This cohort study assesses whether associations exist between patient-physician race, ethnicity, and language concordance and direct oral anticoagulant (DOAC) treatment initiation and persistence among patients with nonvalvular atrial fibrillation.
Key Points
Question
Are there associations between patient-physician race, ethnicity, and language concordance and direct oral anticoagulant (DOAC) therapy initiation and persistence among patients with nonvalvular atrial fibrillation (NVAF)?
Findings
In this cohort study of 59 340 adults with NVAF, patient-physician race and ethnicity concordance was associated with increased odds of both DOAC therapy initiation and persistence. Patient-physician language concordance was associated with increased odds of DOAC therapy persistence.
Meaning
Findings from this study suggest that patient-physician race, ethnicity, and language concordance may facilitate more equitable use of DOACs among patients with NVAF.
Abstract
Importance
While direct oral anticoagulants (DOACs) are recommended for treatment of patients with nonvalvular atrial fibrillation (NVAF) to reduce risk of mortality and stroke, studies have documented racial and ethnic disparities in DOAC use.
Objective
To evaluate the associations between patient-physician race, ethnicity, and language concordance and DOAC treatment initiation and persistence.
Design, Setting, and Participants
This retrospective cohort study was conducted using Kaiser Permanente Southern California electronic health records from January 1, 2012, through December 31, 2022. Adults 18 years of age or older with a confirmed diagnosis of NVAF and a CHA2DS2-VASc (congestive heart failure; hypertension; age ≥75 years; diabetes; prior stroke, transient ischemic attack, or thromboembolism; vascular disease; age 65-74 years; and sex category) score of 2 or higher were followed up from their first NVAF diagnosis to 60 days for a DOAC prescription order or from the first DOAC prescription fill to 1 year. Analysis was performed between April 2024 and January 2026.
Exposures
Patient-physician race, ethnicity, and language concordance.
Main Outcomes and Measures
Associations of patient-physician race, ethnicity, and language concordance with DOAC therapy initiation (defined as the receipt of a DOAC prescription within 60 days from an initial NVAF diagnosis) and persistence (defined as 1 year of continuous DOAC prescription fills without a gap of 60 or more days) were determined using multivariable logistic regression models, accounting for age, sex, patient race and ethnicity, health insurance type, and neighborhood deprivation index.
Results
Of 59 340 DOAC therapy–eligible patients (mean [SD] age, 76.9 [9.1] years; 32 889 [55.4%] females; 12 440 [21.0%] Hispanic, 4803 [8.1%] non-Hispanic Asian, 5374 [9.1%] non-Hispanic Black, and 35 645 [60.1%] non-Hispanic White), 34 281 (57.8%) had a DOAC prescription order. Patient-physician race and ethnicity concordance was associated with increased odds of DOAC therapy initiation (odds ratio [OR], 1.19 [95% CI, 1.14-1.24]). Among 34 215 patients with a DOAC prescription fill, 45.3% were persistent with DOAC therapy for 1 year. Both patient-physician race and ethnicity (OR, 1.08 [95% CI, 1.03-1.14]) and language (OR, 1.14 [95% CI, 1.04-1.24]) concordance were associated with increased odds of DOAC therapy persistence. Among Hispanic patients, patient-physician race and ethnicity concordance (OR, 1.39 [95% CI, 1.22-1.58]) was associated with increased odds of DOAC therapy initiation, and language concordance (OR, 1.12 [95% CI, 1.01-1.24]) was associated with increased odds of DOAC therapy persistence. Neither patient-physician race and ethnicity (OR, 1.04 [95% CI, 0.78-1.39] for initiation; OR, 1.03 [95% CI, 0.72-1.47] for persistence) nor language (OR, 0.61 [95% CI, 0.24-1.55] for initiation; OR, 1.62 [95% CI, 0.53-4.93] for persistence) concordance was associated with DOAC therapy initiation or persistence among non-Hispanic Black patients.
Conclusions and Relevance
In this cohort study, patient-physician race and ethnicity concordance was associated with increased odds of DOAC therapy initiation, whereas race, ethnicity, and language concordance were associated with increased odds of DOAC therapy persistence. These findings suggest patient-physician concordance as a potential pathway to support more equitable anticoagulation care.
Introduction
Nonvalvular atrial fibrillation (NVAF) is a common subtype of atrial fibrillation (AF).1 NVAF is associated with up to a 5-fold increased risk of ischemic stroke compared with individuals without AF.2,3 For patients with NVAF, the CHA2DS2-VASc (congestive heart failure; hypertension; age ≥75 years; diabetes; prior stroke, transient ischemic attack, or thromboembolism; vascular disease; age 65-74 years; and sex category) score is commonly used to assess stroke risk and guide anticoagulation therapy decisions. For men and women with a CHA2DS2-VASc score greater than or equal to 2 and 3, respectively, clinical guidelines generally recommend anticoagulation with direct oral anticoagulants (DOACs) over warfarin to reduce the risk of mortality and stroke and to minimize the bleeding risk.4,5
Notably, the literature highlights substantial racial and ethnic disparities in the initiation of DOAC therapy among patients with NVAF. Multiple large registry and cohort studies have demonstrated that Hispanic and non-Hispanic Black patients are less likely to initiate anticoagulant therapy (either warfarin or DOACs), and among those treated with any anticoagulant therapy, this patient subpopulation is less likely to receive DOACs compared with their non-Hispanic White counterparts.6,7,8,9 In contrast, the literature regarding DOAC therapy persistence is mixed. While some studies have demonstrated an association between racial and ethnic minority groups and early DOAC discontinuation or nonadherence,10,11 others have found minimal racial and ethnic disparities in DOAC therapy persistence.9 Importantly, these racial and ethnic disparities in AF management may also contribute to differences in clinical outcomes. Prior studies have demonstrated that non-Hispanic Black patients with AF have disproportionately higher rates of ischemic stroke compared with their non-Hispanic White counterparts and are at greater risk for associated adverse outcomes such as stroke, heart failure, coronary heart disease, and all-cause mortality.12,13
Taken together, these findings highlight ongoing inequities in anticoagulation management for patients with NVAF. Limited English proficiency, cultural differences, and clinician implicit bias have all been proposed as contributing factors6,9,14 and may be especially relevant in the management of NVAF, in which decisions regarding anticoagulation often involve comprehensive and nuanced discussions surrounding stroke prevention, bleeding risk, and sustained medication adherence despite the frequently asymptomatic nature of the disease. Patient-physician race, ethnicity, and language concordance may help mitigate these influences. These concordances—potentially reflecting communication, mutual cultural affinity, and rapport-building—have been associated with improved medication initiation and adherence in a variety of clinical settings among historically marginalized groups15,16,17,18; thus, their association with DOAC therapy initiation and persistence warrants further exploration. We hypothesized that patient-physician race, ethnicity, and language concordance would be associated with increased odds of DOAC therapy initiation and persistence.
