Key Points
Question
Are race and ethnicity associated with oral anticoagulant use and long-term outcomes in patients with atrial fibrillation?
Findings
In this cohort study including 69 553 patients hospitalized with atrial fibrillation in the Get With The Guidelines–Atrial Fibrillation registry, after adjustment for socioeconomic status and community-level social determinants of health, Black patients with atrial fibrillation were less likely than White patients to receive anticoagulant therapy at discharge and experienced significantly higher rates of adverse outcomes at 1 year, including stroke and mortality.
Meaning
In the era of increased availability of effective, guideline-based therapy, racial inequities persisted in anticoagulant use and outcomes, and interventions to reduce these inequities is key to improving quality of atrial fibrillation care.
This cohort study compares oral anticoagulation use at discharge and atrial fibrillation–related outcomes by race and ethnicity in the Get With The Guidelines–Atrial Fibrillation registry.
Abstract
Importance
Oral anticoagulation (OAC) is underprescribed in underrepresented racial and ethnic group individuals with atrial fibrillation (AF). Little is known of how differential OAC prescribing relates to inequities in AF outcomes.
Objective
To compare OAC use at discharge and AF-related outcomes by race and ethnicity in the Get With The Guidelines–Atrial Fibrillation (GWTG-AFIB) registry.
Design, Setting, and Participants
This retrospective cohort analysis used data from the GWTG-AFIB registry, a national quality improvement initiative for hospitalized patients with AF. All registry patients hospitalized with AF from 2014 to 2020 were included in the study. Data were analyzed from November 2021 to July 2022.
Exposures
Self-reported race and ethnicity assessed in GWTG-AFIB registry.
Main Outcomes and Measures
The primary outcome was prescription of direct-acting OAC (DOAC) or warfarin at discharge. Secondary outcomes included cumulative 1-year incidence of ischemic stroke, major bleeding, and mortality postdischarge. Outcomes adjusted for patient demographic, clinical, and socioeconomic characteristics as well as hospital factors.
Results
Among 69 553 patients hospitalized with AF from 159 sites between 2014 and 2020, 863 (1.2%) were Asian, 5062 (7.3%) were Black, 4058 (5.8%) were Hispanic, and 59 570 (85.6%) were White. Overall, 34 113 (49.1%) were women; the median (IQR) age was 72 (63-80) years, and the median (IQR) CHA2DS2-VASc score (calculated as congestive heart failure, hypertension, age 75 years and older, diabetes, stroke or transient ischemic attack, vascular disease, age 65 to 74 years, and sex category) was 4 (2-5). At discharge, 56 385 patients (81.1%) were prescribed OAC therapy, including 41 760 (74.1%) receiving DOAC. OAC prescription at discharge was lowest in Hispanic patients (3010 [74.2%]), followed by Black patients (3935 [77.7%]) Asian patients (691 [80.1%]), and White patients (48 749 [81.8%]). Black patients were less likely than White patients to be discharged while taking any anticoagulant (adjusted odds ratio, 0.75; 95% CI, 0.68-0.84) and DOACs (adjusted odds ratio, 0.73; 95% CI, 0.65-0.82). In 16 307 individuals with 1-year follow up data, bleeding risks (adjusted hazard ratio [aHR], 2.08; 95% CI, 1.53-2.83), stroke risks (aHR, 2.07; 95% CI, 1.34-3.20), and mortality risks (aHR, 1.22; 95% CI, 1.02-1.47) were higher in Black patients than White patients. Hispanic patients had higher stroke risk (aHR, 2.02; 95% CI, 1.38-2.95) than White patients.
Conclusions and Relevance
In a national registry of hospitalized patients with AF, compared with White patients, Black patients were less likely to be discharged while taking anticoagulant therapy and DOACs in particular. Black and Hispanic patients had higher risk of stroke compared with White patients; Black patients had a higher risk of bleeding and mortality. There is an urgent need for interventions to achieve pharmacoequity in guideline-directed AF management to improve overall outcomes.
Introduction
Atrial fibrillation (AF) is the most common arrhythmia, affecting up to 60 million adults worldwide.1,2,3,4 AF increases the risk of all-cause mortality and is associated with high rates of cardiovascular morbidity, including ischemic stroke.5,6,7,8 Oral anticoagulation (OAC) for AF, particularly direct-acting oral anticoagulants (DOACs), significantly reduces the risk of stroke and is the standard of care in patients with moderate to severe risk of ischemic stroke.9,10,11,12
Prior research has demonstrated that underrepresented racial and ethnic group individuals with AF, including American Indian or Alaska Native, Asian, Black, and Hispanic individuals, are less likely than White individuals to be treated with any form of OAC, particularly DOACs.13,14,15,16,17 The gravity of these inequities is amplified by the fact that individuals from underrepresented racial and ethnic groups with AF have higher rates of stroke and mortality than White individuals.18,19,20 Nevertheless, there remains an opportunity to examine whether racial and ethnic inequities in AF management extend to contemporary OAC prescribing.21,22,23,24
The American Heart Association (AHA) Get With The Guidelines (GWTG) program has demonstrated quality improvement in the treatment of several cardiovascular diseases, including heart failure, myocardial infarction, and stroke. The GWTG-Atrial Fibrillation (GWTG-AFIB) program was launched in 2013 to improve quality of AF care,25 yet little is known of how equity in OAC prescribing and long-term AF outcomes have changed since program implementation. Our primary aim was to compare OAC use at hospital discharge by race and ethnicity for patients in the GWTG-AFIB registry. In secondary analyses, we examined patterns of OAC prescribing over time by race and ethnicity and postdischarge outcomes, including ischemic stroke, major bleeding, and mortality.
Methods
Data Sources
The data used in these analyses were collected from the GWTG-AFIB program, a national, voluntary quality improvement initiative started in January 2013 by the AHA in partnership with the Heart Rhythm Society.15,25 The goal of GWTG-AFIB is to improve cardiovascular health and outcomes in patients with AF including in the use of class I recommended stroke reduction therapies.11,26 The GWTG-AFIB program, its component data elements, and the AF performance measures have been described previously and in the eMethods in the Supplement.25 This study was approved by the Institutional Review Board of Duke University. Informed consent was waived as the analysis used deidentified data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
To assess our secondary outcomes, we used data from the US Centers of Medicare & Medicaid Services (CMS). The linking procedure between the GWTG registry and CMS has been previously described.27,28 The CMS data include inpatient claims along with the corresponding denominator file from 2014 to 2018. The inpatient files contain institutional claims with costs covered by Medicare Part A and additional encrypted beneficiary identifiers, including demographic information, dates of service, diagnosis-related groups, and International Classification of Diseases, Ninth Revision (ICD-9) and ICD-10 codes. The denominator files contain encrypted beneficiary identifiers, demographic characteristics, date of death, and information pertaining to program enrollment.
