared with men. Third, women were more likely than men to experience prolonged hospitalization and had a 2-fold increase in the risk of in-hospital death, although the absolute risk difference was small.
The first trial evaluating the safety and efficacy of the Watchman device in 2009, PROTECT AF (Watchman Left Atrial Appendage System for Embolic Protection in Patients With AF), demonstrated that LAAO was noninferior to warfarin in preventing the long-term composite of stroke, systemic embolism, and cardiovascular death among patients with AF and a CHADS2 (congestive heart failure; hypertension; age 75 years or older; diabetes mellitus; and prior stroke, transient ischemic attack, or thromboembolism) score of 1 or more5; however, there was a high up-front procedural risk (major complications for 49 of 463 patients [10.6%]), including 22 patients (4.8%) with serious pericardial effusion, 16 (3.5%) with major bleeding, and 5 (1.1%) with procedure-related ischemic stroke. In 2014, the PREVAIL (Prospective Randomized Evaluation of the Watchman Left Atrial Appendage Closure Device in Patients with AF vs Long-term Warfarin Therapy) trial confirmed the therapeutic noninferiority of the Watchman device as an alternative to oral anticoagulation with respect to the long-term composite outcome and further reported a significantly lower procedure-related complication rate of 4.2% (11 of 265) compared with the PROTECT AF trial.6 Specifically, pericardial effusions requiring drainage decreased to 1.9% (5 of 265), and only 1 patient (0.4%) each had a procedural-related stroke and major bleeding. Compared with the pivotal trials, the rates of device-related adverse events are lower in the LAAO Registry, with 1394 of 49 357 patients (2.8%) experiencing a major adverse event, of whom 385 (0.8%) developed a pericardial effusion requiring drainage and 593 (1.2%) had major bleeding. Although these findings are overall promising, we highlight markedly disparate sex-based outcomes in a subgroup analysis that has not been extensively previously studied.
Only the PROTECT AF trial reported a sex-specific subgroup analysis showing a significantly reduced risk of the composite end point for men, but no difference was observed for women (men: relative risk, 0.32; 95% CI, 0.13-0.77; women: relative risk, 1.47; 95% CI, 0.52-4.11).5 Moreover, neither trial reported sex-specific safety events. The present study expands the current knowledge of LAAO-related adverse events by exploring sex differences in a broad range of adverse events in a large and contemporary registry. We found that women comprise a larger proportion of patients receiving LAAO implants in real-world practice than that represented in clinical trial data (41% vs approximately 30%, respectively). Furthermore, while the LAAO Registry consists of patients who are older with a higher risk of thromboembolism and bleeding than patients in the pivotal trials, these findings are amplified when comparing the sexes in the LAAO Registry.5,6,11 Compared with men in the LAAO Registry, women are older, more likely to have prior stroke, have uncontrolled hypertension, and report a high fall risk as an indication for the device.5,6 These observed differences may be due to the inclusion criteria of the trials requiring a CHA2DS2-VASc score of 1 or more among patients who are candidates for long-term anticoagulation, whereas for reimbursement purposes, CMS requires a CHA2DS2-VASc score of 3 or more among patients who are unable to receive long-term anticoagulation.13 Although these differing requirements may likely select for a higher-risk cohort in the real world, the reasons for a higher proportion of women receiving LAAO implant in the LAAO Registry, who also may have a higher overall comorbidity risk, remain unclear. It may be a result of sex bias in clinical trials, a delay in referral, or sex differences in the natural history or management of AF leading up to receipt of the LAAO implant.14 Further efforts, including increasing participation of women and more generalizable inclusion criteria in clinical trials, are needed to understand the appropriateness of patient selection for LAAO. Given the ongoing concern regarding the underrepresentation of women in cardiovascular device clinical trials, the generalizability of sex-based differences in trial data is limited, leading to the reliance on large registries to uncover these disparities.15
The most important finding of the present study is the difference in sex-based in-hospital outcomes among patients undergoing LAAO. Compared with men, women were at a 2-fold higher risk of any adverse event and of major adverse events, largely owing to higher rates of pericardial effusion requiring drainage and major bleeding. Women also were more likely to experience prolonged hospital stay and death, albeit the absolute risk difference of death was minimal. These findings are consistent with those of prior studies demonstrating an increased risk of adverse events for women undergoing invasive procedures.9,16,17 For 21 091 patients undergoing AF ablation, of whom 29% were women, there was a significantly higher rate of cardiac tamponade (3.8% vs 2.9%; P < .001) and hemorrhage (2.7% vs 2.0%; P < .001) for women compared with men.10 In a study of 9281 patients using the National Readmission Database, Osman et al18 demonstrated a higher risk of major in-hospital events (as defined by death, major vascular complication, or pericardial complication) among women than men (11.4% vs 6.7%; P < .01) after receipt of Watchman implants from 2015 to 2017. We expand on the prior work with a markedly larger cohort and prospectively collected data that are subject to a rigorous event adjudication process and highlight a markedly higher risk of adverse events for women vs men undergoing LAAO compared with prior invasive cardiac procedures.