Methods
The retrospective cohort study protocol was approved by the Kaiser Permanente Southern California (KPSC) institutional review board. A waiver of informed consent was granted because the research involved no more than minimal risk to research participants. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline, including explicit reporting of cohort selection, variable definitions, and statistical methods. Analysis was performed between April 2024 and January 2026.
Study Setting
A retrospective cohort study was conducted using data from KPSC electronic health records. KPSC currently serves more than 4.9 million members across Southern California through their own facilities, largely reflecting the populations residing in the region.
Study Cohort
We included KPSC members older than 18 years with a confirmed diagnosis of NVAF between January 1, 2012, and December 31, 2022. The first diagnosis of NVAF was defined as the index date. Among those individuals, we included only patients with a CHA2DS2-VASc score greater than or equal to 2, consistent with prior observational studies19,20 and reflecting earlier guideline recommendations and evolving clinical practice from 2012 through 2022. We also required continuous insurance coverage at least 1 year prior to and 60 days after the index date. The CHA2DS2-VASc score was calculated by assigning 1 point each for female sex, age 65 to 74 years at the index date, hypertension, diabetes, vascular disease, and congestive heart failure, and assigning 2 points each for age greater than or equal to 75 years and history of stroke, transient ischemic attack, or thromboembolism using the data 1 year prior to the index date.21 To focus on anticoagulation therapy initiation following incident NVAF diagnosis, individuals with a history of NVAF were excluded. Individuals with valvular disease, including moderate to severe mitral stenosis or mechanical heart valve, as well as those with bleeding disorders or known coagulopathies were also excluded. Additionally, patients with potential contraindications to DOACs were excluded, including those with active cerebrovascular, gastrointestinal, or other bleeding events within 6 months, severe liver disease, recent cancer treatment, or a history of heparin-induced thrombocytopenia. Pregnant individuals or individuals residing in skilled nursing facilities or hospices within 60 days of the index date were also excluded. Definitions of exclusion criteria and study variables are listed in eTable 1 in Supplement 1. For DOAC therapy persistence analysis, we further restricted the study population to individuals with at least 1 DOAC prescription fill within 60 days from the DOAC prescription order and 1 year continuous eligibility from the first DOAC prescription fill (Figure).
Figure. Flow Diagram of the Study Cohort.
CHA2DS2-VASc score represents congestive heart failure; hypertension; age ≥75 years; diabetes; prior stroke, transient ischemic attack, or thromboembolism; vascular disease; age 65-74 years; and sex category. DOAC indicates direct oral anticoagulant; NVAF, nonvalvular atrial fibrillation.
DOAC Therapy Initiation
DOAC therapy initiation was defined as a DOAC prescription order (dabigatran, apixaban, rivaroxaban, and edoxaban) from a physician within 60 days from an initial diagnosis of NVAF. The 60-day cutoff was chosen to account for the usual dispensing interval for anticoagulants and to accommodate patients who required additional evaluation or shared decision-making discussions.22,23
DOAC Therapy Persistence
DOAC therapy persistence was defined as patients having 1 year continuous DOAC use from their initial fill of the DOAC prescription without a refill gap of 60 days or greater. The 60-day gap is frequently used to define DOAC therapy persistence and represents a balance between capturing clinically meaningful interruptions and accounting for variability in prescription refills.24,25,26
Race, Ethnicity, and Language Concordance
Patient and physician self-reported race, ethnicity, and language data were obtained from electronic health records or administrative files. We identified the physician who initially diagnosed the patient as having NVAF and assessed whether the patient and diagnosing physician reported the same race. For patients, race and ethnicity were categorized as Hispanic, multiracial, Native American or Alaska Native, non-Hispanic Asian, non-Hispanic Black, non-Hispanic White, Native Hawaiian or Other Pacific Islander, other, and unknown based on self-reported data following the US Office of Management and Budget standards for the classification of federal data on race and ethnicity. For clinicians, race and ethnicity were similarly based on self-reported data, and the same categories were available except for multiple race and ethnicity. Patient-physician language concordance was defined as the physician reporting fluency in the patient’s preferred language. Individuals with missing race, ethnicity, or language information were categorized as unknown and treated as nonconcordant in analyses.
Statistical Analysis
We used descriptive statistics (mean and SD for continuous variables, and count and percentages for categorical variables) to summarize patient sociodemographic and clinical characteristics. Univariate and multivariable logistic regression models were used to determine the associations between patient-physician race, ethnicity, and language concordance and DOAC therapy initiation and persistence. Covariates for multivariable models were selected a priori based on a conceptual framework incorporating patient-level demographic characteristics (age, sex, race, and ethnicity), social determinants of health (insurance type and neighborhood deprivation index quartile), and clinical risk factors (CHA2DS2-VASc score for persistence analysis) known to influence anticoagulation initiation and persistence. Insurance type was included as a categorical variable (commercial only, commercial and Medicare, Medicaid only, Medicaid and Medicare, Medicare only, private pay, other, or unknown, with Medicare only as the reference group). Neighborhood deprivation index was modeled in quartiles with the least deprived quartile as the reference group. We also conducted prespecified subgroup analyses for Hispanic and non-Hispanic Black patients to assess whether associations differed within these populations. The unadjusted and adjusted odds ratios (ORs) and 95% CIs from univariate and multivariable models were reported, with statistical significance defined as a 2-sided P < .05. All analyses were performed using SAS, version 9.4 (SAS Institute Inc).
