Health Care Utilization Among Post-Stroke Women: Associations with Perceived Racism

Abstract

Purpose:

Post-stroke health care is crucial for recovery and preventing complications. This study explored the association between perceived racism and health care utilization among post-stroke women, highlighting the disparate utilization patterns of marginalized racial/ethnic women.

Methods:

The 2022 Centers for Disease Control and Prevention’s Behavioral Risk Factor Surveillance System Survey was used to characterize health care utilization among women with stroke diagnoses across different racial groups: White, Black, Hispanic, and Asian/Asian, Islander, Alaska Native/Other/Multiracial. Logistic regression analysis characterized the association between three types of health care utilizations—having a health insurance plan, a health care provider, and receiving a routine checkup—and perceived race-based treatment in the medical sector.

Results:

The sample included 10,184 post-stroke women: 75.01% White, 12.42% Black, 6.77% Hispanic, and 5.80% Asian/AIAN/Other/multiracial. Only 2.95% of White women reported worse treatment, compared with 10.74% of Black, 13.19% of Asian/AIAN/Other/multiracial, and 8.85% of Hispanic women. Regression analysis revealed that Hispanic women (odds ratio [OR] = 0.61, confidence interval [CI] = 0.17, 0.84) were less likely to have a health plan, though those receiving similar or better treatment had higher odds (OR = 1.55, CI = 1.62, 3.90). Black women were less likely to have a routine checkup (OR = 0.51, CI = 0.14, 0.77), but those receiving similar or better treatment had higher odds (OR = 2.72, CI = 1.64, 11.63). Women from other racial/ethnic groups, Black, and Hispanic women had lower odds of having a personal provider (ORs ranging from 0.17 to 0.69), though those receiving similar or better treatment had higher odds (ORs ranging from 1.79 to 3.39).

Conclusions:

Perceived differences in treatment in the medical sector among marginalized post-stroke women were significantly associated with lower health care utilization.

Background

Stroke is a leading cause of disability and mortality worldwide, with 795,000 cases annually in the United States.¹ Non-Hispanic (NH) Black adults face nearly twice the risk of stroke compared with their White counterparts.^2
–4 Black individuals also experience worse outcomes during recovery.^5
–7

Women experience 53% of all strokes.¹ Yoon and Bushnell reviewed that women face a higher lifetime risk of stroke compared with men, with worse outcomes and different patterns of risk factors.^8,9 Women’s longer lifespans predispose them to strokes at older ages with more severe outcomes.^9

–12 Consequently, women suffer more stroke-related mortality, illness, and long-term disability compared with men.^9,10,12,13

Factors contributing to this higher prevalence among women include longer lifespan, older age, and comorbidities such as hypertension, heart disease, and diabetes mellitus.^9,11,14,15 Conditions such as antiphospholipid antibodies and pregnancy-related factors also influence stroke likelihood.¹⁶

Race and sex intersect to influence stroke outcomes.^9,11,12 Black women have a higher predisposition to comorbidities and severity of stroke.^7,9,17 Despite the high prevalence among women and disparities in care, researchers have limitedly studied these complex factors among diverse women. Traditional factors such as socioeconomic status, geography, and health care access do not fully explain the differences in health care utilization and stroke outcomes.^7,10,14,15 The knowledge gap risks exacerbating health disparities, particularly for women from diverse racial and socioeconomic backgrounds.^7,9,11 This gap can lead to unequal care, poorer health outcomes, and increased mortality.^11,14 Disparities significantly impact the quality of life for women who experience stroke, leading to physical disabilities, cognitive impairments, and exacerbation by inadequate health care support.^10,18,19

Women from minoritized backgrounds face unique challenges in stroke recovery due to cultural norms, socioeconomic disparities, and health care access issues.^7,10,19 Societal perceptions and stereotypes affecting Black and Hispanic women further impact their experiences as stroke survivors.^20

–23 Researchers have observed disparities in response to stroke symptoms and treatment, in addition to these culturally based challenges.^5,7,24 Ikeme⁵ found that White patients used emergency services more frequently and arrived within 3 hours of symptom onset more often. These disparities in immediate response significantly impact stroke outcomes. Evidence suggests that stroke survivors from minority groups often receive subpar care due to gender and race, affecting their willingness to seek emergency treatment.^5,25,26 Addressing biases at the intersection of race and gender in health care is essential, as race also influences communication quality, treatment, and support systems.^14,25,27

Previous research has explored the intersection of race and sex in stroke outcomes.^28,29 Hosman²⁹ emphasizes the need for further research on interactions between sex and race to ensure equity in outcomes. A recent focus is on how external factors, such as racism, influence stroke management and outcomes. Racism in health care, defined as “unequal power relations based on beliefs about ‘race’ and related systems,” contributes to inequitable patterns of access, utilization, and quality while diminishing patient satisfaction and trust.³⁰ The perception of racism impacts care access and perceptions of health care systems.³¹ Understanding outcomes with racial disparities must include how perceived racism affects outcomes.³² The Centers for Disease Control and Prevention’s (CDC) Behavioral Risk Factor Surveillance System (BRFSS) provides data to investigate health care utilization among post-stroke women. This study aims to explore the relationship between perceived racism and post-stroke health care utilization, offering insights into factors contributing to long-term stroke outcomes and addressing disparities among women.

Methods

Data

This study utilized the 2022 BRFSS—a state-based telephone survey of more than 400,000 respondents representing the noninstitutionalized, adult population with a landline or cellular phone.³³ The BRFSS contains information on individual health-related behaviors, health conditions, and use of health services.^34,35 The CDC analyzes BRFSS data to provide localized health information that can help public health practitioners to (1) identify local emerging health problems, (2) identify populations at risk, (3) evaluate local responses, and (4) efficiently allocate resources to specific needs.

To account for the use of cellular and landline telephone data collection methods and the different sociodemographic composition of individual states, data weights are calculated through an iterative proportional fitting process known as “raking.” BRFSS weighting methodology accounts for survey design, demographic variation, the probability of selection, and nonresponse bias and non-coverage errors and helps make sample data more representative of the population from which the data were collected. If a state had demographic variation between geographic regions or county-level population densities, then additional margins were included in the raking to produce the final weight assigned to each respondent.

All information contained in the BRFSS is based on self-report. Individuals respond to survey questions concerning health conditions, race/ethnicity, health insurance, health care utilization, health providers, and individual/household characteristics. Individual/household characteristics included self-reported age (truncated at 80), sex (male, female), marital status (married, non-married), education level (less than high school, high school or equivalent, college graduate and above), residing in a rural or urban county, and household income (<$25,000, $25,000 to <$50,000, ≥$50,000). Individual respondents classified their race and ethnicity into one of eight categories including NH White, NH Black, NH American Indian or Alaskan Native only, NH Asian only, NH Native Hawaiian or other Pacific Islander only, NH Other race only, NH Multiracial, and Hispanic. Due to sample size limitations, NH American Indian or Alaskan Native only, NH Asian only, NH Native Hawaiian or other Pacific Islander only, NH Other race only, and NH Multiracial were combined into a single category resulting in four primary groups.

