Race and ethnic disparities in health are among the most well-documented findings in both the biomedical and social sciences [1]. These disparities are a consequence of systemic racism and xenophobia, which impose structural barriers to the receipt of timely high-quality care, and constrain access to health-enhancing social environments, including safe neighborhoods, schools, and workplaces [2]. Likewise, many studies have linked gender bias and discrimination to both quality of and access to health care (e.g., [3], which may exacerbate gender gaps in chronic conditions, functional limitations, and depression [4]. Researchers also have identified interpersonal dynamics within and beyond health care settings that bear on patient well-being, documenting how stigma [5], perceived discrimination [6], and multiple forms of racism [2] and sexism [4] contribute to health disparities.

Despite extensive research on race and gender disparities in health, surprisingly little attention has focused on the complex ways race/ethnicity and gender may intersect to shape microlevel interactions in health care settings. Theoretical writings on intersectionality underscore that race and gender have multiplicative effects on one’s lived experience, including the degree to which interactions with health care providers and gatekeepers are marked by respect, care, and patience versus disrespect, invalidation, and abruptness [7]. Further, this work typically relies on population-based data and does not focus on the distinctive circumstances of patients receiving care in low-income settings. Female patients and patients belonging to ethnic or racial minority groups are especially likely to receive care in publicly funded health care settings, given systemic racism and sexism in the United States, which renders women and ethnic minorities especially vulnerable to poverty. Thus, it is important to understand disparities in the perceived quality of care received in these contexts.

Using data from the 2014 Health Center Patient Survey (HCPS), we draw on an intersectionality framework to examine how race and gender intersect to affect perceptions of disrespectful treatment by health care providers and health center staff in federally qualified health centers (FQHCs). Our large data set allows us to contrast the experiences of white, Black, Hispanic, Asian, and Native American men and women, providing a nuanced exploration of the multiplicative influences of race/ethnicity and gender. We further evaluate the extent to which these associations are accounted for by two potential mechanisms: language fluency and health. Documenting whether and how race and gender intersect to shape patient experiences within FQHCs is an important goal. Disrespectful or demeaning exchanges may discourage economically and socially vulnerable patients from receiving timely preventive or curative care, rendering them even more vulnerable to compromised health and longevity.

1 Background

1.1 Disrespectful encounters in healthcare settings

Researchers have documented race, ethnic and gender differences in patient satisfaction across a range of settings including nationally representative samples (e.g., [8] and specific populations such as patients seeking care in the Veterans’ Administration health care system [9]. While overall levels of satisfaction have been explored extensively, we are unaware of studies focusing specifically on perceptions of respectful treatment. Respectful treatment is an important dimension of health care, especially in settings that serve historically marginalized populations. Philosophers characterize respect as a core dimension of social justice; respect and recognition give individuals a sense of self-worth and the confidence to act autonomously [10]. Psychologists elaborate that respect may be bestowed at the categorical level, such that one’s group membership like their race or gender, may trigger disrespectful treatment from individual actors and the social institutions in which they are embedded. Disrespectful encounters may cause feelings of delegitimization, invisibility, and dehumanization [11]. Drawing on this conceptualization, we explore disparities in patients’ perceptions of four dimensions of respectful treatment: feeling respected, being listened to, receiving sufficient time from health providers, and receiving information in a way that one could understand by health care providers. We also consider interactions with health care center staff (e.g., receptionists, clerks) who serve as gatekeepers in clinical settings, focusing on whether a patient felt they were treated respectfully.

Disrespectful treatment is a form of microaggression, or the “brief and commonplace daily verbal, behavioral, or environmental indignities, whether intentional or unintentional, that communicate hostile, derogatory, or negative racial slights and insults toward people of color” ([12], 271). In health care settings, microaggressions may be experienced when “treatment providers who are in positions of authority inadvertently marginalize members of minority groups through culturally insensitive interactions” ([13], 2). Conversely, high-quality interactions between provider and patient are “characterized by mutual respect, openness, and a balance in their respective roles in decision-making” ([14], 90). Microaggressions are an established mechanism contributing to Black-white disparities in health and are implicated in Black persons’ higher rates of hypertension, poor self-rated health, and other harmful health outcomes [15, 16].