When complications were considered individually, pericardial effusion requiring drainage and major bleeding were the main factors associated with the major adverse events observed in women, a finding previously observed in patients undergoing AF ablation and transcatheter aortic valve implant.10,18,19,20 Although our study was not equipped to identify the underlying mechanisms explaining the increased risk of in-hospital adverse events for women undergoing LAAO, several plausible explanations exist. Anatomical differences between women and men, such as smaller vessel diameter, thinner myocardial wall, or a more friable LAA in women, may predispose women to increased risk of complications.21 In the present study, women had a slightly smaller LAA maximal width (21.3 mm vs 21.7 mm; P < .001); however, other characteristics, such as left atrial size, LAA morphology, and number of device deployment attempts, may have important risk implications with regard to pericardial effusion that are not captured in the registry. Moreover, clinician inexperience may be associated with the risk of adverse events, particularly if those clinicians are more likely to perform LAAO for women who may be inherently at increased risk owing to anatomical issues or underlying comorbid conditions. Next, although not directly captured in the registry, women may be frailer than men, as supported by their older age and higher likelihood of prior stroke, anemia, and identification as a high fall risk, a subjective definition left to clinician discretion. Frailty, a measurement that covers several domains, such as nutritional, functional, and cognitive status, has become an important tool in estimating mortality and morbidity after invasive procedures; however, no data currently exist for those undergoing LAAO.22 Last, while we found that women were less likely to be prescribed clinical trial–recommended oral anticoagulant plus single antiplatelet therapy, whether less-aggressive antithrombotic therapy in women after LAAO affects outcomes will require clinical trial data comparing variations of antithrombotic therapy and long-term follow-up.
Although we cannot fully elucidate the causes of the increased risk of adverse events experienced by women, several strategies may mitigate procedural risk, including ultrasonography-guided venous access, preprocedural imaging to evaluate cardiac function, LAA size and morphology to guide equipment and device selection, developing proficiency with LAAO devices, and continued development of safer devices.23,24 We must continue to advocate for increased participation of women in clinical trials to better inform clinical decision-making and adequately delineate sex-based safety and efficacy outcomes.
Limitations
Our study must be interpreted in the context of several limitations. First, the findings are based on observational registry data, and thus causal inferences cannot be made. Although registry data are more generalizable regarding adverse events compared with trials owing to larger sample size, multicenter data, and less restriction on implant indications, reflecting real-world practice, the results of this study should not result in differing sex-based thresholds for LAAO implant. Second, although we adjusted for potential confounders, there may be unmeasured confounders that influence the risk associations not captured in the data collection form. Third, the present analysis was limited to in-hospital outcomes. Long-term results are needed provide further insights into sex-based differences in outcomes.
Conclusions
In the largest, contemporary registry of patients undergoing LAAO implant, to our knowledge, women were more likely than men to experience in-hospital adverse outcomes after multivariable adjustment, including a 2-fold higher risk of major adverse events. Further research is needed to identify the reasons for sex-based differences in outcomes and the strategies to reduce the risk of adverse events among women with AF receiving the Watchman implant.
References
- 1.Chugh SS, Havmoeller R, Narayanan K, et al. Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study. Circulation. 2014;129(8):837-847. doi: 10.1161/CIRCULATIONAHA.113.005119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ball J, Carrington MJ, McMurray JJ, Stewart S. Atrial fibrillation: profile and burden of an evolving epidemic in the 21st century. Int J Cardiol. 2013;167(5):1807-1824. doi: 10.1016/j.ijcard.2012.12.093 [DOI] [PubMed] [Google Scholar]
- 3.January CT, Wann LS, Calkins H, et al. ; Writing Group Members . 2019 AHA/ACC/HRS Focused update of the 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. Heart Rhythm. 2019;16(8):e66-e93. doi: 10.1016/j.hrthm.2019.01.024 [DOI] [PubMed] [Google Scholar]
- 4.Stoddard MF, Dawkins PR, Prince CR, Ammash NM. Left atrial appendage thrombus is not uncommon in patients with acute atrial fibrillation and a recent embolic event: a transesophageal echocardiographic study. J Am Coll Cardiol. 1995;25(2):452-459. doi: 10.1016/0735-1097(94)00396-8 [DOI] [PubMed] [Google Scholar]