Results
Study Cohort
The study cohort consisted of 59 340 patients with a mean (SD) age of 76.9 (9.1) years. Females comprised 55.4% of the cohort (n = 32 889); males, 44.6% (n = 26 451). The cohort included 12 440 [21.0%] Hispanic, 4803 [8.1%] non-Hispanic Asian, 5374 [9.1%] non-Hispanic Black, and 35 645 [60.1%] non-Hispanic White adults; the remaining 2.6% of the cohort was composed of 831 adults who self-reported other race and ethnicity and 247 adults whose race and ethnicity were unknown (Table 1). A majority of the study cohort spoke English (53 333 [89.9%]) as their primary language; other languages included Spanish (4788 [8.1%]), Tagalog (174 [0.3%]), Korean (100 [0.2%]), Mandarin (139 [0.2%]), and Vietnamese (115 [0.2%]).
Table 1. Characteristics of the Study Population for DOAC Therapy Initiation by Racial and Ethnic Group.
| Characteristic | Participants, No. (%) | ||||||
|---|---|---|---|---|---|---|---|
| Total (N = 59 340) | Non-Hispanic White (n = 35 645) | Non-Hispanic Black (n = 5374) | Hispanic (n = 12 440) | Non-Hispanic Asian (n = 4803) | Other (n = 831)a | Unknown (n = 247)b | |
| Age at index date, mean (SD), y | 76.9 (9.1) | 77.6 (8.7) | 76.1 (10.1) | 76.0 (9.7) | 75.9 (9.0) | 74.0 (10.0) | 74.3 (8.8) |
| Sex | |||||||
| Female | 32 889 (55.4) | 19 486 (54.7) | 3305 (61.5) | 6933 (55.7) | 2615 (54.4) | 441 (53.1) | 109 (44.1) |
| Male | 26 451 (44.6) | 16 159 (45.3) | 2069 (38.5) | 5507 (44.3) | 2188 (45.6) | 390 (46.9) | 138 (55.9) |
| Insurance type | |||||||
| Commercial only | 7560 (12.7) | 3692 (10.4) | 1092 (20.3) | 1734 (13.9) | 808 (16.8) | 174 (20.9) | 60 (24.3) |
| Commercial and Medicare | 342 (0.6) | 184 (0.5) | 48 (0.9) | 78 (0.6) | 25 (0.5) | 7 (0.8) | 0 |
| Medicaid only | 468 (0.8) | 115 (0.3) | 71 (1.3) | 191 (1.5) | 78 (1.6) | 11 (1.3) | 2 (0.8) |
| Medicaid and Medicare | 2564 (4.3) | 496 (1.4) | 252 (4.7) | 1450 (11.7) | 313 (6.5) | 52 (6.3) | 1 (0.4) |
| Medicare only | 48 039 (81.0) | 30947 (86.8) | 3897 (72.5) | 8904 (71.6) | 3535 (73.6) | 576 (69.3) | 180 (72.9) |
| Private pay | 348 (0.6) | 204 (0.6) | 12 (0.2) | 75 (0.6) | 42 (0.9) | 11 (1.3) | 4 (1.6) |
| Other | 7 (0.0) | 4 (0.0) | 1 (0.0) | 1 (0.0) | 1 (0.0) | 0 | 0 |
| Unknown | 12 (0.0) | 3 (0.0) | 1 (0.0) | 7 (0.1) | 1 (0.0) | 0 | 0 |
| NDI quartile | |||||||
| 1 (Least deprived) | 15 110 (25.5) | 11 854 (33.3) | 415 (7.7) | 1293 (10.4) | 1303 (27.1) | 173 (20.8) | 72 (29.1) |
| 2 | 14 895 (25.1) | 10 373 (29.1) | 771 (14.3) | 2167 (17.4) | 1285 (26.8) | 239 (28.8) | 60 (24.3) |
| 3 | 14 781 (24.9) | 8143 (22.8) | 1500 (27.9) | 3565 (28.7) | 1270 (26.4) | 236 (28.4) | 67 (27.1) |
| 4 | 14 486 (24.4) | 5226 (14.7) | 2680 (49.9) | 5409 (43.5) | 944 (19.7) | 181 (21.8) | 46 (18.6) |
| Language concordance | 54 253 (91.4) | 35 314 (99.1) | 5351 (99.6) | 8454 (68.0) | 4107 (85.5) | 791 (95.2) | 236 (95.5) |
| Race and ethnicity concordance | 15 478 (26.1) | 11 483 (32.2) | 235 (4.4) | 1337 (10.7) | 2394 (49.8) | 19 (2.3) | 10 (4.0) |
| Race and ethnicity and language concordance | 14 859 (25.0) | 11 381 (31.9) | 234 (4.4) | 1192 (9.6) | 2023 (42.1) | 19 (2.3) | 10 (4.0) |
Abbreviations: DOAC, direct oral anticoagulant; NDI, neighborhood deprivation index.
Other includes multiracial, Native American or Alaska Native, Native Hawaiian or Other Pacific Islander, and self-reported other.
Unknown includes missing or self-reported unknown.
For patients included in the study cohort, 10 579 diagnosing physicians were identified, of whom 3538 (33.4%) were female, 4929 (46.6%) were male, and 2112 (20.0%) were unknown; 831 (7.9%) were Hispanic, 3560 (33.7%) were non-Hispanic Asian, 335 (3.2%) were non-Hispanic Black, and 3140 (29.7%) were non-Hispanic White; 2620 physicians’ race and ethnicity (24.8%) were unknown. Most physicians (n = 8878 [83.9%]) spoke English alone; other language fluencies in addition to English included Spanish (836 [7.9%]), Mandarin (159 [1.5%]), Farsi (107 [1.0%]), Vietnamese (87 [0.8%), Korean (62 [0.6%]), and Tagalog (47 [0.4%]).