Study Cohort
Between January 1, 2014, and June 30, 2020, there were 80 989 patients hospitalized with AF as a primary or secondary diagnosis who were discharged from 161 sites participating in the GWTG-AFIB registry. We excluded those with mechanical prosthetic heart valve (n = 1235), those receiving comfort care only (n = 2856), those whose discharge disposition was missing or not documented, those who transferred to another facility or left against medical advice (n = 1849), and those who died during hospitalization (n = 257). We also excluded those missing information on oral anticoagulation use at admission or discharge (n = 2525; 1 site); those missing information on age, sex, or medical history (n = 120); and those missing race and ethnicity data or who identified as other race (n = 2594; 1 site) (Figure 1).
Figure 1. Flow of Patients in the Get With The Guidelines–Atrial Fibrillation (GWTG-AFIB) Study.
Between January 1, 2014, and June 30, 2020, there were 80 989 patients hospitalized with atrial fibrillation who were discharged from 161 sites participating in the GWTG-AFIB registry. After clinical and discharge disposition exclusions, the study cohort for anticoagulation outcomes was 69 553 from 159 sites. To assess postdischarge outcomes, we included only those with a hospital discharge between January 1, 2014, and June 30, 2018, and those with data linked to the US Centers for Medicare & Medicaid Services. This resulted in a cohort of 16 307 patients to examine postdischarge outcomes.
To assess postdischarge outcomes, we further subset the in-hospital analysis cohort to include only those with a hospital discharge between January 1, 2014, and June 30, 2018 (the dates for which we have discharge data with at least 6 months of follow-up). We excluded individuals younger than 65 years and those without data linked to CMS. We further excluded individuals with in-hospital death, hospice, transfers, left against medical advice, or missing discharge status (Figure 1).
Study Outcomes
The primary outcome was OAC prescription at discharge by race and ethnicity. Discharge medication was ascertained using medical record review abstraction and was considered present if there was documentation of warfarin, apixaban, dabigatran, edoxaban, or rivaroxaban medication.25 Secondary outcomes were examined in the CMS linkage cohort and included incidence of ischemic stroke, major bleeding, and all-cause mortality within 1 year of discharge by race and ethnicity and by anticoagulation status at discharge (ie, no OAC, any OAC, warfarin vs DOAC). Using previously described methods, ischemic stroke and major bleeding were defined by established ICD-9 and ICD-10 diagnosis codes from CMS inpatient claims data (eTable 1 in the Supplement).29,30 It was required that these codes were in the primary (ischemic stroke) or primary or secondary position (major bleeding).31 Major bleeding included intracerebral bleed, gastrointestinal bleed, pericardial bleed, and procedure-related bleeding. Mortality was determined using the date of death recorded in CMS data. In sensitivity analyses, we examined primary and secondary outcomes by CHA2DS2-VASc stroke risk score (calculated as congestive heart failure, hypertension, age 75 years and older, diabetes, stroke or transient ischemic attack, vascular disease, age 65 to 74 years, and sex category) of less than 2 or 2 and greater based on guidelines for anticoagulation in AF.11,26
Independent Variables and Baseline Characteristics
We examined several patient sociodemographic and clinical as well as hospital characteristics considered determinants of the association between our primary independent variables (ie, race and ethnicity) and our study outcomes. Race and ethnicity were identified by self-report as a required multiple-choice data element in the GWTG-AFIB registry.25,32,33 This variable was defined in mutually exclusive categories as non-Hispanic Asian, non-Hispanic Black, Hispanic (any race), and non-Hispanic White (the referent group).
We identified patient age and sex. Socioeconomic status factors included insurance status, categorized as private/health maintenance organization, Medicare, Medicare-private health maintenance organization, Medicaid, or uninsured. Patient 5-digit zip code and census block data captured in the GWTG-AFIB registry was used to link individual patients to the 2019 US Census Bureau and the American Community Survey, through which we examined median household income, median home value, high school and college graduate rate, and unemployment rate.34,35,36
We assessed patient clinical variables using abstraction from medical record review. These included type of AF, CHA2DS2-VASc stroke risk score, Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) bleeding risk score37 and previous medical history, including anemia, chronic obstructive pulmonary disease, coronary artery disease, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke or transient ischemic attack, previous hemorrhage, diabetes, hemodialysis, heart failure, left ventricular ejection fraction less than 40%, hypertension, chronic kidney disease stage of III or greater, dialysis, obstructive sleep apnea, thyroid disease, cardiomyopathy, smoking, alcohol use disorder, illicit drug use, and mean body mass index. We also assessed the presence of antiplatelet and antiarrhythmic agents at admission as well as year of AF admission.
Hospital characteristics included whether the patient was hospitalized at an academic or teaching hospital, hospital with access to clinical cardiac electrophysiology specialist care, number of beds, rural location, and geographic region. We also examined the rate of underrepresented racial and ethnic group patients hospitalized with AF at each GWTG-AFIB site to assess for hospital-level variation in OAC use by race and ethnicity.
Statistical Analysis
Baseline patient characteristics were described overall and by race and ethnicity using proportions for categorical variables and medians with IQRs for continuous variables. Differences in baseline characteristics were compared with Pearson χ2 test, if applicable, for categorical variables and Wilcoxon rank-sum test for continuous variables. We accounted for site clustering in all of the models, which were fit at the patient level.
To assess the associations between race, ethnicity, temporal trends, and OAC use at discharge, we used logistic regression with generalized estimating equations, which account for within-hospital correlation. Models adjusted for the patient-level sociodemographic characteristics, clinical factors (including admission year), and hospital factors. We also tested an AF type × race and ethnicity group interaction term into these models to assess differences in OAC use by race, ethnicity, and type of AF.
Postdischarge clinical outcomes were assessed in the Medicare linkage cohort. Event rates within 1 year of discharge were calculated beginning on date of index discharge, using the Kaplan-Meier method or the cumulative incidence function for nonfatal outcomes.38 Differences in cumulative incidence were compared using log-rank test for mortality and Gray test for nonfatal outcomes. Outcomes were compared using Cox proportional hazards models with robust variance estimation to account for within-hospital clustering and adjustment for the same set of variables for the discharge OAC outcomes as well as adjustment for interaction between type of OAC at discharge and race and ethnicity groups. The cause-specific model was used to account for the competing risk of mortality for nonfatal outcomes.
Multiple imputation was used to handle missing data in the models; 5 imputed data sets were created using the fully conditional specification method.15 Hospital characteristics and type of AF were not imputed. Patients at sites with missing hospital characteristics were excluded from models. Most variables had missing rates less than 10%, except estimated glomerular filtration rate less than 60 mL/min/1.73 m2 (21.6%). All tests were 2-tailed, and statistical significance was set at P < .05. All analyses were performed with SAS version 9.4 (SAS Institute).