Each respondent was asked, “Has a doctor, nurse, or other health professional ever told you that you had a stroke?” and they selected either yes or no. They were also asked a series of questions regarding their health insurance, health care receipt, and health care provider. The three survey items listed below with their answer choices were selected as indicators of individual health care utilization. 1.

Do you currently have some form of health insurance? (Yes, No)

About how long has it been since you last visited a doctor for a routine checkup? (Within the last year, >1 years ago/never)

Do you have one person (or a group of doctors) whom you think of as your personal health care provider? (Yes, No)

While only utilized by a handful of states in previous years, 28 states and the District of Columbia included the Reactions to Race (R2R) Module—a series of six questions and a prologue designed to assess socially assigned race, race consciousness, perceptions of treatment by race/ethnicity in health care and work settings, and experiences of emotional and physical symptoms resulting from race-based treatment—in 2022.³⁶ This analysis included one item from the R2R that intended to measure the impact of “socially assigned race” or how the perception of race by others may be a determinant of unequal treatment, including treatment when seeking health care.³⁷ To facilitate interpretation, response categories were collapsed into two groups.

1. Within the past 12 months, when seeking health care, do you feel your experiences were worse than, the same as, or better than people of other races? (1-Worse than other races, 2-The same as other races/Better than other races/Worse than some races, Better than others/Only encountered people of the same race)

Sample

The analysis sample was limited to respondents between the ages of 18 and 80 who identified as female, reported having been diagnosed with stroke, and completed the R2R module in 2022. These inclusion criteria resulted in a sample of 10,184 post-stroke women.

Analysis

First, descriptive statistics were calculated for demographic, perception of treatment by race in health care, and health care utilization variables for all four racial/ethnic groups. Chi-squared and t-tests assessed between-group differences in characteristics, outcomes, and perceived treatment. Second, logistic regression estimated the relationship between the three health care utilization parameters, race/ethnicity, and perceived racial treatment accounting for heterogeneity in individual/household characteristics. Interaction terms between race/ethnicity and indicators of perceived racial treatment were also included to determine whether utilization differed within groups. Odds ratios were also calculated to quantify the magnitude of the associations. All analyses incorporated individual-level weights as well as stratification and sampling parameters to account for the complex sampling framework of BRFSS. All analyses were conducted in SAS 9.4 using selected survey analytic commands that account for the clustering and stratification of the data sampling approach.

Sensitivity analyses

To test the robustness of findings, two additional analyses were performed. First, since unobserved heterogeneity between individuals with and without a health plan may have confounded results, the sample was restricted to only respondents who reported having a health plan. Models testing the association between perceived racism and having a doctor and having a recent checkup were re-estimated on the restricted data. Second, since racial and ethnic minorities reported lower rates of health care utilization, but higher rates of perceived racism, generalized structural equation models (GSEM) tested the mediating effect of perceived racism on the association between race/ethnicity and the likelihood of having a provider, a recent checkup, and a health plan. GSEM first fits an initial path model using an indicator of race/ethnicity, the potential mediator (perceived racism), and the outcome (provider, checkup, or health plan) to assess the crude association. After controlling for other potentially confounding factors, full mediation analysis models were fitted for all three health care utilization outcomes. The mediation was hypothesized to be partial mediation in which an exposure variable has both direct and indirect effects on the outcome where the direct effect is not mediated, but an indirect effect is transmitted through the mediator variable. In this analysis, the total effect was estimated and then the direct and indirect effects were determined using the product-of-coefficients test, and the portion explained by the mediator was estimated.

Results

Univariate statistics for all individual/household characteristics, perceived racial treatment, and health care utilization variables for the four racial/ethnic groups are reported in Table 1 along with between group chi-square and t-test results. The sample consisted of 10,184 post-stroke women women—75% White, 12% Black, 7% Other races, and 6% Hispanic—with mean ages ranging between 66.65 (standard deviation [SD] = 12.16) for White and 58.18 (SD = 14.43) for Hispanic women (t = 22.31, p < 0.0001). Nearly 90% of White women lived in urban areas compared with 95% of Black, 90% of Other, and 95% of Hispanic women (χ² = 63.37, p = 0.0003). Most women were not married (White 62.00%; Black 83.86%; 70.32% Other; Hispanic 60.32%; χ² = 29.69, p < 0.0001). Roughly one-third of White and Other women were college graduates compared with 20.12% and 22.10% of Black and Hispanic women (χ² = 45.70, p < 0.0001). Nearly 35% of White women lived in households earning $50,000 and above compared with 17.98%, 25.85%, and 21.47% of Black, Other, and Hispanic women, respectively (χ² = 27.85, p < 0.0001).

Table 1.

Sample Characteristics and Test of Between-Group Differences

	White(N = 7,639, 75.01%)		Black(N = 1,264, 12.42%)		Asian/AIAN/Other(N = 693, 6.80%)		Hispanic(N = 588, 5.77%)		Group difference
	Mean	Std	Mean	Std	Mean	Std	Mean	Std	t	p-Value
Age	66.54	12.16	61.51	12.60	61.47	14.12	58.18	14.43	−22.31	<0.0001

	N	PCT	N	PCT	N	PCT	N	PCT	χ²	p-Value
Have a health plan	7,171	97.45	1,148	97.64	641	96.65	501	89.03	4.44	0.68
Health checkup within last 1 year	6,917	91.67	1,178	91.18	594	89.12	514	91.04	19.07	<0.0001
Have personal health care provider	7,318	96.57	1,190	95.05	648	94.68	512	85.99	92.60	<0.0001
Treated worse than other races when seeking health care	56	2.95	72	10.74	23	13.19	11	8.85	33.46	<0.0001
Treated better, the same, neither better nor worse than other races when seeking health care	2,487	97.05	419	89.26	138	86.81	105	91.16
<$25,000	1,864	34.14	485	56.30	246	47.49	212	50.10	27.85	<0.0001
$25,000 to <$50,000	1,903	31.81	284	25.72	164	26.66	131	28.43
$50,000+	1,848	34.06	192	17.98	128	25.85	95	21.47
Did not graduate high school	593	8.41	193	15.50	79	9.03	150	29.78	45.70	<0.0001
Graduated high school/Attended college or technical school	4,893	63.68	802	64.38	402	59.56	307	48.12
Graduated from college or technical school	2,122	27.91	267	20.12	208	31.40	128	22.10
Married	2,866	38.00	239	16.14	205	29.68	225	39.68	29.69	<0.0001
Not married	4,724	62.00	1,020	83.86	477	70.32	359	60.32
Urban counties	6,359	89.08	1,131	94.51	504	89.90	471	94.85	63.37	0.0003
Rural counties	1,266	10.92	109	5.49	153	10.10	29	5.15

Perceived racial treatment differed significantly between racial/ethnic groups with 10.74% of Black, 13.19% of Other, and 8.85% of Hispanic women reporting worse treatment than other races when seeking health care compared with only 2.95% of White women (χ² = 33.46, p < 0.0001). There was no significant difference between the portion of each group that reported having a health plan (White 97.45%; Black 97.64%; Other 96.65%; Hispanic 89.03; χ² = 4.44, p = 0.68). However, the percent of those having a health plan (White 91.67%; Black 91.18%; Other 89.12%; Hispanic 91.04%; χ² = 19.07, p < 0.0001) and a personal health care provider (White 96.57%; Black 95.05%; Other 94.68%; Hispanic 85.99%; χ² = 92.60 p < 0.0001) was significantly different.