In the study cohort, 34 281 patients (57.8%) initiated treatment with a DOAC within 60 days from an initial NVAF diagnosis. By race and ethnic groups, the corresponding percentages were 21.7% for Hispanic (n = 7449), 8.5% for non-Hispanic Asian (n = 2905), 8.4% for non-Hispanic Black (n = 2866), 59.5% for non-Hispanic White (n = 20 407), 1.5% for other (n = 518), and 0.4% for unknown (n = 136) race and ethnicity.
Of 59 340 patients, 15 478 (26.1%) were race and ethnicity concordant with their physician, and 54 253 (91.4%) were language concordant with their physician (Table 1). Race and ethnicity concordance was observed for 1337 Hispanic patients (10.7%), 2394 non-Hispanic Asian patients (49.8%), 235 non-Hispanic Black patients (4.4%), and 11 483 non-Hispanic White patients (32.2%). Language concordance was highest among non-Hispanic Black (5351 [99.6%]) and non-Hispanic White (35 314 [99.1%]) adults, followed by non-Hispanic Asian (4107 [85.5%]), and Hispanic (8454 [68.0%]) adults.
Concordance and DOAC Therapy Initiation
Univariate analysis showed that patient-physician race and ethnicity concordance (unadjusted OR, 1.19 [95% CI, 1.14-1.23]) was associated with increased odds of DOAC therapy initiation, whereas language concordance (unadjusted OR, 1.00 [95% CI, 0.94-1.06]) was not (Table 2). After adjusting for covariates, the results continued to show that patient-physician race and ethnicity concordance (adjusted OR [AOR], 1.19 [95% CI, 1.14-1.24]), but not patient-physician language concordance (AOR, 1.04 [95% CI, 0.97-1.12]), was associated with increased odds of DOAC therapy initiation. In the Hispanic subgroup (n = 12 440), only patient-physician race and ethnicity concordance (AOR, 1.39 [95% CI, 1.22-1.58]) was associated with increased odds of DOAC therapy initiation (language concordance AOR, 1.01 [95% CI, 0.96-1.12]) (eTable 2 in Supplement 1). Neither patient-physician race and ethnicity concordance (AOR, 1.04 [95% CI, 0.78-1.39]) nor language concordance (AOR, 0.61 [95% CI, 0.24-1.55]) was associated with DOAC therapy initiation in the non-Hispanic Black subgroup (n = 5374) (eTable 3 in Supplement 1).
Table 2. Patient-Physician Concordance and DOAC Therapy Initiation.
| Exposure | Race and ethnicity concordance model | Language concordance model |
|---|---|---|
| Univariate model, crude odds ratios (95% CI)a | ||
| Patient-physician race and ethnicity concordance vs discordance | 1.19 (1.14-1.23) | NA |
| Patient-physician language concordance vs discordance | NA | 1.00 (0.94-1.06) |
| Multivariable model, adjusted odds ratios (95% CI)b | ||
| Patient-physician race and ethnicity concordance vs discordance | 1.19 (1.14-1.24) | NA |
| Patient-physician language concordance vs discordance | NA | 1.04 (0.97-1.12) |
| Patient age group at index date, y | ||
| 75-84 vs <75 | 0.32 (0.30-0.33) | 0.32 (0.30-0.33) |
| ≥85 vs <75 | 0.16 (0.16-0.17) | 0.16 (0.16-0.17) |
| Patient sex, female (compared with male) | 0.60 (0.58-0.62) | 0.60 (0.58-0.62) |
| Patient race and ethnicity (compared with non-Hispanic White) | ||
| Hispanic | 1.09 (1.04-1.15) | 1.06 (1.01-1.12) |
| Native American or Alaska Native | 0.99 (0.67-1.46) | 0.94 (0.64-1.38) |
| Non-Hispanic Asian | 0.96 (0.89-1.02) | 0.99 (0.93-1.06) |
| Non-Hispanic Black | 0.84 (0.79-0.90) | 0.80 (0.75-0.86) |
| Native Hawaiian or Other Pacific Islander | 0.86 (0.69-1.07) | 0.81 (0.65-1.02) |
| Multiracialc | 0.86 (0.55-1.36) | 0.82 (0.52-1.28) |
| Otherd | 1.45 (1.06-1.99) | 1.38 (1.00-1.89) |
| Unknown | 0.60 (0.45-0.80) | 0.57 (0.42-0.76) |
| Patient insurance type (vs Medicare only) | ||
| Commercial only | 1.73 (1.62-1.84) | 1.72 (1.62-1.83) |
| Commercial and Medicare | 0.88 (0.70-1.11) | 0.88 (0.70-1.11) |
| Medicaid | 1.03 (0.84-1.26) | 1.04 (0.85-1.28) |
| Medicaid and Medicare | 1.39 (1.27-1.52) | 1.40 (1.28-1.54) |
| Private pay | 5.42 (3.52-8.33) | 5.43 (3.53-8.35) |
| Other | 2.26 (0.26-19.30) | 2.37 (0.27-20.55) |
| Unknown | 0.78 (0.23-2.61) | 0.78 (0.23-2.60) |
| NDI (vs Q1, least deprived) | ||
| Q2 | 0.90 (0.86-0.95) | 0.90 (0.86-0.94) |
| Q3 | 0.91 (0.87-0.96) | 0.91 (0.86-0.95) |
| Q4 | 0.83 (0.78-0.87) | 0.82 (0.78-0.87) |
| Unknown Q | 0.65 (0.39-1.09) | 0.64 (0.38-1.08) |
Abbreviations: DOAC, direct oral anticoagulant; NA, not applicable; NDI, neighborhood deprivation index; Q, quartile.
Univariate logistic regression.
Multivariable logistic regression, adjusted for patient age, sex, race and ethnicity, insurance type, and neighborhood deprivation index quartile.
Individuals who self-identified with more than 1 race or ethnicity.
Self-reported other.