Results
Baseline Characteristics
The final study cohort included 69 553 patients hospitalized with AF from 159 sites, of whom 863 (1.2%) were Asian, 5062 (7.3%) were Black, 4058 (5.8%) were Hispanic, and 59 570 (85.6%) were White (Table). Overall, 34 113 (49.1%) were women, and the median (IQR) age was 72 (63-80) years. The median (IQR) CHA2DS2-VASc score was 4.0 (2.0-5.0), and 61 523 (88.5%) had a CHA2DS2-VASc score of 2 or greater. Overall, 29 845 patients (42.9%) had paroxysmal AF, 17 521 (25.2%) had persistent or permanent AF, and 14 545 (20.9%) were experiencing their first detected AF diagnosis.
Table. Baseline Patient Characteristics in Patients Hospitalized With Atrial Fibrillation (AF) by Race and Ethnicitya.
| Variable | No. (%)b | ||||
|---|---|---|---|---|---|
| Overall (N = 69 553) | Asian (n = 863) | Black (n = 5062) | Hispanic (n = 4058) | White (n = 59 570) | |
| Sociodemographic characteristics | |||||
| Age, median (IQR), y | 72.0 (63.0-80.0) | 73.0 (63.0-81.0) | 64.0 (54.0-74.0) | 69.0 (59.0-79.0) | 72.0 (64.0-81.0) |
| Sex | |||||
| Male | 35 440 (51.0) | 419 (48.6) | 2546 (50.3) | 1999 (49.3) | 30 476 (51.2) |
| Female | 34 113 (49.0) | 444 (51.5) | 2516 (49.7) | 2059 (50.7) | 29 094 (48.8) |
| Insurance | |||||
| Private | 26 768 (38.5) | 343 (39.8) | 1547 (30.6) | 1407 (34.7) | 23 471 (39.4) |
| Medicaid | 6668 (9.6) | 195 (22.6) | 1262 (24.9) | 1068 (26.3) | 4143 (7.0) |
| Medicare | 16 154 (23.2) | 120 (13.9) | 790 (15.6) | 544 (13.4) | 14 700 (24.7) |
| Medicare-private | 14 767 (21.2) | 90 (10.4) | 872 (17.2) | 7251 (7.9) | 13 080 (22.0) |
| No insurance | 1448 (2.1) | 19 (2.2) | 231 (4.6) | 229 (5.6) | 969 (1.6) |
| Household income, median (IQR), $ | 60 085 (47 623-76 457) | 84 254 (63 225-116 783) | 43 706 (34 274-57 330) | 46 865 (37 183-65 441) | 61 447 (49 865-77 702) |
| Home value of owner-occupied units, median (IQR), $ | 172 800 (125 300-265 000) | 467 200 (234 700-790 700) | 130 500 (90 600-189 500) | 172 800 (95 700-322 500) | 175 200 (130 100-264 400) |
| High school graduation rate, median (IQR) | 91.3 (86.0-94.6) | 91.2 (85.8-95.4) | 85.5 (80.1-90.9) | 79.6 (70.3-89.4) | 91.9 (87.3-95.0) |
| College graduation rate, median (IQR) | 28.1 (19.2-41.5) | 40.4 (28.1-55.4) | 21.9 (14.6-31.2) | 18.1 (11.8-30.5) | 29.3 (20.0-42.5) |
| Unemployment rate, median (IQR) | 4.5 (3.4-6.2) | 4.2 (3.4-5.5) | 6.9 (5.0-10.3) | 5.9 (4.5-8.4) | 4.3 (3.2-5.9) |
| Clinical characteristics | |||||
| Type of AF | |||||
| Paroxysmal | 29 843 (42.9) | 332 (38.4) | 2196 (43.4) | 1612 (39.7) | 25 703 (43.2) |
| Persistent | 12 526 (18.0) | 169 (19.6) | 698 (13.8) | 597 (14.7) | 11 062 (18.6) |
| Permanent | 4995 (7.2) | 80 (9.3) | 276 (5.4) | 223 (5.5) | 4416 (7.4) |
| First detected | 14 545 (20.9) | 195 (22.6) | 1240 (24.5) | 938 (23.1) | 12 172 (20.4) |
| Unable to determine | 7644 (11.0) | 87 (10.1) | 652 (12.9) | 688 (17.0) | 6217 (10.4) |
| CHA2DS2-VASc score, median (IQR)c | 4.0 (2.0-5.0) | 4.0 (3.0-5.0) | 4.0 (2.0-5.0) | 4.0 (2.0-5.0) | 4.0 (2.0-5.0) |
| ORBIT-AF score, median (IQR)d | 2.0 (1.0-3.0) | 2.0 (1.0-4.0) | 2.0 (1.0-3.0) | 2.0 (1.0-3.0) | 2.0 (1.0-3.0) |
| Anemia | 7750 (11.1) | 36 (15.8) | 806 (15.9) | 439 (10.8) | 6369 (10.7) |
| COPD | 12 122 (17.4) | 78 (9.0) | 853 (16.9) | 499 (12.3) | 10 692 (18.0) |
| Coronary artery disease | 19 743 (28.4) | 245 (28.4) | 1173 (23.2) | 1045 (25.8) | 17 289 (29.0) |
| Prior myocardial infarction | 6893 (9.9) | 75 (8.7) | 510 (10.1) | 284 (7.0) | 6024 (10.1) |
| Prior PCI | 8038 (11.6) | 110 (12.8) | 447 (8.8) | 420 (10.4) | 7061 (11.9) |
| Prior stroke or TIA | 9519 (13.7) | 133 (15.4) | 851 (16.8) | 521 (12.8) | 8014 (13.5) |
| Prior hemorrhage | 2927 (4.2) | 47 (5.5) | 226 (4.5) | 117 (2.9) | 2537 (4.3) |
| Prior AF procedure | |||||
| Cardioversion | 14 041 (20.2) | 159 (18.4) | 697 (13.8) | 419 (10.3) | 12 766 (21.4) |
| Ablation | 7312 (10.5) | 71 (8.2) | 440 (8.7) | 228 (5.6) | 6573 (11.0) |
| Surgical maze | 533 (0.8) | 6 (0.7) | 28 (0.6) | 21 (0.5) | 478 (0.8) |
| Peripheral vascular disease | 4251 (6.1) | 22 (2.6) | 276 (5.5) | 192 (4.7) | 3761 (6.3) |
| Diabetes | 19 760 (28.3) | 296 (34.3) | 1911 (37.8) | 1653 (40.7) | 15 810 (26.5) |
| Heart failure | 19 539 (28.1) | 258 (29.9) | 1935 (38.2) | 1083 (26.7) | 16 263 (27.3) |
| Left ventricular ejection fraction <40% | 9567 (13.8) | 124 (14.4) | 1088 (21.5) | 627 (15.5) | 7728 (13.0) |
| Hypertension | 53 552 (77.0) | 674 (78.1) | 4303 (85.0) | 3269 (80.6) | 45 306 (76.1) |
| Chronic kidney disease stage ≥III | 25 835 (37.1) | 311 (36.0) | 1825 (36.1) | 1516 (37.4) | 22 183 (37.2) |
| Dialysis dependent | 1044 (1.5) | 27 (3.1) | 261 (5.2) | 180 (4.4) | 576 (1.0) |
| Obstructive sleep apnea | 12 297 (17.7) | 109 (12.6) | 925 (18.3) | 471 (11.6) | 10 792 (18.1) |
| Thyroid disease | 12 921 (18.6) | 159 (18.4) | 538 (10.6) | 673 (16.6) | 11 551 (19.4) |