Results from the regression model estimating the association between having a health plan, perceived racial treatment, and race/ethnicity are listed in Table 2. Post-stroke women who did not graduate from high school (odds ratio [OR] = 0.02, confidence interval [CI] = 0.00, 0.15) as well as those living in household earning below $25,000 (OR = 0.07, CI = 0.01, 0.42) were less likely to have a health plan than their counterparts. Asian/AIAN/Other/multiracial women (OR = 6.11, CI = 1.47, 9.94) as well as those in nonurban/rural counties (OR = 5.47, CI = 2.14, 11.92) were more likely than the reference group to have a health plan. Compared with White women, Hispanic women were less likely to have a health plan (OR = 0.61, CI = 0.17, 0.84). However, Hispanic women who perceived being treated similar or better than women of other races were more likely to have a health plan compared with those Hispanic women who were treated worse than women of other races.

Table 2.

Likelihood of Having a Health Plan Among Women with Stroke

	Std err		t	Pr	OR	95% CI
Intercept	6.80	3.87	1.76	0.08
Age	−0.39	0.96	−0.41	0.68	0.67	0.10	4.40
Did not graduate from high school	−2.01	0.47	−4.31	<0.0001	0.02	0.00	0.15
High school graduate/Some college/Technical degree	0.00	0.38	0.01	0.99	0.14	0.02	0.90
Income $25,000 to <$50,000	−0.49	0.49	−1.01	0.31	0.13	0.02	1.02
Income <$25,000	−1.10	0.38	−2.93	0.00	0.07	0.01	0.42
Married	−0.59	0.42	−1.40	0.16	0.31	0.06	1.60
Lives in rural county	1.37	0.50	2.71	0.01	5.47	2.14	11.92
Asian/AIAN/Other	4.90	0.54	9.05	<0.0001	6.11	1.47	9.94
Black	−0.17	0.41	−0.41	0.68	0.72	0.10	5.19
Hispanic	−2.01	0.28	−7.18	<0.0001	0.61	0.17	0.84
Treated better, the same, neither better nor worse as other races when seeking health care	−0.73	0.28	−1.56	0.10	0.94	0.85	9.94
Asian/AIAN/Other*Treated better, the same, neither better nor worse as other races when seeking health care	−3.65	0.56	−6.52	<0.0001	0.79	0.18	0.83
Black*Treated better, the same, neither better nor worse as other races when seeking health care	0.57	0.41	1.38	0.17	2.62	1.38	17.89
Hispanic*Treated better, the same, neither better nor worse as other races when seeking health care	0.95	0.36	2.63	0.01	1.55	1.62	3.90

Reference categories: Education (College graduate and above), Income (≥$50,000), Marital status (not married), Urbanicity (Lives in urban county), Race/ethnicity (White, non-Hispanic), Racial treatment (Treated worse than other races).

Dependent variable: Has a health plan.

Estimates indicate the relative likelihood of having a health plan.

Regression procedures used BRFSS stratification and cluster parameters to account for complex survey design.

Estimates weighted to reflect a representative population.

Boldface indicates significance at the 95% confidence level.

BRFSS, Behavioral Risk Factor Surveillance System; CI, confidence interval; OR, odds ratio; PR, probability.

Results from the regression model estimating the association between having a recent checkup, perceived racial treatment, and race/ethnicity are listed in Table 3. Age (OR = 4.23, CI = 1.84, 9.70), rurality (OR = 2.26, CI = 1.08, 4.74), and income (OR = 2.29, CI = 1.19, 4.39) were significantly related to the likelihood of having a recent checkup. Compared with White women, those of other races (OR = 1.15, CI = 1.14, 3.01) were more likely to have had a checkup, while Black women were less likely (OR = 0.51, CI = 0.14, 0.77). However, perceived racial treatment was significant among post-stroke Black women with those who perceived their treatment as like or better than other races being more likely (OR = 2.72, CI = 1.64, 11.63) than those who perceived worse treatment to have had a recent checkup.

Table 3.

Likelihood of Having a Routine Checkup in the Last Year

	Std err		t	Pr\|	OR	95% CI
Intercept	−3.33	1.69	−1.97	0.05
Age	1.44	0.42	3.40	0.00	4.23	1.84	9.70
Did not graduate from high school	−0.16	0.26	−0.61	0.54	0.76	0.32	1.80
High school graduate/Some college/Technical degree	0.04	0.18	0.24	0.81	0.93	0.52	1.67
Income $25,000 to <$50,000	0.40	0.18	2.18	0.03	2.29	1.19	4.39
Income <$25,000	0.04	0.19	0.20	0.84	1.60	0.82	3.14
Married	−0.07	0.12	−0.54	0.59	0.88	0.54	1.43
Lives in rural county	0.41	0.19	2.16	0.03	2.26	1.08	4.74
Asian/AIAN/Other	0.86	0.40	2.15	0.03	1.15	1.14	3.01
Black	−0.11	0.34	−2.34	0.05	0.51	0.14	0.77
Hispanic	−0.59	0.43	−1.37	0.17	0.39	0.06	2.43
Treated better, the same, neither better nor worse as other races when seeking health care	0.09	0.21	0.42	0.67	1.58	0.48	5.17
Asian/AIAN/Other*Treated better, the same, neither better nor worse as other races when seeking health care	−0.74	0.40	−1.86	0.06	0.27	0.04	1.79
Black*Treated better, the same, neither better nor worse as other races when seeking health care	0.41	0.34	2.22	0.02	2.72	1.64	11.63
Hispanic*Treated better, the same, neither better nor worse as other races when seeking health care	0.19	0.42	0.47	0.64	1.76	0.24	12.95

Dependent variable: Visited the doctor for a routine checkup within the last year.

Estimates indicate the relative likelihood of having a checkup in the last year.

Regression procedures used BRFSS stratification and cluster parameters to account for complex survey design.

Estimates weighted to reflect a representative population.

Boldface indicates significance at the 95% confidence level.

BRFSS, Behavioral Risk Factor Surveillance System; CI, confidence interval; OR, odds ratio; Pr, Probability; Std err, standard error.