Microaggressions may be gender-based as well as race-based. Women are more likely than men to be characterized as hypochondriacal, their illnesses are more likely to be misdiagnosed as psychosomatic, and their subjective experiences of pain and illness discounted [14]. Painful and debilitating symptoms that affect women specifically, such as endometriosis, fibromyalgia, and migraines have been historically dismissed, minimized, or ignored by health care providers [17–19]. However, it is unclear whether invalidating encounters that have been documented among women and ethnic/racial minority patients differ along intersectional lines. That is, are there multiplicative effects of race/ethnicity and gender on perceived respectful treatment in clinical encounters? Our analysis of microaggressions in health care may reveal and inform practices to address “processes of oppression or privilege, and policies of institutional practices” [20].

1.2 Intersectional approaches to understanding health care encounters

Intersectional approaches reveal how systems of inequality based on gender, race, ethnicity, class and other axes of social stratification intersect to shape personal experiences. All forms of inequality are conceptualized as mutually reinforcing and must be examined simultaneously to prevent one form of inequality from reinforcing another [7]. As a corrective to the “mutually exclusive” approach to understanding health care microaggressions [21], we contrast 10 distinct race-gender categories to identify the multiplicative effects of race/ethnicity and gender on patients’ experiences of (dis)respectful health care encounters: Non-Hispanic white, non-Hispanic Black, non-Hispanic Asian, non-Hispanic American Indian or Alaskan native, and Hispanic men and women.

We hypothesize that women and Black, Asian, Hispanic, and Native American patients will report less respectful treatment in health care settings relative to their male and white counterparts, respectively, consistent with prior writings on race- and gender-related microaggressions [12]. However, we further anticipate that the magnitude of gender disparities may differ on the basis of race/ethnicity, and racial/ethnic disparities may differ on the basis of gender, revealing how individuals’ multiple social locations intersect and shape their access to privilege and power [7, 5].). We do not generate specific hypotheses for each of the ten race/gender groups, although we do offer several examples to suggest how race/ethnicity and gender may intersect in health care settings.

Black women historically have been viewed as physically less “delicate” and “feminine” than their white or Asian counterparts, contributing to health care providers neglecting or minimizing their health care concerns, a process that has been implicated in delayed detection and poorer prognoses for Black women’s chronic illnesses including breast cancer [22]. As such, Black women may be more likely to report being disrespected in clinical settings, relative to other women and Black men. Asian American men, by contrast, have been marginalized historically within hierarchies of hegemonic masculinity, a function of racial stereotypes labeling them as passive, weak, or effeminate. These feelings of marginalization may be amplified for Asian men as well as women receiving care in low-income health care settings, as economic precarity may violate the “model minority” stereotype, which portrays Asians as successful and self-reliant [23]. Thus, we explore how race and gender intersect to influence self-reported experiences of perceived disrespectful treatment from health care providers and staff in a large nationally represented sample of patients receiving care in low-income health settings.

1.3 Understanding disparate treatment in low-income health settings

Our analyses focus specifically on low-income health care settings supported by the federal Health Resources and Services Administration (HRSA), also referred to as federally qualified health centers (FQHCs). FQHCs were first established in 1965 under President Lyndon B. Johnson’s War on Poverty programs and are funded through Section 330 of the Public Health Service Act. They are an essential site of health care for women and Black and Hispanic patients, given their heightened risk of poverty. Most studies of disrespectful or discriminatory health care encounters have used national population samples like the Behavioral Risk Factor Surveillance System (BRFSS) and have considered the patient’s economic status as either a statistical control or as a potential mechanism contributing to race/ethnic differences in perceived discriminatory treatment [24]. However, our objective is to explicitly examine whether low-income health care settings—a space in which women and ethnic/racial minority patients disproportionately seek care—also perpetuate race- and gender- based microaggressions.