| Cardiomyopathy | 8359 (12.0) | 131 (15.2) | 886 (17.5) | 442 (10.9) | 6900 (11.6) |
| Antiplatelet agent on admission | 5280 (7.6) | 78 (9.0) | 346 (6.8) | 311 (7.7) | 4545 (7.6) |
| Antiarrhythmic on admission | 11 663 (16.8) | 140 (16.2) | 691 (13.7) | 496 (12.2) | 10 336 (17.4) |
| Anticoagulation on admission | 35 207 (50.6) | 428 (49.6) | 2211 (43.7) | 1562 (38.5) | 31 006 (52.1) |
| DOAC | 23 602 (33.9) | 292 (33.8) | 1415 (28.0) | 1033 (25.5) | 20 862 (35.0) |
| Warfarin | 11 605 (16.7) | 136 (15.8) | 796 (15.7) | 529 (13.0) | 10 144 (17.0) |
| Smoker | 7001 (10.1) | 38 (4.4) | 903 (17.8) | 385 (9.5) | 5675 (9.5) |
| Alcohol dependence | 2907 (4.2) | 15 (1.7) | 310 (6.1) | 204 (5.0) | 2378 (4.0) |
| Illicit drug use | 1203 (1.7) | 2 (0.2) | 383 (7.6) | 116 (2.9) | 702 (1.2) |
| Body mass index, median (IQR)e | 29.3 (25.2-34.7) | 24.8 (22.3-28.0) | 31.2 (26.0-32.7) | 30.0 (26.0-35.1) | 29.2 (25.1-30.5) |
| Admission year | |||||
| 2014 | 3059 (4.4) | 42 (4.9) | 269 (5.3) | 348 (8.6) | 2400 (4.0) |
| 2015 | 8531 (12.3) | 72 (8.3) | 568 (11.2) | 456 (11.2) | 7435 (12.5) |
| 2016 | 11 327 (16.3) | 147 (17.0) | 814 (16.1) | 743 (18.3) | 9623 (16.2) |
| 2017 | 13 750 (19.8) | 153 (17.7) | 933 (18.4) | 778 (19.2) | 11 886 (20.0) |
| 2018 | 14 124 (20.3) | 210 (24.3) | 967 (19.1) | 861 (21.2) | 12 086 (20.3) |
| 2019 | 14 946 (21.5) | 191 (22.1) | 1190 (23.5) | 754 (18.6) | 12 811 (21.5) |
| 2020 | 3816 (5.5) | 48 (5.6) | 321 (6.3) | 118 (2.9) | 3329 (5.6) |
| Hospital characteristics | |||||
| Teaching hospital | 55 742 (80.1) | 698 (80.9) | 4549 (89.9) | 3617 (89.1) | 46 878 (78.7) |
| Cardiac EP services | 9741 (14.0) | 70 (8.1) | 435 (8.6) | 148 (3.7) | 9088 (15.3) |
| Rural location | 4581 (6.6) | 44 (5.1) | 42 (0.8) | 16 (0.4) | 4479 (7.5) |
| Hospital size ≥500 beds | 25 735 (39.0) | 231 (27.0) | 2889 (59.2) | 2633 (66.5) | 19 982 (35.4) |
| Region | |||||
| Northeast | 25 032 (36.0) | 2032 (3.5) | 1276 (25.2) | 911 (22.5) | 22 642 (38.0) |
| Midwest | 13 047 (18.8) | 116 (13.4) | 1152 (22.8) | 351 (8.7) | 11 428 (19.2) |
| South | 22 722 (32.7) | 101 (11.7) | 2313 (45.7) | 1934 (47.7) | 18 374 (30.8) |
| West | 7979 (11.5) | 443 (51.3) | 277 (5.5) | 858 (21.1) | 6401 (10.8) |
Abbreviations: COPD, chronic obstructive pulmonary disease; DOAC, direct-acting oral anticoagulant; EP, electrophysiology; ORBIT, Outcomes Registry for Better Informed Treatment of Atrial Fibrillation; PCI, percutaneous coronary intervention; TIA, transient ischemic attack.
Race and ethnicity were identified by self-report as a required multiple-choice data element in the Get With The Guidelines–Atrial Fibrillation registry.
P values are based on χ2 rank-based group means score statistics for all continuous and ordinal row variables. All P values were significant at P < .001 except for chronic kidney disease (P = .26). All tests treat the column variable as nominal.
Calculated as congestive heart failure, hypertension, age 75 years and older, diabetes, stroke or transient ischemic attack, vascular disease, age 65 to 74 years, and sex category.
ORBIT indicates a score composed of points for older age (≥74 years), bleeding history, insufficient kidney function, and treatment with antiplatelet.
Calculated as weight in kilograms divided by height in meters squared.
Anticoagulant Use at Discharge
On discharge, 56 385 patients (81.1%) were prescribed OAC therapy, including 41 760 (74.1%) receiving DOAC and 14 625 (25.9%) receiving warfarin. OAC prescription at discharge was lowest in Hispanic patients (3010 [74.2%]), followed by Black patients (3935 [77.7%]), Asian patients (691 [80.1%]), and White patients (48 749 [81.8%]). In adjusted models, Black patients were significantly less likely than White patients to be discharged taking any anticoagulant (adjusted odds ratio, 0.75; 95% CI, 0.68-0.84) and DOACs in particular (adjusted odds ratio, 0.73; 95% CI, 0.65-0.82); this difference was not observed in other racial and ethnic groups (Figure 2A). Figure 2B and eTable 2 in the Supplement demonstrate that racial and ethnic differences in OAC receipt at discharge were also present in the subset of patients with a CHA2DS2-VASc score of 2 or greater. OAC use also differed across key medical comorbidities, including congestive heart failure, chronic and end-stage kidney disease, and types of AF (ie, paroxysmal, persistent, and permanent) (eTable 3 in the Supplement). Notably, interaction terms for race, ethnicity, and AF type were nonsignificant, indicating similar differences in any OAC and DOAC initiation across AF types for all racial and ethnic groups. Racial and ethnic patient representation at the hospital level was not strongly associated with anticoagulant use overall nor among patients from underrepresented racial or ethnic groups (eFigure 1 in the Supplement).