Results from the regression model estimating the association between having a personal health care provider, perceived racial treatment, and race/ethnicity are listed in Table 4. Results for age (OR = 19.25, CI = 4.42, 83.95), education (OR = 0.18, CI = 0.04, 0.81), and rurality (OR = 17.66, CI = 1.96, 159.06) were like those observed among those utilization measures discussed previously. On average, Other (OR = 0.62, CI = 0.47, 0.87), Black (OR = 0.69, CI = 0.07, 0.90), and Hispanic (OR = 0.17, CI = 0.06, 0.45) women were less likely than White women to have a personal health care provider. Women who were treated similar or better than those of other races were more likely to have a provider (OR = 1.11, CI = 1.05, 1.24) compared with those who were treated worse and within subgroup effects of racial treatment were also significant. Compared with their counterpart who were treated worse, Asian/AIAN/Other/multiracial (OR = 1.79, CI = 1.25, 2.47), Black (OR = 2.69, CI = 1.59, 12.26), and Hispanic (OR = 3.39, CI = 1.50, 7.65) women who were treated similarly to other races were more likely to have a personal health care provider.

Table 4.

Likelihood of Personal Health Care Provider

	Std err		t	Pr	OR	95% CI
Intercept	−11.17	3.07	−3.64	0.00
Age	2.96	0.75	3.94	<0.0001	19.25	4.42	83.95
Did not graduate from high school	−1.72	0.77	−2.23	0.03	0.18	0.04	0.81
High school graduate/Some college/Technical degree	−0.91	0.67	−1.37	0.17	0.40	0.11	1.49
Income $25,000 to <$50,000	0.45	0.78	0.58	0.56	1.57	0.34	7.23
Income <$25,000	−0.57	0.76	−0.75	0.45	0.56	0.13	2.51
Married	0.35	0.61	0.58	0.56	1.42	0.43	4.67
Lives in rural county	2.87	1.12	2.56	0.01	17.66	1.96	159.06
Asian/AIAN/Other	−3.21	1.30	−10.14	<0.0001	0.62	0.47	0.87
Black	−2.18	1.32	−2.65	0.01	0.69	0.07	0.90
Hispanic	−4.55	0.84	−5.43	<0.0001	0.17	0.06	0.45
Treated better, the same, neither better nor worse as other races when seeking health care	2.74	0.61	4.49	<0.0001	1.11	1.05	1.24
Asian/AIAN/Other*Treated better, the same, neither better nor worse as other races when seeking health care	4.93	0.73	20.49	<0.0001	1.79	1.25	2.47
Black*Treated better, the same, neither better nor worse as other races when seeking health care	2.08	0.59	3.55	0.00	2.69	1.59	12.26
Hispanic*Treated better, the same, neither better nor worse as other races when seeking health care	3.49	0.69	5.06	<0.0001	3.39	1.50	7.65

Dependent variable: Has one person or a group of doctors considered to be personal health care provider.

Estimates indicate the relative likelihood of having a personal health care provider.

Regression procedures used BRFSS stratification and cluster parameters to account for complex survey design.

Estimates weighted to reflect a representative population.

Boldface indicates significance at the 95% confidence level.

BRFSS, Behavioral Risk Factor Surveillance System; CI, confidence interval; OR, odds ratio; Pr, Probability; Std err, standard error.

Sensitivity analyses

Results from the regression model estimating the association between perceived racism and the likelihood of having a recent checkup/health care provider among the sample of respondents who reported having a health plan are listed in Table 5. Like estimates presented above, Black women (OR = 2.42, CI = 2.20, 3.39) who perceived being treated similar to or better than women of other races were more likely to have had a routine checkup within the last year compared with their counterparts. Black (OR = 2.62, CI = 1.88, 3.80), Asian/AIAN/Other/multiracial (OR = 2.94, CI = 1.77, 5.90), and Hispanic (OR = 1.91, CI = 1.79, 2.01) women with a health plan who perceived equal or better race-based treatment were more likely to have a personal health care provider compared with their counterparts who perceived worse racial treatment.

Table 5.

Robustness Test Among Respondents with a Health Plan

	Likelihood of having a routine checkup in the last year							Likelihood of personal health care provider
	Std err		t-Value	Pr > \|t\|	OR	95% CI		Std err		t-Value	Pr > \|t\|	OR	95% CI
Intercept	−3.33	1.92	−1.73	0.08				−10.41	3.23	−3.23	0.00
Age	1.24	0.45	2.74	0.01	3.45	1.42	8.39	3.13	0.83	3.76	0.00	22.88	4.47	117.20
Did not graduate from high school	0.02	0.50	0.03	0.98	1.02	0.38	2.69	−0.41	1.07	−0.38	0.70	0.67	0.08	5.46
High school graduate/Some college/Technical degree	−0.04	0.31	−0.12	0.91	0.96	0.53	1.77	−0.96	0.69	−1.39	0.16	0.38	0.10	1.48
Income $25,000 to <$50,000	0.80	0.35	2.26	0.02	2.22	1.11	4.44	0.34	0.86	0.40	0.69	1.41	0.26	7.62
Income <$25,000	0.42	0.36	1.15	0.25	1.52	0.74	3.10	−0.13	0.76	−0.17	0.87	0.88	0.20	3.89
Married	−0.21	0.26	−0.81	0.42	0.81	0.49	1.35	0.15	0.68	0.23	0.82	1.16	0.31	4.38
Lives in rural county	0.93	0.35	2.61	0.01	2.52	1.26	5.05	2.79	1.27	2.21	0.03	16.28	1.36	194.65
Asian/AIAN/Other	1.58	1.08	1.46	0.14	2.29	0.70	7.45	13.67	0.56	24.58	<0.0001	32.88	28.06	35.30
Black	−0.24	0.98	−0.25	0.81	1.02	0.35	3.00	0.43	0.42	1.03	0.30	1.00	0.24	4.12
Hispanic	11.30	0.89	1.68	0.10	5.91	4.58	8.35	0.30	0.66	0.46	0.64	0.89	0.37	2.12
Treated better, the same, neither better nor worse as other races when seeking health care	0.26	0.71	0.37	0.71	0.06	0.03	0.13	1.50	0.38	3.91	<0.0001	1.96	1.12	3.42
Asian/AIAN/Other*Treated better, the same, neither better nor worse as other races when seeking health care	−1.50	1.20	−1.25	0.21	0.29	0.04	1.99	1.59	0.69	2.31	0.02	2.94	1.77	5.90
Black*Treated better, the same, neither better nor worse as other races when seeking health care	0.52	1.09	2.48	0.04	2.42	2.20	3.39	0.86	1.18	2.73	0.05	2.62	1.88	3.80
Hispanic*Treated better, the same, neither better nor worse as other races when seeking health care	−11.49	1.09	−0.56	0.42	0.08	0.19	6.27	0.85	0.62	2.37	0.02	1.91	1.79	2.01

Estimates indicate the relative likelihood of having a checkup/provider.

Regression procedures used BRFSS stratification and cluster parameters to account for complex survey design.

Estimates weighted to reflect a representative population.

Boldface indicates significance at the 95% confidence level.

BRFSS, Behavioral Risk Factor Surveillance System; CI, confidence interval; OR, odds ratio; Pr, Probability; Std err, standard error.