Concordance and DOAC Therapy Persistence
Within the study cohort, 34 215 patients met the criteria for DOAC therapy persistence, and 45.3% were persistent with DOAC therapy for 1 year. By race and ethnic groups, the corresponding percentages were 50.3% for Hispanic, 42.0% for non-Hispanic Asian, 53.7% for non-Hispanic Black, 42.8% for non-Hispanic White, 43.9% for other, and 43.3% for unknown race and ethnicity.
The DOAC therapy persistence substudy population characteristics are described in eTable 4 in Supplement 1. Univariate analysis demonstrated that race and ethnicity concordance (unadjusted OR, 1.20 [95% CI, 1.14-1.26]) and language concordance (unadjusted OR, 1.33 [95% CI, 1.23-1.43]) were associated with increased odds of DOAC therapy persistence (Table 3). Multivariable analysis demonstrated similar results; both patient-physician race and ethnicity concordance (AOR, 1.08 [95% CI, 1.03-1.14]) and language concordance (AOR, 1.14 [95% CI, 1.04-1.24]) were associated with increased odds of DOAC therapy persistence. In the Hispanic subgroup (n = 7462), patient-physician language concordance (AOR, 1.12 [95% CI, 1.01-1.24]), but not race and ethnicity concordance (AOR, 1.07 [95% CI, 0.93-1.23]), was associated with increased odds of DOAC therapy persistence (eTable 5 in Supplement 1). By comparison, neither patient-physician race and ethnicity concordance (AOR, 1.03 [95% CI, 0.72-1.47]) nor language concordance (AOR, 1.62 [95% CI, 0.53-4.93]) were associated with DOAC therapy persistence in the non-Hispanic Black subgroup (n = 2862) (eTable 6 in Supplement 1).
Table 3. Patient-Physician Concordance and DOAC Therapy Persistence.
| Exposure | Race and ethnicity concordance model | Language concordance model |
|---|---|---|
| Univariate model, crude odds ratios (95% CI)a | ||
| Patient-physician race and ethnicity concordance vs discordance | 1.20 (1.14-1.26) | NA |
| Patient-physician language concordance vs discordance | NA | 1.33 (1.23-1.43) |
| Multivariable model, adjusted odds ratios (95% CI)b | ||
| Patient-physician race and ethnicity concordance vs discordance | 1.08 (1.03-1.14) | NA |
| Patient-physician language concordance vs discordance | NA | 1.14 (1.04-1.24) |
| Patient age group at index date (vs <65), y | ||
| 65-74 | 1.26 (1.14-1.38) | 1.26 (1.15-1.39) |
| 75-84 | 1.18 (1.06-1.31) | 1.19 (1.07-1.32) |
| ≥85 | 0.93 (0.83-1.05) | 0.94 (0.84-1.06) |
| Patient sex, female (compared with male) | 1.15 (1.10-1.21) | 1.16 (1.10-1.21) |
| Patient race and ethnicity (compared with non-Hispanic White) | ||
| Hispanic | 0.84 (0.79-0.89) | 0.85 (0.80-0.91) |
| Native American or Alaska Native | 1.21 (0.76-1.94) | 1.18 (0.74-1.90) |
| Non-Hispanic Asian | 1.05 (0.91-1.14) | 1.08 (1.00-1.17) |
| Non-Hispanic Black | 0.74 (0.68-0.81) | 0.72 (0.67-0.78) |
| Native Hawaiian or Other Pacific Islander | 1.03 (0.79-1.34) | 1.01 (0.77-1.31) |
| Multiracialc | 0.78 (0.45-1.35) | 0.76 (0.44-1.31) |
| Otherd | 0.95 (0.69-1.33) | 0.93 (0.67-1.29) |
| Unknown | 1.02 (0.69-1.50) | 0.99 (0.67-1.46) |
| Patient insurance type (vs Medicare only) | ||
| Commercial only | 0.95 (0.89-1.03) | 0.95 (0.89-1.03) |
| Commercial and Medicare | 1.13 (0.85-1.49) | 1.12 (0.84-1.48) |
| Medicaid | 0.76 (0.59-0.96) | 0.77 (0.61-0.98) |
| Medicaid and Medicare | 0.95 (0.85-1.05) | 0.98 (0.87-1.09) |
| Private pay | 0.83 (0.66-1.05) | 0.84 (0.66-1.06) |
| Other | 0.81 (0.16-4.04) | 0.80 (0.16-4.01) |
| Unknown | 0.14 (0.02-1.13) | 0.14 (0.02-1.16) |
| NDI (vs Q1, least deprived) | ||
| Q2 | 0.95 (0.89-1.01) | 0.94 (0.89-1.00) |
| Q3 | 0.92 (0.86-0.98) | 0.92 (0.86-0.97) |
| Q4 | 0.78 (0.73-0.83) | 0.78 (0.73-0.84) |
| Unknown Q | 0.80 (0.41-1.57) | 0.80 (0.41-1.56) |
| CHA2DS2-VASc score (vs 2)e | ||
| 3 | 0.96 (0.91-1.02) | 0.96 (0.91-1.02) |
| 4 | 0.96 (0.89-1.03) | 0.96 (0.89-1.03) |
| ≥5 | 0.82 (0.75-0.90) | 0.82 (0.75-0.90) |
Abbreviations: DOAC, direct oral anticoagulant; NA, not applicable; NDI, neighborhood deprivation index; Q, quartile.
Univariate logistic regression.
Multivariable logistic regression, adjusted for patient age, sex, race and ethnicity, insurance type, and neighborhood deprivation index quartile, and CHA2DS2-VASc score.
Individuals who self-identified with more than 1 race or ethnicity.
Self-reported other.
CHA2DS2-VASc score represents congestive heart failure; hypertension; age ≥75 years; diabetes; prior stroke, transient ischemic attack, or thromboembolism; vascular disease; age 65-74 years; and sex category.
Discussion
In this retrospective cohort study of 59 340 DOAC therapy–eligible patients with NVAF, patient-physician race and ethnicity concordance was associated with increased odds of DOAC therapy initiation, whereas race, ethnicity, and language concordance were associated with increased odds of DOAC therapy persistence. Taken together, these results suggest that race, ethnicity, and language concordance may play important roles in DOAC-related care processes although their relative influence may vary across clinical contexts.