Figure 2. Adjusted Odds Ratios (aORs) for Anticoagulant Therapy at Discharge by Race and Ethnicity for Patients With Atrial Fibrillation.
Using logistic regression modeling adjusted for patient sociodemographic and clinical factors as well as hospital factors, Black patients had significantly lower receipt of any anticoagulant on discharge, including warfarin and direct-acting oral anticoagulation (DOAC) therapy compared with White patients in both the overall cohort (A) and when examining only patients with a CHA2DS2-VASc stroke risk score (calculated as congestive heart failure, hypertension, age 75 years and older, diabetes, stroke or transient ischemic attack, vascular disease, age 65 to 74 years, and sex category) of 2 or greater (B).
Anticoagulant Use Over Time
Overall, OAC use at discharge increased from 67.2% in 2014 to 87.4% in 2020 (P < .001) (eFigure 2 in the Supplement). OAC use at discharge increased from 2014 to 2020 by an absolute 23.5% in Asian patients, 20.7% in Black patients, 18.2% in Hispanic patients, and 19.7% in White patients (P = .58). Among those discharged while taking OAC, 938 (45.6%) were discharged while taking DOAC in 2014 compared with 2868 (86.0%) in 2020. This increase in DOAC use was accompanied by a reciprocal decrease in warfarin use from 54.4% to 14.0%. In adjusted analyses, there were no significant interactions between race and ethnicity and admission year with respect to the odds of any OAC or DOAC receipt at discharge, indicating similar changes in anticoagulation over time in all racial and ethnic groups.
Clinical Outcomes at 1 Year
There were 16 307 individuals with CMS records available to examine postdischarge outcomes (eTable 4 in the Supplement). Overall, 1-year cumulative incidence of ischemic stroke was 1.9% (95% CI, 1.7-2.1) and was higher in Black patients (3.8%; 95% CI, 2.4-5.6) and Hispanic patients (3.3%; 95% CI, 2.1-5.0) than White patients (1.7%; 95% CI, 1.5-2.0) (eFigure 3 in the Supplement). Adjusted hazard ratios (aHRs) for stroke were higher in Black patients (aHR, 2.07; 95% CI, 1.33-3.20) and Hispanic patients (aHR, 2.02; 95% CI, 1.38-2.95) than White patients (Figure 3). Higher incidence of stroke was present in Black patients discharged while taking DOAC (aHR, 2.47; 95% CI, 1.43-4.27) and in Hispanic patients discharged without any OAC (aHR, 2.75; 95% CI, 1.20-6.28) or warfarin (aHR, 3.17; 95% CI, 1.56-6.41) compared with White patients.
Figure 3. Incidence of Clinical Outcomes at 1 Year by Race and Ethnicity and Anticoagulation Status for Patients With Atrial Fibrillation.
All analyses used Cox proportional hazards models, accounting for within-hospital clustering and adjusting for sociodemographic, clinical, and hospital factors. Compared with White patients, Black and Hispanic patients had significantly higher rates of ischemic stroke, and Black patients had higher rates of bleeding and mortality. Stroke incidence in Hispanic patients and mortality incidence in Black patients were higher when discharged without anticoagulation. aHR indicates adjusted hazard ratio; DOAC, direct-acting oral anticoagulants; NA, not available; OAC, oral anticoagulant.
Overall, 1-year cumulative incidence of major bleeding was 5.5% (95% CI, 5.2-5.9), with incidence higher in Black patients (11.3%; 95% CI, 8.9-14.1) and Hispanic patients (6.7%; 95% CI, 4.8-9.0) than White patients (5.3%; 95% CI, 4.9-5.6). Black patients had significantly higher bleeding risk than White patients (aHR, 2.08; 95% CI, 1.53-2.83). This bleeding risk was higher in Black patients discharged without any OAC (aHR, 3.15; 95% CI, 2.12-4.68) or when discharged while taking warfarin (aHR, 1.81; 95% CI, 1.04-3.15) or DOAC (aHR, 1.74; 95% CI, 1.10-2.74) compared with White patients (Figure 3).
Overall, 1-year cumulative incidence of mortality was 16.0% (95% CI, 15.4-16.6). Mortality risk was highest in Black patients (19.4%; 95% CI, 16.2-22.5), followed by Hispanic patients (18.3%; 95% CI, 15.3-21.3), Asian patients (16.6%; 95% CI, 11.1-22.2), and White patients (15.8%; 95% CI, 15.2-16.4) (Figure 3). Black patients had significantly higher risk of mortality compared with White patients (aHR, 1.22; 95% CI, 1.02-1.47). Black patients discharged without any OAC had significantly higher rates of mortality than White patients (aHR, 1.99; 95% CI, 1.54-2.57). When discharged while taking warfarin (aHR, 0.94; 95% CI, 0.67-1.30) or DOAC (aHR, 0.93; 95% CI, 0.67-1.28), the Black-White mortality difference was not present.
Discussion
In a national registry of more than 69 000 patients with AF, we observed 3 main findings. First, there were racial and ethnic differences in OAC use at discharge, with Black and Hispanic patients less likely to be prescribed anticoagulation and DOACs in particular, even among those with a high stroke risk. Second, these differences were seen despite a significant and similar absolute increase in OAC use over time observed in all racial and ethnic groups. Third, we observed racial and ethnic differences in clinical outcomes at 1 year, including higher rates of stroke, bleeding, and mortality in Black vs White patients and higher rates of stroke in Hispanic vs White patients.
Using contemporary data from a large national quality improvement AF registry, our findings extend prior research in detecting inequities in AF management and outcomes. In an analysis of the ORBIT-AF II trial, Black patients were significantly less likely than White patients to initiate any OAC, particularly DOACs.21 A similar finding was observed in a national cohort of patients receiving care in the Veterans Health Administration, which also found that Hispanic patients had lower DOAC use compared with White patients.13 Although the magnitude of racial and ethnic differences in the prior 2 studies were similar to the current study, these analyses focused on outpatient AF management and did not include an assessment of clinical outcomes. Moreover, few studies have examined racial and ethnic differences in AF outcomes by anticoagulation status as was done in the current study.14,23 An analysis of the Atherosclerosis Risk in Communities (ARIC) study did observe a 2-fold increase in stroke and mortality rates when comparing Black with White individuals with AF. Notably, the ARIC study was a smaller, community-based cohort limited to participants from 4 regions of the US and did not adjust for AF treatments, including anticoagulation.19
The observation that Black and Hispanic patients with AF are less likely to receive stroke-reducing anticoagulation, especially newer, more effective DOACs, is important for further understanding the barriers to equitable implementation of guideline-directed cardiovascular care.39 Furthermore, the finding that these differences in OAC receipt were further associated with higher rates of long-term outcomes and mortality has direct implications for quality of AF care.