Table 6 presents results of the GSEM assessing whether perceived racism mediates racial and ethnic differences in post-stroke health care utilization. Perceived racism among Asian/AIAN/Other/multiracial participants showed a minimal, insignificant effect on the likelihood of having a personal doctor. While the total effect (14.60, standard error [SE] = 1.45) was significant, it was transmitted primarily through the direct effect (14.51, SE = 1.06). Among Black respondents, perceived racism only had a small, insignificant, indirect effect (0.18, SE = 0.82) on the association between race and the likelihood of having a personal doctor, while the total effect (1.76, SE = 0.91) was significant and transmitted directly (1.58, SE = 1.36). Similarly, the likelihood of having had a routine checkup within the last year among Black respondents was small and insignificant (0.12, SE = 0.57) showing that the direct (0.83, SE = 0.66) and indirect (−0.71, 0.66) slightly offset one another.

Table 6.

Mediation Analysis of Perceived Racism in Health Care Utilization

	Have a personal doctor (0,1)						Had a routine checkup within the last year (0,1)						Has a health plan (0,1)
	Std err		t	Pr	95% CI		Std err		t	Pr	95% CI		Std err		t	Pr	95% CI
Intercept	−3.16	1.87	−1.69	0.09	−6.82	0.51	−5.11	1.36	−3.75	0.00	−7.79	−2.44	−2.33	2.63	−0.88	0.38	−7.49	2.84
Age	1.62	0.38	4.31	0.00	0.88	2.35	1.63	0.29	5.55	0.00	1.05	2.21	1.08	0.78	1.39	0.17	−0.45	2.61
Did not graduate from high school	−1.92	0.98	−1.96	0.05	−3.85	0.01	−0.74	0.94	−0.79	0.43	−2.59	1.10	−1.42	1.11	−1.28	0.20	−3.60	0.76
High school graduate/Some college/Technical degree	−0.45	0.26	−1.72	0.09	−0.97	0.06	0.10	0.26	0.36	0.72	−0.42	0.61	0.13	0.41	0.31	0.76	−0.68	0.93
Income $25,000 to <$50,000	−0.77	0.29	−2.69	0.01	−1.33	−0.21	−0.25	0.24	−1.08	0.28	−0.72	0.21	0.02	0.73	0.03	0.98	−1.41	1.45
Income <$25,000	−0.13	0.33	−0.39	0.70	−0.77	0.52	0.16	0.25	0.65	0.52	−0.33	0.65	0.70	0.73	0.97	0.33	−0.72	2.13
Married	0.23	0.29	0.77	0.44	−0.35	0.80	0.01	0.21	0.05	0.96	−0.40	0.42	0.76	0.54	1.42	0.16	−0.29	1.82
Lives in rural county	−0.11	0.34	−0.31	0.76	−0.78	0.56	0.52	0.25	2.04	0.04	0.02	1.02	0.95	0.54	1.77	0.08	−0.10	2.00
Asian/AIAN/Other	−1.58	1.36	−1.16	0.25	−4.25	1.09	16.92	1.17	14.40	0.00	14.62	19.22	16.59	1.43	11.58	0.00	13.78	19.39
Black	14.51	1.06	13.65	0.00	12.43	16.59	−0.47	0.94	−0.50	0.61	−2.32	1.37	16.32	1.22	10.37	0.22	13.56	18.18
Hispanic	−1.83	1.43	−1.28	0.20	−4.64	0.98	0.93	1.25	0.75	0.45	−1.51	3.38	−1.53	1.50	−1.02	0.31	−4.47	1.41
Treated better, the same, neither better nor worse as other races	0.24	1.09	0.22	0.83	−1.90	2.37	0.95	0.59	1.60	0.11	−0.22	2.12	0.41	0.98	0.42	0.68	−1.51	2.33
Asian/AIAN/Other*Treated better, the same, neither better nor worse as other races	1.37	1.43	0.96	0.34	−1.43	4.18	0.75	0.64	1.17	0.24	−0.51	2.01	1.29	1.09	1.18	0.24	−0.86	3.43
Black*Treated better, the same, neither better nor worse as other races	5.20	1.13	13.44	0.00	2.99	7.42	0.83	0.66	2.25	0.02	1.47	2.12	1.06	1.11	0.95	0.34	−1.12	3.24
Hispanic*Treated better, the same, neither better nor worse as other races	0.11	1.49	0.08	0.94	−2.81	3.04	1.08	0.66	1.63	0.10	−0.22	2.37	−1.08	1.02	−1.06	0.29	−3.07	0.92
	Treated better, the same, neither better nor worse than other races (0,1)						Treated better, the same, neither better nor worse than other races (0,1)						Treated better, the same, neither better nor worse than other races (0,1)
Intercept	−0.75	1.97	−0.38	0.70	−4.61	3.10	−0.75	1.97	−0.38	0.70	−4.61	3.10	−0.75	1.97	−0.38	0.70	−4.61	3.10
Age	1.38	0.48	2.91	0.00	0.45	2.32	1.38	0.48	2.91	0.00	0.45	2.32	1.38	0.48	2.91	0.00	0.45	2.32
Did not graduate from high school	−3.35	1.23	−2.72	0.01	−5.77	−0.94	−3.35	1.23	−2.72	0.01	−5.77	−0.94	−3.35	1.23	−2.72	0.01	−5.77	−0.94
High school graduate/Some college/Technical degree	−0.40	0.50	−0.79	0.43	−1.38	0.59	−0.40	0.50	−0.79	0.43	−1.38	0.59	−0.40	0.50	−0.79	0.43	−1.38	0.59
Income $25,000 to <$50,000	−0.86	0.49	−1.76	0.08	−1.82	0.10	−0.86	0.49	−1.76	0.08	−1.82	0.10	−0.86	0.49	−1.76	0.08	−1.82	0.10
Income <$25,000	−0.38	0.58	−0.65	0.52	−1.51	0.76	−0.38	0.58	−0.65	0.52	−1.51	0.76	−0.38	0.58	−0.65	0.52	−1.51	0.76
Married	1.15	0.76	1.51	0.13	−0.34	2.65	1.15	0.76	1.51	0.13	−0.34	2.65	1.15	0.76	1.51	0.13	−0.34	2.65
Lives in rural county	−0.19	0.87	−0.21	0.83	−1.89	1.52	−0.19	0.87	−0.21	0.83	−1.89	1.52	−0.19	0.87	−0.21	0.83	−1.89	1.52
Asian/AIAN/Other	−0.75	0.48	−1.54	0.12	−1.69	0.20	0.40	0.63	0.63	0.53	−0.84	1.63	−0.75	0.48	−1.54	0.12	−1.69	0.20
Black	0.40	0.63	0.63	0.53	−0.84	1.63	−0.75	0.48	−1.54	0.12	−1.69	0.20	−0.17	0.77	−0.22	0.83	−1.67	1.33
Hispanic	−0.17	0.77	−0.22	0.83	−1.67	1.33	−0.17	0.77	−0.22	0.83	−1.67	1.33	0.40	0.63	0.63	0.53	−0.84	1.63
Total Effect
Asian/AIAN/Other	14.60	1.45	10.09	0.00	11.77	17.44	1.13	1.03	1.10	0.27	−0.89	3.14	1.45	1.48	0.98	0.33	−1.44	4.34
Black	1.76	0.91	−1.93	0.05	0.03	3.54	0.12	0.57	0.21	0.83	−0.99	1.23	0.76	0.66	1.15	0.25	−0.53	2.04
Hispanic	−1.87	1.30	−1.44	0.15	−4.42	0.67	0.91	0.94	0.97	0.33	−0.93	2.76	−1.15	0.88	−1.30	0.20	−2.88	0.59
Indirect Effect
Asian/AIAN/Other	0.09	0.46	0.21	0.84	−0.80	0.99	0.38	0.63	0.60	0.55	−0.86	1.62	0.16	0.47	0.34	0.73	−0.77	1.09
Black	0.18	0.82	−0.22	0.83	−1.79	1.43	0.71	0.66	−1.08	0.28	−2.00	0.58	−0.30	0.76	−0.40	0.69	−1.78	1.18
Hispanic	−0.04	0.27	−0.15	0.88	−0.58	0.50	−0.16	0.73	−0.22	0.83	−1.60	1.28	−0.07	0.37	−0.19	0.85	−0.79	0.65
Direct Effect
Asian/AIAN/Other	14.51	1.06	13.65	0.00	12.43	16.59	0.75	0.64	1.17	0.24	−0.50	2.01	1.29	1.09	1.18	0.24	−0.86	3.43
Black	1.58	1.36	1.16	0.25	1.09	4.24	0.83	0.66	2.25	0.02	0.47	2.12	1.06	1.11	0.95	0.34	−1.12	3.24
Hispanic	1.83	1.43	−1.28	0.20	−4.64	0.97	1.08	0.66	1.63	0.10	−0.22	2.37	−1.08	1.02	−1.06	0.29	−3.07	0.92