Understanding perceived disrespectful treatment based on race and gender among FQHC patients is an important and policy-relevant goal. More than 1400 FQHCs in the U.S. serve more than 29 million patients, with the total number of centers increasing by more than 80 percent between 2007 and 2014 [25]. This growth was largely attributable to the Affordable Care Act, which established the Community Health Center Fund (CHCF) to support the expansion of FQHCs between 2011 and 2015, allocating $11 billion over five years [26]. In 2014, the year in which the Health Center Patient Survey (HCPS) was conducted, slightly more than 10 percent of all Americans and 14 percent of all ethnic minorities received care at FQHCs, with rates ranging from 4.4 percent among Asian Americans and 5.5 percent among whites, to over 10 percent among Blacks, Hispanics, and Native Americans [27].

Most FQHCs patients live beneath the federal poverty line, lack health insurance, or are Medicaid-dependent, and a disproportionate share have complex health care needs. FQHCs also provide essential preventive services including flu and COVID-19 shots, mammograms, and breast cancer screenings, as well as some reproductive health services for low-income women. Given the significant and otherwise unmet health care needs of patients at FQHCs, perceived disrespectful treatment may discourage these already vulnerable patients from seeking timely preventive or curative care. Thus, we use data from the 2014 HCPS to examine the extent to which race and gender intersect to affect patients’ perceptions of disrespectful treatment from health care providers and staff in low-income health care settings.

1.4 Potential explanatory mechanisms: language fluency and health

Two potential mechanisms may partially account for race/gender disparities in perceived disrespectful treatment: language fluency and health. Language fluency varies based on one’s racial and ethnic background and may contribute to disparities in such treatment. Just 8 percent of Blacks and 15 percent of Whites report that they speak a language other than English at home, whereas rates are somewhat higher among Native Americans (28 percent) and dramatically higher among Hispanics (75 percent) and Asians (77 percent) [28].

Persons with limited English proficiency are particularly likely to report disrespectful treatment from health care providers, including not being understood, or having providers present information too quickly to easily comprehend. They also report other disadvantageous health outcomes which are associated with disrespectful treatment, including lower satisfaction with health care and higher rates of medical errors [29]. The 2014 HCPS was administered in five languages (English, Spanish, Mandarin and Cantonese Chinese, Korean, and Vietnamese), although a measure for language of survey administration was not made publicly available (See HRSA 2016 for detail on language of survey administration). Thus, we use self-reported language fluency as an indirect measure, recognizing that many patients who speak fluent English may also speak another language at home.

Self-rated health also is a plausible mechanism linking race/gender and perceived disrespectful treatment in clinical settings. Race and ethnic disparities in physical and, to a lesser extent, mental health are widely documented. Although patterns vary somewhat across outcomes, whites, Asians, and Hispanics report better self-rated health, fewer health symptoms, and greater longevity relative to their Black and Native American counterparts, with disparities detected among both men and women [30].

Across all racial and ethnic groups, women report higher rates of depression and anxiety, with especially pronounced gaps among whites [31]. Persons with poorer mental health are more likely to report dissatisfaction with their health care encounters, both because depressed affect negatively biases subjective appraisals of interpersonal interactions and persons with compromised mental health are vulnerable to stigmatization [32].

The link between self-rated physical health and perceived disrespectful treatment is less clear and may be bidirectional. Persons who feel they have been treated disrespectfully by health care providers may avoid seeking care until they feel their health problems are severe enough to necessitate a clinical visit. Alternatively, persons with poorer self-rated health may be treated disrespectfully by health care providers and staff because providers believe poor health is a sign of irresponsibility and poor compliance with provider recommendations [33]. Thus, we evaluate the extent to which observed associations between race/gender and disrespectful encounters persist after language, self-rated physical health, and mental health symptoms are controlled.