The mechanisms by which racial and ethnic inequities in anticoagulation exist may include overattributing risk to patients from underrepresented racial and ethnic groups with adverse clinical and socioeconomic factors, including the presence of end-stage kidney disease, low income level, rurality or distance to care, limited AF awareness, and poor insurance access.22,40,41,42,43 These factors are particularly important when considering initiation of more costly DOAC therapy.44 Clinician implicit bias in prescribing has been observed in studies beyond AF management45,46 and, while difficult to measure in a large cohort analysis like this, may contribute to inequitable treatment. The undertreatment of Hispanic patients may also represent an opportunity to further explore limited English proficiency and immigration status as determinants of AF management and outcomes. Other factors such as cultural differences in perceptions of risk (ie, the need for OAC) or differential quality of cross-cultural shared decision-making are important to consider.47,48 Whereas health system–level differences, such as availability of electrophysiologists or hospital size, did not mediate the racial inequities observed in our analysis, such factors may reflect differential access to high-quality cardiovascular care across racial and ethnic groups.
Although ischemic stroke rates were overall higher in Black patients compared with White patients at 1 year, we also observed that stroke incidence was higher in Black than White patients discharged while taking DOACs. This observation is counter to prior literature and suggests the need to better understand OAC postdischarge adherence rates across racial groups as well as stroke risk factor modification among Black patients who may have higher rates of multiple stroke risk factors.41,49 Further qualitative research examining how clinicians assess stroke and bleeding risk among patients with AF from different racial and ethnic groups may help provide clarity to this study finding. Additionally, the finding that higher rates of bleeding were present in Black patients regardless of anticoagulation status represents an important opportunity for future research examining the safety of DOACs in underrepresented racial and ethnic groups.
Limitations
There are limitations to our study. First, data were collected by medical record review and claims data and are dependent on the accuracy and completeness of documentation and abstraction. Additionally, vital signs and laboratory values at discharge were not assessed in this analysis. Second, we were limited in the determination of type of AF for 11% of patients, whether AF occurred during hospitalization (eg, postprocedural), the duration or burden of AF, or whether AF was precipitated by a secondary extracardiac cause, all of which may influence anticoagulant decision-making. Third, because of the registry design, GWTG-AFIB attracts participating hospitals with an interest in quality improvement, potentially leading to selection bias. Fourth, because of the sample size, the analysis was limited in its assessment of patients who identified as American Indian, Alaska Native, or other racial groups. Even among the underrepresented racial and ethnic groups studied in this analysis (ie, Asian, Black, and Hispanic individuals), the proportion of the cohort was lower than their representation in the US population, which may have underestimated the differences in outcomes observed and suggests an opportunity to improve AF detection and awareness in these populations.50 Fifth, the registry is limited in the availability of individual-level clinician factors (eg, assessment of contraindications) and patient decision-making factors related to anticoagulation as well as social determinants that may influence anticoagulant receipt, such as housing instability or experiences with racism in health care. Such individual-level factors may also result in differential hospitalization for AF, which was not assessed. While an examination of how specific socioeconomic factors may intersect with race and ethnicity to influence anticoagulant receipt is key to future research, such an assessment was beyond the scope of this analysis. Sixth, postdischarge outcome assessments were confined to a shorter time frame in the subset of patients with available linkage to Medicare data. Similarly, postdischarge outpatient anticoagulant prescriptions were not assessed in this analysis. Seventh, while we used a broad definition for ischemic stroke using ICD-9 and ICD-10 codes (eTable 1 in the Supplement), we were limited in our assessment of stroke subtype and etiology. Eighth, as an observational analysis, there is concern for residual confounding, and we are unable to determine causality of the observed inequities.
Conclusions
In a national registry of patients hospitalized with AF, Black and Hispanic patients compared with White patients were less likely to be prescribed anticoagulation at hospital discharge, particularly DOACs. These differences persisted despite substantial improvements in overall OAC prescribing over time. Black patients also had higher rates of AF-related adverse outcomes, including stroke (also higher in Hispanic patients), bleeding, and mortality. Interventions to reduce racial and ethnic inequities in anticoagulation and AF outcomes are critical to improving quality of AF care.
eMethods.
eTable 1. ICD-9-CM and ICD-10-CM Diagnosis and Procedure Codes for Study Outcomes
eTable 2. Adjusted Odds Ratios for Anticoagulant Therapy at Discharge by Race and Ethnicity and CHA2DS2-VASc Score for Patients With Atrial Fibrillation
eTable 3. Rates of Oral Anticoagulant Prescriptions at Discharge by Race and Ethnicity and Medical Comorbidity, Atrial Fibrillation Type, and Diagnosis Position
eTable 4. Baseline Patient Characteristics in Patients Hospitalized With Atrial Fibrillation Enrolled in Medicare by Race and Ethnicity
eFigure 1. Hospital-Level Anticoagulant Use at Discharge for Patients With Atrial Fibrillation by Percentage of Non-White Patients
eFigure 2. Rates of Oral Anticoagulant Prescriptions at Discharge for Patients With Atrial Fibrillation by Race and Ethnicity, 2014-2020
eFigure 3. Cumulative incidence of Ischemic Stroke, Major Bleeding, and All-Cause Mortality by Race and Ethnicity Over 1 Year
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.Virani SS, Alonso A, Aparicio HJ, et al. ; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics—2021 update: a report from the American Heart Association. Circulation. 2021;143(8):e254-e743. doi: 10.1161/CIR.0000000000000950 [DOI] [PubMed] [Google Scholar]
- 3.Kornej J, Benjamin EJ, Magnani JW. Atrial fibrillation: global burdens and global opportunities. Heart. Published online January 28, 2021. doi: 10.1136/heartjnl-2020-318480 [DOI] [PubMed] [Google Scholar]