Generalized structural equation model.

Survey design parameters and sampling weight used in the generation of estimates.

Boldface indicates significance at 95% confidence interval.

CI, confidence interval; OR, odds ratio; Pr, Probability; Std err, standard error.

Discussion

Study aim and research question

This study aimed to investigate how race affects the likelihood of having a routine checkup, a health plan, and a personal doctor among post-stroke women. Specifically, it explored differences in race, health care utilization, and health status among women who have had a stroke. The study focused on understanding how racial variations influence responses to race and health care utilization, providing valuable insights into the underlying factors contributing to health inequities.

Key findings

The sample used in this study was 75% White,12% Black, and 6% Hispanic. While this BRFSS module was not administered nationally, the demographic portions represented herein resemble the national population of women with stroke in the United States.^11,38 The findings revealed significant disparities in health care utilization based on perceived treatment. Individuals who felt they were treated poorly had less favorable outcomes in accessing health care compared with those who perceived their treatment as equitable or better. These disparities highlight the critical role of social determinants of health (SDOH), which encompass “the conditions in which people are born, grow, live, and age” and the “wider set of forces and systems shaping the conditions of daily life.”¹⁵ Non-medical factors such as race, health care trust, patient-physician communication, and access to services are pivotal in shaping health outcomes and disparities. The observed effect sizes, represented by odds ratios, underscore the significant associations between perceived treatment and health care utilization, reinforcing the impact of equitable treatment on health care access and outcomes.

Mechanisms explaining the findings

Several mechanisms may explain these findings: health care trust, patient–physician communication, and addressing SDOH. Perceived treatment that is the same or better may reinforce health care trust, leading to increased engagement. Effective communication can mitigate perceived biases, and factors such as socioeconomic status and access to care can reduce barriers to health care utilization. The inclusion of control variables in the analysis, such as age, marital status, education level, income, and health insurance coverage, provided a clearer understanding of the relationships between outcome variables and predictors, highlighting the nuanced interplay of demographic and socioeconomic factors in shaping health outcomes.

Contextual factors and intersectionality

Contextual factors such as racism, health care trust, and patient–physician communication are crucial in understanding health care utilization. For example, perceived racial treatment significantly influences whether individuals trust the health care system, affecting their engagement with health care services. While sensitivity analyses indicated that perceived racism was not a significant mediating factor in health care utilization, disparities in stroke recovery and incidence highlight systemic challenges faced by different racial and ethnic groups. Intersectionality—considering the overlapping impacts of race, gender, and socioeconomic status—complicates these disparities further.³⁹ This approach suggests that a nuanced understanding is necessary to address these issues effectively. By addressing these contextual factors and recognizing the interplay of various identities, the study underscores the need for targeted interventions to mitigate disparities in health care utilization among post-stroke women from diverse backgrounds. The concept of intersectionality highlights that individuals experience overlapping systems of discrimination and privilege based on factors such as race, gender, socioeconomic status, and health condition. In this study, the absence of racism as a mediating factor may reflect the complex interplay of other intersecting influences, such as gendered expectations of caregiving, systemic biases specific to women’s health, or the compounded effects of socioeconomic disadvantage. These factors might independently or collectively shape health care utilization patterns, diminishing the direct mediating role of racism in the observed disparities.

Strengths

The study’s design incorporates several strengths that enhance its contribution to understanding health care experiences among post-stroke women. It acknowledges the intersectionality of race, gender, age, and socioeconomic status in shaping health outcomes, offering a nuanced analysis of disparities. By adjusting for relevant covariates, the study minimizes the influence of extraneous factors, thereby strengthening the validity of its findings. The utilization of established data sources, such as the BRFSS and R2R module, bolsters the reliability and generalizability of the results. The large sample size further enhances the generalizability, making the findings more applicable to a broader population. Additionally, the use of standardized measures for assessing perceived treatment and health care utilization ensures data reliability and consistency across different groups.

Limitations

There are notable limitations in the study that must be considered when evaluating its results and implications for understanding health care utilization among post-stroke women. One significant limitation is the reliance on self-reported data, which can introduce subjectivity and potential response bias into perceptions of treatment. Respondents in this national survey reported whether or not they received a diagnosis of stroke. A clinical diagnosis of stroke reported in a medical record would be optimal, yet this reporting approach is common in national surveys. Additionally, we acknowledge that even though all respondents were stroke survivors, the time frame post-stroke is potentially highly variable with some occurring many years prior to the survey. However, when considering that the 10-year stroke recurrence rate is >11%,⁴⁰ management of stroke risk persists long-term and for most throughout their lifespan. Selection bias is another concern, as the sample may not fully represent all post-stroke women, potentially affecting the applicability of the findings. Methodologically, the study’s cross-sectional design limits the ability to infer causality, restricting conclusions about cause-and-effect relationships. Additionally, the findings may not be generalizable to other populations or settings due to the specific sample and geographic focus.

The accuracy of standardized measures for assessing perceived treatment and health care utilization depends on participants’ ability to recall and honestly report their experiences. Furthermore, the R2R module’s single item may not fully capture the complexity of perceptions of racial treatment. Although covariates were adjusted for in the analysis, there may be unmeasured or residual confounding variables that could influence the associations between perceived racial treatment and health care utilization. Lastly, while the BRFSS is a valuable dataset, it has inherent methodological constraints, such as the exclusion of institutionalized individuals and potential non-coverage of certain populations, which may limit the comprehensiveness of the data.