Prior studies have evaluated associations between patient-physician race and language concordance and improved medication adherence and disease management across a variety of chronic disease settings, including diabetes, hypertension, and hyperlipidemia.15,16,17,18 The present study extends the existing concordance literature to anticoagulation care by examining DOAC therapy initiation and persistence among patients with NVAF. Beyond this extension, examining this question is particularly relevant given the well-established racial and ethnic disparities in AF management and outcomes.
Medication initiation is a complex process in which patients and physicians collaboratively weigh potential benefits and risks, integrate patient values and preferences, and align treatment decisions with clinical evidence and individual circumstances. Several barriers to anticoagulation initiation in the setting of AF have been identified, including patients’ poor understanding of the condition, fear of adverse effects, and concerns related to cost.27,28 The patient-physician relationship may also influence this process, as poor rapport and insufficient shared decision-making can lead to hesitancy or refusal to initiate therapy.28,29 The current study found that patient-physician race and ethnicity concordance was associated with increased odds of DOAC therapy initiation, suggesting that race and ethnicity concordance may help foster the trust and alignment necessary for patients to feel comfortable proceeding with anticoagulation therapy.
The study period (January 2012 through December 2022) spans a time of substantial change in anticoagulation practice. In the earlier years, DOACs were considered a relatively novel therapy, but they became a predominant, guideline-recommended strategy for NVAF in the later years. Accordingly, DOAC therapy initiation in the earlier years may reflect treatment selection over warfarin, whereas in the later years it more closely represents uptake of preferred therapy. By the end of the study period, DOACs had become the dominant (over 95%) anticoagulation strategy following NVAF diagnosis; thus, the observed associations should be interpreted primarily as differences in receipt of guideline-recommended therapy. In this context, patient-physician concordance may influence not only treatment initiation but also the likelihood of receiving guideline-concordant care.
Compared with initiation of DOAC therapy, DOAC therapy persistence reflects treatment continuity and is affected by ongoing factors such as evolving patient motivation and continued trust in the patient-physician relationship.30,31 The present study observed that both patient-physician race and ethnicity and language concordance were independently associated with increased odds of DOAC therapy persistence. This finding aligns with broader evidence that patient-physician concordance can strengthen different dimensions of the therapeutic relationship. In addition to the potential benefits conferred by race and ethnicity concordance, patient-physician language concordance supports clear, ongoing communication about diagnoses and treatment plans and is associated with improved patient understanding, satisfaction, and trust.32,33 That both forms of concordance are associated with DOAC therapy persistence underscores the notion that sustained engagement relies on relational trust and effective communication over time.
Subgroup analysis in Hispanic and non-Hispanic Black populations conducted in the current study revealed nuanced findings. Among Hispanic patients, race and ethnicity concordance was associated with increased DOAC therapy initiation; in contrast, language concordance was associated with increased DOAC therapy persistence. One possible explanation for this finding is the considerable heterogeneity that exists within Hispanic populations regarding genetic ancestry, country of origin, and culture. For example, Hispanic populations are admixed, with varying proportions of Native American, European, and African ancestry34,35,36; similarly, Hispanic subgroups differ in their immigration histories, with proportions of foreign-born individuals varying widely among subgroups.37 Consequently, race and ethnicity concordance may be a less consistent marker of enduring interpersonal alignment in this group. Language concordance, however, may be a more salient and relevant determinant of DOAC therapy persistence by facilitating improved communication regarding adverse effects and evolving patient concerns, thereby reinforcing understanding and sustained engagement with therapy. In contrast, we did not observe an association between race and ethnicity or language concordance and DOAC therapy initiation or persistence among non-Hispanic Black adults. This finding may partially reflect limited statistical power due to the relatively small number of non-Hispanic Black physicians and language-discordant encounters within this subgroup. However, it is also possible that race and ethnicity concordance serves as an imperfect proxy for the complex interpersonal and structural factors that influence care. Prior studies have demonstrated that non-Hispanic Black patients disproportionately experience health care–related mistrust, perceived discrimination, and longstanding structural inequities,38,39,40 all of which may influence treatment decisions and longitudinal engagement with care independent of physician race or ethnicity. Consequently, shared racial or ethnic identity alone may be insufficient to overcome the broader factors contributing to the disparities in anticoagulation management observed in this population.
Collectively, the findings of this study support continued efforts to improve culturally responsive care and to diversify the physician workforce while also acknowledging that concordance alone may not be sufficient to fully address the broader drivers of inequities in anticoagulation management. Potential future strategies include strengthening mentorship and pathway programs, supporting the recruitment and retention of physicians from historically underrepresented communities, and expanding multilingual communication training.
Limitations
This study has several limitations. As with any retrospective cohort study, this study is susceptible to biases related to the quality of the gathered data. Unmeasured factors, including bleeding risk and potential patient selection of clinicians, may contribute to residual confounding. Furthermore, race, ethnicity, and language concordance were based on administrative data, which may not fully capture the nuances of cultural or linguistic alignment. DOAC therapy initiation was defined as a new prescription order within 60 days following NVAF diagnosis, and patients were not required to be naive to anticoagulation therapy. Therefore, some cases may reflect continuation of prior therapy rather than true first-time initiation. Other clinician characteristics (eg, specialty, age, and years in practice) may influence DOAC prescribing and persistence patterns; however, these data were not available and may contribute to residual confounding. Moreover, DOAC therapy persistence may reflect longitudinal care and may be influenced by interactions with multiple clinicians beyond the initial diagnosing physician. In this study, concordance was defined based on the diagnosing physician and therefore may not capture the full complexity of the ongoing care relationship. Finally, given that the study cohort consisted only of individuals enrolled in KPSC, the generalizability of these results to populations outside of this health care system may be limited.