1.5 Other influences on perceived disrespectful encounters

We adjust all multivariable analyses for demographic and socioeconomic covariates that may confound an observed statistical association between the race/gender categories and our outcomes. Demographic covariates include age/cohort, marital status, and urban/rural health center location. Racial and ethnic groups in the U.S. vary with respect to age, such that older cohorts tend to include larger shares of white persons relative to younger cohorts, given shifts in birth rates, mortality rates, and migration patterns [34]. Reports of perceived discrimination and disrespectful treatment also vary based on age/cohort. Although older adults are vulnerable to ageism, current cohorts of young and midlife adults are more sensitive to mistreatment by service providers, given rising levels of awareness of microaggressions among more recent birth cohorts [35]. Racial and ethnic groups differ with respect to marital status, where Blacks and Native Americans are less likely than whites, Hispanics, and Asians to be currently married, and a higher proportion of men than women tend to be married across all racial and ethnic groups [36]. Marital status may bear on experiences of perceived disrespectful treatment in health care settings, especially among unmarried women of color seeking reproductive health services [37]. Analyses also are adjusted for urban/rural status. Residents of rural areas typically have fewer health care options and cannot easily change their care site if treated disrespectfully, whereas urban residents have more local options [38]

All analyses are further adjusted for educational attainment and poverty status, recognizing the socioeconomic diversity among patients seeking care at FQHCs [25]. Hispanics have higher rates of high school dropout relative to other racial and ethnic groups, and Blacks, Hispanics, and Native Americans have lower rates of college graduation relative to whites and Asians [39]. Blacks, Hispanics, and Native Americans are more than twice as likely as whites and Asians to have household incomes beneath the federal poverty line, with women faring consistently worse than men in all racial/ethnic groups [40]. Persons with lower levels of education and income, in turn, report disrespectful treatment by health care providers at higher rates than their counterparts with greater socioeconomic resources, with these disparities most pronounced among whites [41]. Thus, all multivariable analyses are adjusted for educational attainment, poverty status, age, marital status, and urban/rural status.

2 Methods

2.1 Data

Data are from the Health Center Patient Survey (HCPS), a cross-sectional survey administered in 2014, supported by the Health Resources and Services Administration (HRSA) of the U.S. Department of Health and Human Services. The survey produces cross-sectional, nationally representative data on community-dwelling patients served by health centers funded through Section 330 of the Public Health Service Act, including those supported through four Bureau of Primary Health Care (BPHC) grant programs: Community Health Center (CHC), Migrant Health Center (MHC), Health Center for the Homeless (HCH), and Public Housing Primary Care (PHPC).

The HCPS used a comprehensive three-stage nested probability sampling design.

First-stage sampling units were health center grantees, second-stage sampling units were eligible sites, and third-stage sampling units were eligible patients with at least one visit in the past year to eligible sites. First-stage sampling was stratified by funding stream, patient volume, census region, urban/rural location, and number of sites per grantee. Third-stage sampling involved selecting a random sample of patients for interview when they entered a health center and registered with the receptionist to receive services; 91.4% of those eligible completed interviews.

Data were collected from a total of 169 grantees, 520 health centers, and 7002 patients. The survey was conducted in five languages: Spanish, Chinese (Mandarin and Cantonese), Korean, and Vietnamese. The HCPS oversampled patients who self-identified as American Indian or Alaska Native, Asian, or Native Hawaiian or Pacific Islander, as well as those ages 65 or older. Data collection took place between September 2014 and April 2015. Computer-assisted personal interviews were conducted by trained field interviews and lasted about 50 min. Interview questions were drawn from other federal health surveys including the National Health Interview Survey and National Health and Nutrition Examination Survey (See [42] for detail on sampling scheme and administration). We limit our analytic sample (n = 5385) to persons ages 18 and older, and persons who identify as heterosexual. A full exploration of the ways that sexual orientation intersects with race, ethnicity, and gender is beyond the scope of this analysis.