While the use of standardized measures for perceived treatment and health care utilization is beneficial, the accuracy of these measures depends on participants’ ability to recall and honestly report their experiences. Furthermore, the R2R Module’s single item may not capture the full complexity of perceptions of racial treatment. Although covariates were adjusted for in the analysis, there may be unmeasured or residual confounding variables that could influence the associations between perceived racial treatment and health care utilization.

Furthermore, not all potentially confounding factors were available in the BRFSS data. For example, the timing, frequency, and intensity of health care utilization may vary throughout the stroke recovery process and relate to an individual’s insurance status at the time of the stroke event. Unfortunately, neither time post stroke nor insurance status at the time of stroke were available in the dataset utilized for this analysis.

Finally, while the BRFSS is a valuable dataset, it has inherent methodological constraints, such as the exclusion of institutionalized individuals and potential non-coverage of certain populations, which may affect the comprehensiveness of the data.

To address the study’s limitations and enhance the validity, reliability, and generalizability of the findings, we employed several strategies. We adjusted for covariates in the logistic regression analysis to control for potential confounding variables, isolating the effect of perceived racial treatment on health care utilization and improving result accuracy. Utilizing a large sample of over 10,000 post-stroke women strengthened the robustness and generalizability of the findings, reducing sampling error and providing more reliable estimates. We applied data weighting using an iterative proportional fitting process, known as “raking,” to adjust for demographic variation, selection probability, and nonresponse bias, ensuring that the sample more accurately represents the broader population. We employed standardized measures for assessing perceived treatment and health care utilization to maintain consistency in data collection and analysis, enhancing reliability and comparability across groups. We incorporated complex sampling adjustments by addressing the BRFSS’s sampling framework with individual-level weights and specific survey analytic commands in SAS 9.4, reducing biases associated with the survey design. Finally, we used descriptive and inferential statistics, including chi-squared tests, t-tests, and logistic regression, to provide a detailed understanding of demographic and perception variables, and employed interaction terms in the logistic regression analysis to examine within-group differences. These strategies effectively mitigated methodological and bias-related limitations, thereby improving the study’s validity, reliability, and generalizability.

Implications for practice

This study provides strong evidence of an association between perceived same or better treatment and improved health care utilization among post-stroke women. Addressing perceived treatment plays a crucial role in enhancing health care access and outcomes, especially within the context of SDOH. Effective patient–physician communication, particularly in intercultural settings, mitigates perceived racial biases and boosts health care utilization. To apply these findings, health care providers should focus on improving patient–physician communication and receive training in cultural competence to better understand and meet the diverse needs of patients. They must implement targeted interventions to address the specific concerns and barriers faced by post-stroke women from various racial and ethnic backgrounds. Additionally, health care systems need to adopt policies that promote equity and reduce biases in health care delivery. Strategies such as building patient trust, addressing socioeconomic barriers, and engaging community-based approaches will further enhance health care access and utilization. Regularly monitoring and evaluating health care practices will help identify persistent disparities and ensure that interventions remain effective, ultimately leading to better health care outcomes for post-stroke women from diverse backgrounds.

Future research directions

Future research should concentrate on several key areas to deepen our understanding of health care utilization among post-stroke women. Researchers need to conduct longitudinal studies to track how perceived racial treatment impacts long-term health care utilization and recovery outcomes, establishing causal relationships. Evaluating targeted interventions that aim to improve patient–physician communication and reduce racial biases will reveal effective strategies for enhancing trust in the health care system. Expanding research to include diverse racial and ethnic groups will provide a more comprehensive view of disparities. Researchers should use intersectional frameworks to examine how overlapping identities influence health care experiences and incorporate qualitative methods to gain deeper insights into patient perceptions. Developing and validating comprehensive patient experience surveys will guide targeted interventions. Additionally, researchers need to assess the impact of policy changes on reducing health care disparities and investigate specific mechanisms through which racial treatment affects trust. Exploring the role of digital health tools and comparing how different countries address health care disparities will offer valuable insights into effective practices and policies.

Conclusion

This study reaffirms the significant impact of perceived racism on health care utilization, shedding light on how treatment disparities hinder health care access. It emphasizes the intersectionality of race, health care trust, and patient–physician communication as key social determinants of health. The findings underscore the urgent need for systemic interventions to address health care inequities, enhance patient engagement, and promote equitable health care access across diverse populations.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Authors’ Contributions

M.J.: Contributed to the data curation, analysis, interpretation, writing, and editing of this article. C.E.: Contributed to the conceptualization, contextualization, interpretation, writing, and editing of this article. A.M.: Contributed to the conceptualization of this article.

Funding Information

No funding was received for this article.

References

Martin

, Aday

, Almarzooq

, et al.; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee. 2024 heart disease and stroke statistics: A report of US and global data from the American Heart Association. Circulation, 2024; 149(8):e347–e913; doi: 10.1161/CIR.0000000000001209

Towfighi

, Boden-Albala

, Cruz-Flores

, et al.; American Heart Association Stroke Council; Council on Cardiovascular and Stroke Nursing; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; Council on Hypertension; Council on the Kidney in Cardiovascular Disease; and Council on Peripheral Vascular Disease. Strategies to reduce racial and ethnic inequities in stroke preparedness, care, recovery, and risk factor control: A scientific statement from the American Heart Association. Stroke, 2023; 54(7):E371–E388; doi: 10.1161/STR.0000000000000437

Yang

, Tong

, Schieb

, et al. Morbidity and mortality weekly report stroke mortality among black and white adults aged ≥35 years before and during the COVID-19 pandemic-United States, 2015-2021. MMWR Morb Mortal Wkly Rep, 2023; 72(16):431–436.

Tsao

, Aday

, Almarzooq

, et al. Heart disease and stroke statistics-2022 update: A report from the American Heart Association. Circulation, 2022; 145(8):E153–E639; doi: 10.1161/CIR.0000000000001052

Ikeme

, Kottenmeier

, Uzochukwu

, et al. Evidence-based disparities in stroke care metrics and outcomes in the United States: A systematic review. Stroke, 2022; 53(3):670–679; doi: 10.1161/STROKEAHA.121.036263

Cao

, Jain

, Lu

, et al. Secondary stroke risk reduction in black adults: A systematic review. J Racial Ethn Health Disparities, 2023; 10(1):306–318; doi: 10.1007/s40615-021-01221-2

Esparza

, de Almeida

, Verduzco-Gutierrez

. Disparities in care after stroke for black and Hispanic patients: Intersection with socioeconomic status and insurance status—A narrative review. Curr Phys Med Rehabil Rep, 2023; 11(4):435–442; doi: 10.1007/s40141-023-00424-4

Yoon

, Bushnell

. Stroke in women: A review focused on epidemiology, risk factors, and outcomes. J Stroke, 2023; 25(1):2–15; doi: 10.5853/jos.2022.03468

Cherian

. Women and ischemic stroke: Disparities and outcomes. Neurol Clin, 2023; 41(2):265–281; doi: 10.1016/j.ncl.2022.10.001

10.