- 4.Kornej J, Börschel CS, Benjamin EJ, Schnabel RB. Epidemiology of atrial fibrillation in the 21st century: novel methods and new insights. Circ Res. 2020;127(1):4-20. doi: 10.1161/CIRCRESAHA.120.316340 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart study. Circulation. 1998;98(10):946-952. doi: 10.1161/01.CIR.98.10.946 [DOI] [PubMed] [Google Scholar]
- 6.Dieleman JL, Cao J, Chapin A, et al. US health care spending by payer and health condition, 1996-2016. JAMA. 2020;323(9):863-884. doi: 10.1001/jama.2020.0734 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Krishnamurthi N, Francis J, Fihn SD, Meyer CS, Whooley MA. Leading causes of cardiovascular hospitalization in 8.45 million US veterans. PLoS One. 2018;13(3):e0193996. doi: 10.1371/journal.pone.0193996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham study. Stroke. 1991;22(3):312-318. doi: 10.1161/01.STR.22.3.312 [DOI] [PubMed] [Google Scholar]
- 9.Di Pasquale G, Casella G. Antithrombotic strategies for atrial fibrillation: on the threshold of changes? yes. J Thromb Haemost. 2005;3(3):428-432. doi: 10.1111/j.1538-7836.2005.01182.x [DOI] [PubMed] [Google Scholar]
- 10.Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146(12):857-867. doi: 10.7326/0003-4819-146-12-200706190-00007 [DOI] [PubMed] [Google Scholar]
- 11.January CT, Wann LS, Calkins H, et al. 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. J Am Coll Cardiol. 2019;74(1):104-132. doi: 10.1016/j.jacc.2019.01.011 [DOI] [PubMed] [Google Scholar]
- 12.Xian Y, Xu H, O’Brien EC, et al. Clinical effectiveness of direct oral anticoagulants vs warfarin in older patients with atrial fibrillation and ischemic stroke: findings from the Patient-Centered Research Into Outcomes Stroke Patients Prefer and Effectiveness Research (PROSPER) study. JAMA Neurol. 2019;76(10):1192-1202. doi: 10.1001/jamaneurol.2019.2099 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Essien UR, Kim N, Hausmann LRM, et al. Disparities in anticoagulant therapy initiation for incident atrial fibrillation by race/ethnicity among patients in the Veterans Health Administration System. JAMA Netw Open. 2021;4(7):e2114234. doi: 10.1001/jamanetworkopen.2021.14234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sur NB, Wang K, Di Tullio MR, et al. Disparities and temporal trends in the use of anticoagulation in patients with ischemic stroke and atrial fibrillation. Stroke. 2019;50(6):1452-1459. doi: 10.1161/STROKEAHA.118.023959 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Piccini JP, Xu H, Cox M, et al. ; Get With The Guidelines-AFIB Clinical Working Group and Hospitals . Adherence to guideline-directed stroke prevention therapy for atrial fibrillation is achievable. Circulation. 2019;139(12):1497-1506. doi: 10.1161/CIRCULATIONAHA.118.035909 [DOI] [PubMed] [Google Scholar]
- 16.Gu K, Mahtta D, Kaneria A, et al. Racial disparities among Asian Americans with atrial fibrillation: an analysis from the NCDR® PINNACLE registry. Int J Cardiol. 2021;329:209-216. doi: 10.1016/j.ijcard.2020.12.064 [DOI] [PubMed] [Google Scholar]
- 17.Khalid U, Marzec LN, Mantini N, et al. Treatment of AF in American Indians and Alaska Natives: insights from the NCDR PINNACLE-AF registry. J Am Coll Cardiol. 2020;75(21):2749-2750. doi: 10.1016/j.jacc.2020.03.069 [DOI] [PubMed] [Google Scholar]
- 18.Mathur R, Pollara E, Hull S, Schofield P, Ashworth M, Robson J. Ethnicity and stroke risk in patients with atrial fibrillation. Heart. 2013;99(15):1087-1092. doi: 10.1136/heartjnl-2013-303767 [DOI] [PubMed] [Google Scholar]
- 19.Magnani JW, Norby FL, Agarwal SK, et al. Racial differences in atrial fibrillation-related cardiovascular disease and mortality: the Atherosclerosis Risk in Communities (ARIC) study. JAMA Cardiol. 2016;1(4):433-441. doi: 10.1001/jamacardio.2016.1025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sanchez JM, Jolly SE, Dewland TA, et al. Incident strokes among American Indian individuals with atrial fibrillation. J Am Heart Assoc. 2021;10(6):e019581. doi: 10.1161/JAHA.120.019581 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Essien UR, Holmes DN, Jackson LR II, et al. Association of race/ethnicity with oral anticoagulant use in patients with atrial fibrillation: findings from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation II. JAMA Cardiol. 2018;3(12):1174-1182. doi: 10.1001/jamacardio.2018.3945 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Essien UR, Magnani JW, Chen N, Gellad WF, Fine MJ, Hernandez I. Race/ethnicity and sex-related differences in direct oral anticoagulant initiation in newly diagnosed atrial fibrillation: a retrospective study of Medicare data. J Natl Med Assoc. 2020;112(1):103-108. doi: 10.1016/j.jnma.2019.10.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Golwala H, Jackson LR II, Simon DN, et al. ; Outcomes Registry for Better Informed Treatment for Atrial Fibrillation (ORBIT-AF) Registry . Racial/ethnic differences in atrial fibrillation symptoms, treatment patterns, and outcomes: insights from Outcomes Registry for Better Informed Treatment for Atrial Fibrillation Registry. Am Heart J. 2016;174:29-36. doi: 10.1016/j.ahj.2015.10.028 [DOI] [PubMed] [Google Scholar]
- 24.Yong C, Azarbal F, Abnousi F, et al. Racial differences in quality of anticoagulation therapy for atrial fibrillation (from the TREAT-AF study). Am J Cardiol. 2016;117(1):61-68. doi: 10.1016/j.amjcard.2015.09.047 [DOI] [PubMed] [Google Scholar]
- 25.Lewis WR, Piccini JP, Turakhia MP, et al. Get With The Guidelines AFIB: novel quality improvement registry for hospitalized patients with atrial fibrillation. Circ Cardiovasc Qual Outcomes. 2014;7(5):770-777. doi: 10.1161/CIRCOUTCOMES.114.001263 [DOI] [PubMed] [Google Scholar]
- 26.Heidenreich PA, Solis P, Estes NAM III, et al. 2016 ACC/AHA clinical performance and quality measures for adults with atrial fibrillation or atrial flutter: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures. J Am Coll Cardiol. 2016;68(5):525-568. doi: 10.1016/j.jacc.2016.03.521 [DOI] [PubMed] [Google Scholar]
- 27.Saeed O, Zhang S, Patel SR, et al. Oral anticoagulation and adverse outcomes after ischemic stroke in heart failure patients without atrial fibrillation. J Card Fail. 2021;27(8):857-864. doi: 10.1016/j.cardfail.2021.02.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hernandez AF, Mi X, Hammill BG, et al. Associations between aldosterone antagonist therapy and risks of mortality and readmission among patients with heart failure and reduced ejection fraction. JAMA. 2012;308(20):2097-2107. doi: 10.1001/jama.2012.14795 [DOI] [PubMed] [Google Scholar]
- 29.Reeves MJ, Fonarow GC, Smith EE, et al. Representativeness of the Get With The Guidelines-Stroke Registry: comparison of patient and hospital characteristics among Medicare beneficiaries hospitalized with ischemic stroke. Stroke. 2012;43(1):44-49. doi: 10.1161/STROKEAHA.111.626978 [DOI] [PubMed] [Google Scholar]