Conclusions
In this retrospective cohort study of adults with NVAF, patient-physician race and ethnicity concordance was associated with increased odds of DOAC therapy initiation, whereas patient-physician race, ethnicity, and language concordance were associated with increased odds of DOAC therapy persistence. Collectively, these findings highlight the potential influence of patient-physician concordance in improving equity in anticoagulation care; importantly, they also underscore the need for both increased diversity within the physician workforce and further research to clarify how relational and communication-based factors may support equitable DOAC use across diverse populations.
eTable 1. Definitions of Study Variables Used in the Analysis
eTable 2. Patient-Physician Concordance and DOAC Initiation in the Hispanic Subgroup
eTable 3. Patient-Physician Concordance and DOAC Initiation in the NH Black Subgroup
eTable 4. Characteristics of Study Population for DOAC Persistence by Racial/Ethnic Group
eTable 5. Patient-Physician Concordance and DOAC Persistence in the Hispanic Subgroup
eTable 6. Patient-Physician Concordance and DOAC Persistence in the NH Black Subgroup
Data Sharing Statement
References
- 1.Martins RP, Galand V, Colette E, et al. Defining nonvalvular atrial fibrillation: a quest for clarification. Am Heart J. 2016;178:161-167. doi: 10.1016/j.ahj.2016.05.014 [DOI] [PubMed] [Google Scholar]
- 2.Wolf PA, Dawber TR, Thomas HE Jr, Kannel WB. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: the Framingham study. Neurology. 1978;28(10):973-977. doi: 10.1212/WNL.28.10.973 [DOI] [PubMed] [Google Scholar]
- 3.Lip GYH, Murphy RR, Sahiar F, et al. Risk levels and adverse clinical outcomes among patients with nonvalvular atrial fibrillation receiving oral anticoagulants. JAMA Netw Open. 2022;5(8):e2229333. doi: 10.1001/jamanetworkopen.2022.29333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kumbhani DJ, Cannon CP, Beavers CJ, et al. 2020 ACC expert consensus decision pathway for anticoagulant and antiplatelet therapy in patients with atrial fibrillation or venous thromboembolism undergoing percutaneous coronary intervention or with atherosclerotic cardiovascular disease: a report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol. 2021;77(5):629-658. doi: 10.1016/j.jacc.2020.09.011 [DOI] [PubMed] [Google Scholar]
- 5.Joglar JA, Chung MK, Armbruster AL, et al. ; Peer Review Committee Members . 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2024;149(1):e1-e156. doi: 10.1161/CIR.0000000000001193 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Reynolds KR, Khosrow-Khavar F, Dave CV. Racial and ethnic disparities in initiation of direct oral anticoagulants among Medicare beneficiaries. JAMA Netw Open. 2024;7(5):e249465. doi: 10.1001/jamanetworkopen.2024.9465 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.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]
- 8.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]
- 9.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]
- 10.An J, Bider Z, Luong TQ, et al. Long-term medication adherence trajectories to direct oral anticoagulants and clinical outcomes in patients with atrial fibrillation. J Am Heart Assoc. 2021;10(21):e021601. doi: 10.1161/JAHA.121.021601 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Joy M, Williams J, Emanuel S, et al. Trends in direct oral anticoagulant (DOAC) prescribing in English primary care (2014-2019). Heart. 2023;109(3):195-201. doi: 10.1136/heartjnl-2022-321377 [DOI] [PubMed] [Google Scholar]
- 12.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]
- 13.Essien UR, Tang JJ, Kim N, et al. Association of race and ethnicity with stroke and mortality outcomes in atrial fibrillation. JACC Adv. 2025;4(7):101860. doi: 10.1016/j.jacadv.2025.101860 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Daly DJ, Essien UR, Del Carmen MG, et al. Race, ethnicity, sex, and socioeconomic disparities in anticoagulation for atrial fibrillation: a narrative review of contemporary literature. J Natl Med Assoc. 2023;115(3):290-297. doi: 10.1016/j.jnma.2023.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Traylor AH, Schmittdiel JA, Uratsu CS, Mangione CM, Subramanian U. Adherence to cardiovascular disease medications: does patient-provider race/ethnicity and language concordance matter? J Gen Intern Med. 2010;25(11):1172-1177. doi: 10.1007/s11606-010-1424-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Parker MM, Fernández A, Moffet HH, Grant RW, Torreblanca A, Karter AJ. Association of patient-physician language concordance and glycemic control for limited-English proficiency Latinos with type 2 diabetes. JAMA Intern Med. 2017;177(3):380-387. doi: 10.1001/jamainternmed.2016.8648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Cano-Ibáñez N, Zolfaghari Y, Amezcua-Prieto C, Khan KS. Physician-patient language discordance and poor health outcomes: a systematic scoping review. Front Public Health. 2021;9:629041. doi: 10.3389/fpubh.2021.629041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ma A, Campbell J, Sanchez A, Sumner S, Ma M. Racial disparities in medication adherence and the patient-provider relationship: does racial/ethnic concordance matter? Health Serv Res. 2026;61(1):e70040. doi: 10.1111/1475-6773.70040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lubitz SA, Khurshid S, Weng LC, et al. Predictors of oral anticoagulant non-prescription in patients with atrial fibrillation and elevated stroke risk. Am Heart J. 2018;200:24-31. doi: 10.1016/j.ahj.2018.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Manning E, Burns K, Laurie M, Patten L, Ho M, Sandhu A. Factors associated with oral anticoagulant prescription status among patients with a new diagnosis of atrial fibrillation. Clin Cardiol. 2023;46(8):937-941. doi: 10.1002/clc.24077 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lip GYH, Nieuwlaat R, Pisters R, Lane DA, Crijns HJGM. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on Atrial Fibrillation. Chest. 2010;137(2):263-272. doi: 10.1378/chest.09-1584 [DOI] [PubMed] [Google Scholar]