- 5.Holmes DR, Reddy VY, Turi ZG, et al. ; PROTECT AF Investigators . Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial. Lancet. 2009;374(9689):534-542. doi: 10.1016/S0140-6736(09)61343-X [DOI] [PubMed] [Google Scholar]
- 6.Holmes DR Jr, Kar S, Price MJ, et al. Prospective randomized evaluation of the Watchman left atrial appendage closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol. 2014;64(1):1-12. doi: 10.1016/j.jacc.2014.04.029 [DOI] [PubMed] [Google Scholar]
- 7.Ko D, Rahman F, Schnabel RB, Yin X, Benjamin EJ, Christophersen IE. Atrial fibrillation in women: epidemiology, pathophysiology, presentation, and prognosis. Nat Rev Cardiol. 2016;13(6):321-332. doi: 10.1038/nrcardio.2016.45 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Emdin CA, Wong CX, Hsiao AJ, et al. Atrial fibrillation as risk factor for cardiovascular disease and death in women compared with men: systematic review and meta-analysis of cohort studies. BMJ. 2016;532:h7013. doi: 10.1136/bmj.h7013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Peterson PN, Daugherty SL, Wang Y, et al. ; National Cardiovascular Data Registry . Gender differences in procedure-related adverse events in patients receiving implantable cardioverter-defibrillator therapy. Circulation. 2009;119(8):1078-1084. doi: 10.1161/CIRCULATIONAHA.108.793463 [DOI] [PubMed] [Google Scholar]
- 10.Kaiser DW, Fan J, Schmitt S, et al. Gender Differences in Clinical Outcomes after Catheter Ablation of Atrial Fibrillation. JACC Clin Electrophysiol. 2016;2(6):703-710. doi: 10.1016/j.jacep.2016.04.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Freeman JV, Varosy P, Price MJ, et al. The NCDR Left Atrial Appendage Occlusion Registry. J Am Coll Cardiol. 2020;75(13):1503-1518. doi: 10.1016/j.jacc.2019.12.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.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 [DOI] [PubMed] [Google Scholar]
- 13.Centers for Medicare & Medicaid Services. Decision memo for percutaneous left atrial appendage (LAA) closure therapy (CAG-00445N). Accessed November 10, 2020. https://www.cms.gov/medicare-coverage-database/details/nca-decision-memo.aspx?NCAId=281
- 14.Volgman AS, Benjamin EJ, Curtis AB, et al. ; American College of Cardiology Committee on Cardiovascular Disease in Women . Women and atrial fibrillation. J Cardiovasc Electrophysiol. 2020;(Dec). doi: 10.1111/jce.14838 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ghare MI, Chandrasekhar J, Mehran R, Ng V, Grines C, Lansky A. Sex Disparities in Cardiovascular Device Evaluations: Strategies for Recruitment and Retention of Female Patients in Clinical Device Trials. JACC Cardiovasc Interv. 2019;12(3):301-308. doi: 10.1016/j.jcin.2018.10.048 [DOI] [PubMed] [Google Scholar]
- 16.Vlastra W, Chandrasekhar J, García Del Blanco B, et al. Sex Differences in Transfemoral Transcatheter Aortic Valve Replacement. J Am Coll Cardiol. 2019;74(22):2758-2767. doi: 10.1016/j.jacc.2019.09.015 [DOI] [PubMed] [Google Scholar]
- 17.Elayi CS, Darrat Y, Suffredini JM, et al. Sex differences in complications of catheter ablation for atrial fibrillation: results on 85,977 patients. J Interv Card Electrophysiol. 2018;53(3):333-339. doi: 10.1007/s10840-018-0416-1 [DOI] [PubMed] [Google Scholar]
- 18.Osman M, Patel B, Munir MB, et al. Sex-stratified analysis of the safety of percutaneous left atrial appendage occlusion. Catheter Cardiovasc Interv. 2021;97(5):885-892. doi: 10.1002/ccd.29282 [DOI] [PubMed] [Google Scholar]
- 19.Zhang XD, Tan HW, Gu J, et al. Efficacy and safety of catheter ablation for long-standing persistent atrial fibrillation in women. Pacing Clin Electrophysiol. 2013;36(10):1236-1244. doi: 10.1111/pace.12212 [DOI] [PubMed] [Google Scholar]
- 20.Chandrasekhar J, Dangas G, Yu J, et al. ; STS/ACC TVT Registry . Sex-based differences in outcomes with transcatheter aortic valve therapy: TVT registry from 2011 to 2014. J Am Coll Cardiol. 2016;68(25):2733-2744. doi: 10.1016/j.jacc.2016.10.041 [DOI] [PubMed] [Google Scholar]
- 21.Kamiński R, Kosiński A, Brala M, et al. Variability of the left atrial appendage in human hearts. PLoS One. 2015;10(11):e0141901. doi: 10.1371/journal.pone.0141901 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Afilalo J, Lauck S, Kim DH, et al. Frailty in older adults undergoing aortic valve replacement: the FRAILTY-AVR Study. J Am Coll Cardiol. 2017;70(6):689-700. doi: 10.1016/j.jacc.2017.06.024 [DOI] [PubMed] [Google Scholar]
- 23.Kupó P, Pap R, Sághy L, et al. Ultrasound guidance for femoral venous access in electrophysiology procedures-systematic review and meta-analysis. J Interv Card Electrophysiol. 2020;59(2):407-414. doi: 10.1007/s10840-019-00683-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kar S, Doshi SK, Sadhu A, et al. ; PINNACLE FLX Investigators . Primary outcome evaluation of a next-generation left atrial appendage closure device: results from the PINNACLE FLX trial. Circulation. 2021;143(18):1754-1762. doi: 10.1161/CIRCULATIONAHA.120.050117 [DOI] [PubMed] [Google Scholar]