- 30.Hammill BG, Hernandez AF, Peterson ED, Fonarow GC, Schulman KA, Curtis LH. Linking inpatient clinical registry data to Medicare claims data using indirect identifiers. Am Heart J. 2009;157(6):995-1000. doi: 10.1016/j.ahj.2009.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Guimarães PO, Krishnamoorthy A, Kaltenbach LA, et al. Accuracy of medical claims for identifying cardiovascular and bleeding events after myocardial infarction : a secondary analysis of the TRANSLATE-ACS study. JAMA Cardiol. 2017;2(7):750-757. doi: 10.1001/jamacardio.2017.1460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Essien UR, Kim N, Magnani JW, et al. Association of race and ethnicity and anticoagulation in patients with atrial fibrillation dually enrolled in Veterans Health Administration and Medicare: effects of Medicare Part D on prescribing disparities. Circ Cardiovasc Qual Outcomes. 2022;15(2):e008389. doi: 10.1161/CIRCOUTCOMES.121.008389 [DOI] [PubMed] [Google Scholar]
- 33.Hess PL, Hernandez AF, Bhatt DL, et al. Sex and race/ethnicity differences in implantable cardioverter-defibrillator counseling and use among patients hospitalized with heart failure: findings from the Get With The Guidelines-Heart Failure Program. Circulation. 2016;134(7):517-526. doi: 10.1161/CIRCULATIONAHA.115.021048 [DOI] [PubMed] [Google Scholar]
- 34.Miller AL, Simon D, Roe MT, et al. Comparison of delay times from symptom onset to medical contact in blacks versus whites with acute myocardial infarction. Am J Cardiol. 2017;119(8):1127-1134. doi: 10.1016/j.amjcard.2016.12.021 [DOI] [PubMed] [Google Scholar]
- 35.Bucholz EM, Ma S, Normand SL, Krumholz HM. Race, socioeconomic status, and life expectancy after acute myocardial infarction. Circulation. 2015;132(14):1338-1346. doi: 10.1161/CIRCULATIONAHA.115.017009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.US Census Bureau . American Community Survey Question Detail. Accessed February 18, 2022. https://www.census.gov/programs-surveys/acs
- 37.O’Brien EC, Simon DN, Thomas LE, et al. The ORBIT bleeding score: a simple bedside score to assess bleeding risk in atrial fibrillation. Eur Heart J. 2015;36(46):3258-3264. doi: 10.1093/eurheartj/ehv476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Dinse GE, Lagakos SW. Nonparametric estimation of lifetime and disease onset distributions from incomplete observations. Biometrics. 1982;38(4):921-932. doi: 10.2307/2529872 [DOI] [PubMed] [Google Scholar]
- 39.Essien UR, Dusetzina SB, Gellad WF. A policy prescription for reducing health disparities-achieving pharmacoequity. JAMA. 2021;326(18):1793-1794. doi: 10.1001/jama.2021.17764 [DOI] [PubMed] [Google Scholar]
- 40.Waddy SP, Solomon AJ, Becerra AZ, et al. Racial/ethnic disparities in atrial fibrillation treatment and outcomes among dialysis patients in the United States. J Am Soc Nephrol. 2020;31(3):637-649. doi: 10.1681/ASN.2019050543 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Tedla YG, Schwartz SM, Silberman P, Greenland P, Passman RS. Racial disparity in the prescription of anticoagulants and risk of stroke and bleeding in atrial fibrillation patients. J Stroke Cerebrovasc Dis. 2020;29(5):104718. doi: 10.1016/j.jstrokecerebrovasdis.2020.104718 [DOI] [PubMed] [Google Scholar]
- 42.Schaefer JK, Sood SL, Haymart B, et al. Sociodemographic factors in patients continuing warfarin vs those transitioning to direct oral anticoagulants. Blood Adv. 2017;1(26):2536-2540. doi: 10.1182/bloodadvances.2017012377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Yong CM, Liu Y, Apruzzese P, et al. ; ACC PINNACLE Investigators . Association of insurance type with receipt of oral anticoagulation in insured patients with atrial fibrillation: a report from the American College of Cardiology NCDR PINNACLE registry. Am Heart J. 2018;195:50-59. doi: 10.1016/j.ahj.2017.08.010 [DOI] [PubMed] [Google Scholar]
- 44.Rome BN, Gagne JJ, Avorn J, Kesselheim AS. Non-warfarin oral anticoagulant copayments and adherence in atrial fibrillation: a population-based cohort study. Am Heart J. 2021;233:109-121. doi: 10.1016/j.ahj.2020.12.010 [DOI] [PubMed] [Google Scholar]
- 45.Goyal MK, Johnson TJ, Chamberlain JM, et al. ; Pediatric Care Applied Research Network (PECARN) . Racial and ethnic differences in antibiotic use for viral illness in emergency departments. Pediatrics. 2017;140(4):e20170203. doi: 10.1542/peds.2017-0203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Keister LA, Stecher C, Aronson B, McConnell W, Hustedt J, Moody JW. Provider bias in prescribing opioid analgesics: a study of electronic medical records at a hospital emergency department. BMC Public Health. 2021;21(1):1518. doi: 10.1186/s12889-021-11551-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Fang MC, Panguluri P, Machtinger EL, Schillinger D. Language, literacy, and characterization of stroke among patients taking warfarin for stroke prevention: implications for health communication. Patient Educ Couns. 2009;75(3):403-410. doi: 10.1016/j.pec.2008.12.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Chung MK, Fagerlin A, Wang PJ, et al. Shared decision making in cardiac electrophysiology procedures and arrhythmia management. Circ Arrhythm Electrophysiol. 2021;14(12):e007958. doi: 10.1161/CIRCEP.121.007958 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Norby FL, Benjamin EJ, Alonso A, Chugh SS. Racial and ethnic considerations in patients with atrial fibrillation: JACC focus seminar 5/9. J Am Coll Cardiol. 2021;78(25):2563-2572. doi: 10.1016/j.jacc.2021.04.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Essien UR, Kornej J, Johnson AE, Schulson LB, Benjamin EJ, Magnani JW. Social determinants of atrial fibrillation. Nat Rev Cardiol. 2021;18(11):763-773. doi: 10.1038/s41569-021-00561-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eTable 1. ICD-9-CM and ICD-10-CM Diagnosis and Procedure Codes for Study Outcomes
eTable 2. Adjusted Odds Ratios for Anticoagulant Therapy at Discharge by Race and Ethnicity and CHA2DS2-VASc Score for Patients With Atrial Fibrillation
eTable 3. Rates of Oral Anticoagulant Prescriptions at Discharge by Race and Ethnicity and Medical Comorbidity, Atrial Fibrillation Type, and Diagnosis Position
eTable 4. Baseline Patient Characteristics in Patients Hospitalized With Atrial Fibrillation Enrolled in Medicare by Race and Ethnicity
eFigure 1. Hospital-Level Anticoagulant Use at Discharge for Patients With Atrial Fibrillation by Percentage of Non-White Patients
eFigure 2. Rates of Oral Anticoagulant Prescriptions at Discharge for Patients With Atrial Fibrillation by Race and Ethnicity, 2014-2020
eFigure 3. Cumulative incidence of Ischemic Stroke, Major Bleeding, and All-Cause Mortality by Race and Ethnicity Over 1 Year