Jacobs

, Ellis

. Estimating the cost and value of functional changes in communication ability following telepractice treatment for aphasia. PLoS One, 2021; 16(9):e0257462; doi: 10.1371/journal.pone.0257462

11.

Jiménez

, Manson

JAE

, Cook

, et al. Racial variation in stroke risk among women by stroke risk factors. Stroke, 2019; 50(4):797–804; doi: 10.1161/STROKEAHA.117.017759

12.

, Amarilla Vallejo

, Cantalapiedra Calvete

, et al. Stroke outcomes in women: A population-based cohort study. Stroke, 2022; 53(10):3072–3081; doi: 10.1161/STROKEAHA.121.037829

13.

Jimenez

, Williams

, Keenan

, et al. Bilingual randomized controlled trial design, of a telephone-based intervention to promote rehabilitation adherence; A study focus on recruitment of Hispanic children with traumatic brain injury. Contemp Clin Trials, 2023; 135:107362; doi: 10.1016/j.cct.2023.107362

14.

Metcalf

, Zhang

. Racial and ethnic disparities in the usage and outcomes of ischemic stroke treatment in the United States. J Stroke Cerebrovasc Dis, 2023; 32(12):107393; doi: 10.1016/j.jstrokecerebrovasdis.2023.107393

15.

Calancie

, Leng

, Whitsel

, et al. Racial disparities in stroke incidence in the Women’s Health Initiative: Exploring biological, behavioral, psychosocial, and social risk factors. SSM Popul Health, 2024; 25:101570; doi: 10.1016/j.ssmph.2023.101570

16.

Knight

, Branch

, Ortel

. Antiphospholipid syndrome: Advances in diagnosis, pathogenesis, and management. BMJ, 2023; 380:e069717; doi: 10.1136/bmj-2021-069717

17.

Sheehy

, Aparicio

, Palmer

, et al. Perceived interpersonal racism and incident stroke among US black women. JAMA Netw Open, 2023; 6(11):E2343203; doi: 10.1001/jamanetworkopen.2023.43203

18.

Buie

JNJ

, Zhao

, Burns

, et al.; WISSDOM Research Center Study Group. Racial disparities in stroke recovery persistence in the post-acute stroke recovery phase: Evidence from the health and retirement study. Ethn Dis, 2020; 30(2):339–348; doi: 10.1886/5/ED.30.2.339

19.

Love

, Brooks

, Cox

, et al. The effects of racism and resilience on Black stroke- survivor quality of life: Study protocol and rationale for a mixed-methods approach. Front Neurol, 2022; 13:885374; doi: 10.3389/FNEUR.2022.885374

20.

Billups

, Thelamour

, Thibodeau

, et al. On intersectionality: Visualizing the invisibility of Black women. Cogn Res Princ Implic, 2022; 7(1):100; doi: 10.1186/s41235-022-00450-1

21.

Yip

, Cheah

CSL

, Kiang

, et al. Rendered invisible: Are Asian Americans a model or a marginalized minority? Am Psychol, 2021; 76(4):575–581; doi: 10.1037/amp0000857

22.

Hoffman

, Trawalter

, Axt

, et al. Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proc Natl Acad Sci U S A, 2016; 113(16):4296–4301; doi: 10.1073/pnas.1516047113

23.

Jefferies

. The strong black woman: Insights and implications for nursing. J Am Psychiatr Nurses Assoc, 2022; 28(4):332–338; doi: 10.1177/1078390320983900

24.

Levine

, Duncan

, Nguyen-Huynh

, et al. Interventions targeting racial/ethnic disparities in stroke prevention and treatment. Stroke, 2020; 51(11):3425–3432; doi: 10.1161/STROKEAHA.120.030427

25.

Gadson

, Wallace

, Young

, et al. The relationship between health-related quality of life, perceived social support, and social network size in African Americans with aphasia: A cross-sectional study. Top Stroke Rehabil, 2022; 29(3):230–239; doi: 10.1080/10749357.2021.1911749

26.

Stamm

, Royan

, Madsen

. Association of prior stroke with health care perceptions of adequate emergency care in women. Stroke, 2024; 55(2):301–304; doi: 10.1161/STROKEAHA.123.044967

27.

Alkhamees

, Alasqah

. Patient-physician communication in intercultural settings: An integrative review. Heliyon, 2023; 9(12):e22667; doi: 10.1016/j.heliyon.2023.e22667

28.

Howard

, Madsen

, Kleindorfer

, et al. Sex and race differences in the association of incident ischemic stroke with risk factors. JAMA Neurol, 2019; 76(2):179–186; doi: 10.1001/jamaneurol.2018.3862

29.

Hosman

, Engels

, den Ruijter

, et al. Call to action for enhanced equity: Racial/ethnic diversity and sex differences in stroke symptoms. Front Cardiovasc Med, 2022; 9:874239; doi: 10.3389/fcvm.2022.874239

30.

Ben

, Cormack

, Harris

, et al. Racism and health service utilisation: A systematic review and meta-analysis. PLoS One, 2017; 12(12):e0189900; doi: 10.1371/journal.pone.0189900

31.

Adegbembo

, Tomar

, Logan

. Perception of racism explains the difference between Blacks’ and Whites’ level of healthcare trust. Ethn Dis, 2006; 16(4):792–798.

32.

Onukwugha

, McRae

, Castillo

. Cost-related non-utilization of health services and self-perceived reactions to race. Ethn Dis, 2020; 30(3):399–410; doi: 10.1886/5/ED.30.3.399

33.

Schneider

, Clark

, Rakowski

, et al. Evaluating the impact of non-response bias in the Behavioral Risk Factor Surveillance System (BRFSS). J Epidemiol Community Health, 2012; 66(4):290–295.

34.

Erickson

, Dumoulin-Smith

. User guide: A guide to disability statistics from the Behavioral Risk Factors Surveillance System. DigitalCollections@ILR, 2009.

35.

Srebotnjak

, Mokdad

, Murray

. A novel framework for validating and applying standardized small area measurement strategies. Popul Health Metr, 2010; 8(1):26; doi: 10.1186/1478-7954-8-26

36.

Jones

. Confronting institutionalized racism. Phylon, 2002; 50(1/2):7; doi: 10.2307/4149999

37.

White

, Lawrence

, Tchangalova

, et al. Socially-assigned race and health: A scoping review with global implications for population health equity. Int J Equity Health, 2020; 19(1):25; doi: 10.1186/s12939-020-1137-5

38.

Bushnell

, Chaturvedi

, Gage

, et al. Sex differences in stroke: Challenges and opportunities. J Cereb Blood Flow Metab, 2018; 38(12):2179–2191; doi: 10.1177/0271678X18793324

39.

Crenshaw

. Mapping the margins: Intersectionality, identity politics, and violence against women of color. The Public Nature of Private Violence: Women and the Discovery of Abuse, 2013:93–118; doi: 10.2307/1229039

40.

Flach

, Muruet

, Wolfe

CDA

, et al. Risk and secondary prevention of stroke recurrence. Stroke, 2020; 51(8):2435–2444; doi: 10.1161/STROKEAHA.120.028992