- 22.Bezabhe WM, Bereznicki LR, Radford J, et al. Factors influencing oral anticoagulant use in patients newly diagnosed with atrial fibrillation. Eur J Clin Invest. 2021;51(5):e13457. doi: 10.1111/eci.13457 [DOI] [PubMed] [Google Scholar]
- 23.Mulholland RJ, Ciminata G, Quinn TJ, Pollock KG, Lister S, Geue C. Delay and pay? healthcare costs associated with late oral anti-coagulant prescribing in people with atrial fibrillation. Pharmacoeconomics. 2024;42(11):1241-1253. doi: 10.1007/s40273-024-01419-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Coleman CI, Tangirala M, Evers T. Treatment persistence and discontinuation with rivaroxaban, dabigatran, and warfarin for stroke prevention in patients with non-valvular atrial fibrillation in the United States. PLoS One. 2016;11(6):e0157769. doi: 10.1371/journal.pone.0157769 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ozaki AF, Choi AS, Le QT, et al. Real-world adherence and persistence to direct oral anticoagulants in patients with atrial fibrillation: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes. 2020;13(3):e005969. doi: 10.1161/CIRCOUTCOMES.119.005969 [DOI] [PubMed] [Google Scholar]
- 26.Deitelzweig S, Di Fusco M, Kang A, et al. Real-world persistence to direct oral anticoagulants in patients with atrial fibrillation: a systematic review and network meta-analysis. Curr Med Res Opin. 2021;37(6):891-902. doi: 10.1080/03007995.2021.1897555 [DOI] [PubMed] [Google Scholar]
- 27.Ferguson C, Inglis SC, Newton PJ, Middleton S, Macdonald PS, Davidson PM. Barriers and enablers to adherence to anticoagulation in heart failure with atrial fibrillation: patient and provider perspectives. J Clin Nurs. 2017;26(23-24):4325-4334. doi: 10.1111/jocn.13759 [DOI] [PubMed] [Google Scholar]
- 28.Vaanholt MCW, Weernink MGM, von Birgelen C, Groothuis-Oudshoorn CGM, IJzerman MJ, van Til JA. Perceived advantages and disadvantages of oral anticoagulants, and the trade-offs patients make in choosing anticoagulant therapy and adhering to their drug regimen. Patient Educ Couns. 2018;101(11):1982-1989. doi: 10.1016/j.pec.2018.06.019 [DOI] [PubMed] [Google Scholar]
- 29.Ng VWS, Siu CW, Chiu PKC, et al. Understanding the barriers to using oral anticoagulants among long-term aspirin users with atrial fibrillation - a qualitative study. BMC Health Serv Res. 2020;20(1):1084. doi: 10.1186/s12913-020-05947-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wroth TH, Pathman DE. Primary medication adherence in a rural population: the role of the patient-physician relationship and satisfaction with care. J Am Board Fam Med. 2006;19(5):478-486. doi: 10.3122/jabfm.19.5.478 [DOI] [PubMed] [Google Scholar]
- 31.Zeber JE, Manias E, Williams AF, et al. ; ISPOR Medication Adherence Good Research Practices Working Group . A systematic literature review of psychosocial and behavioral factors associated with initial medication adherence: a report of the ISPOR medication adherence & persistence special interest group. Value Health. 2013;16(5):891-900. doi: 10.1016/j.jval.2013.04.014 [DOI] [PubMed] [Google Scholar]
- 32.Diamond L, Izquierdo K, Canfield D, Matsoukas K, Gany F. A systematic review of the impact of patient-physician non-English language concordance on quality of care and outcomes. J Gen Intern Med. 2019;34(8):1591-1606. doi: 10.1007/s11606-019-04847-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lor M, Martinez GA. Scoping review: definitions and outcomes of patient-provider language concordance in healthcare. Patient Educ Couns. 2020;103(10):1883-1901. doi: 10.1016/j.pec.2020.05.025 [DOI] [PubMed] [Google Scholar]
- 34.González Burchard E, Borrell LN, Choudhry S, et al. Latino populations: a unique opportunity for the study of race, genetics, and social environment in epidemiological research. Am J Public Health. 2005;95(12):2161-2168. doi: 10.2105/AJPH.2005.068668 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Bryc K, Velez C, Karafet T, et al. Colloquium paper: genome-wide patterns of population structure and admixture among Hispanic/Latino populations. Proc Natl Acad Sci U S A. 2010;107(suppl 2):8954-8961. doi: 10.1073/pnas.0914618107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Conomos MP, Laurie CA, Stilp AM, et al. Genetic diversity and association studies in US Hispanic/Latino populations: applications in the Hispanic Community Health Study/Study of Latinos. Am J Hum Genet. 2016;98(1):165-184. doi: 10.1016/j.ajhg.2015.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Carvajal DN, Anaya YB, McLean I, et al. We are not all the same: implications of heterogeneity among Latiné/e/x/o/a, Hispanic, and Spanish origin people. Ann Fam Med. 2024;22(3):254-258. doi: 10.1370/afm.3103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Bazargan M, Cobb S, Assari S. Discrimination and medical mistrust in a racially and ethnically diverse sample of California adults. Ann Fam Med. 2021;19(1):4-15. doi: 10.1370/afm.2632 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Cueva KL, Marshall AR, Snyder CR, Young BA, Brown CE. Medical mistrust among black patients with serious illness: a mixed methods study. J Gen Intern Med. 2024;39(14):2747-2754. doi: 10.1007/s11606-024-08997-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Sabater Minarim D, Morgan KM, Buckley K, et al. Racial and ethnic disparities in perceived healthcare discrimination and health outcomes. J Gen Intern Med. 2025;40(11):2548-2557. doi: 10.1007/s11606-025-09627-y [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
eTable 1. Definitions of Study Variables Used in the Analysis
eTable 2. Patient-Physician Concordance and DOAC Initiation in the Hispanic Subgroup
eTable 3. Patient-Physician Concordance and DOAC Initiation in the NH Black Subgroup
eTable 4. Characteristics of Study Population for DOAC Persistence by Racial/Ethnic Group
eTable 5. Patient-Physician Concordance and DOAC Persistence in the Hispanic Subgroup
eTable 6. Patient-Physician Concordance and DOAC Persistence in the NH Black Subgroup
Data Sharing Statement