2.2 Measures

2.2.1 Dependent variables

Respect from health care providers (α = 0.84) is assessed with four questions: In the last 12 months, how often did this doctor or other health professional: (a) listen carefully to you; (b) give you easy to understand information; (c) show respect for what you had to say; and (d) spend enough time with you. Response categories are always, usually, sometimes, never, and don’t know. Less than 0.1 percent responded “don’t know” and these cases were coded system-missing in the publicly available data set. We averaged responses such that higher scores reflect more perceived respectful treatment. Respect from clerks and receptionists is assessed with the single item “In the last 12 months, how often did clerks and receptionists at this health center treat you with courtesy and respect?” Responses were skewed, with 80 percent reporting they were “always” treated with courtesy and respect. We use a dichotomous measure indicating the 20 percent reporting “usually” or less frequent perceived respectful treatment.

2.2.2 Independent variables

We constructed a ten-category measure indicating each respondent’s self-reported gender, race (White, Black, Asian/Pacific Islander, American Indian/Alaskan Native), and Hispanic ethnicity. For ease of interpretation, in our multivariable analyses we present coefficients for each of these categories rather than two-way interaction terms of race/ethnicity by gender. (All results from moderation analyses are available from the authors, and are discussed below).

We consider two potential mechanisms that may account for subgroup differences in perceptions of respectful treatment: language fluency and self-rated health. The HCPS does not collect information on English language fluency; thus, we use an indirect indicator: whether one speaks a language other than English at home. Health encompasses three mental health symptoms and a global measure of self-rated physical health. Respondents indicate how often within the past 30 days they felt anxious, hopeless, and worthless; response categories were all, most, some, a little, and none of the time. Responses were highly skewed, with a small minority reporting “all” or “most” of the time (9.5, 9.6, and 14.4 percent, respectively for worthlessness, hopelessness, and anxious). We constructed dichotomous indicators for each symptom where 1 refers to responses of “most” or “all” of the time (omitted category is “some” or less of the time). Given well-documented gender and race differences in symptoms of anxiety and depression [43], we retain each as a separate measure because an aggregated measure might conceal such differences. Self-rated health refers to whether one rates their overall physical health as excellent (reference group), very good, good, fair or poor. This widely used measure is a robust predictor of mortality [44, 45], and shows strong associations with race, ethnicity, gender, and socioeconomic resources [46].

All multivariable models are adjusted for socioeconomic and demographic covariates. Socioeconomic status encompasses poverty status and educational attainment. Poverty status refers to whether one’s household income is classified as less than 100 percent, 100 to 200 percent, or more than 200 percent (reference group) of the federal poverty line (FPL). The FPL is an economic measure issued annually by the U.S. Department of Health and Human Services to determine income-based eligibility for state-sponsored programs and benefits. We use an annual household income of $25,100 for a family of four as the 100 percent FPL cut point. (For further information, See https://www.healthcare.gov/glossary/federal-poverty-level-fpl/). Education refers to whether one has less than a high school degree, a high school diploma or equivalent, and any education beyond high school (reference group).

Demographic characteristics include age, marital status, and urban/rural residence. Age is recoded into three groups; ages 18 to 44, 45 to 64, and 65 and older, reflecting non-linear associations between age and perceptions of clinical interactions [47]. Marital status refers to whether one is currently married (reference group), divorced/separated, widowed, or never married. Urban/rural status refers to whether the site where one received care is located in an urban or rural (reference group) location.

2.3 Analytic plan

We first present univariate and bivariate analyses for all study measures; we carried out ANOVA to compare all pairs of gender-specific race/gender subgroups and all male–female contrasts for each race/gender subgroup. Contrasting each subgroup with one another, rather than an overall test comparing categories to a single reference group (white men) is consistent with the goals of intersectional analyses. We then estimate multivariable models predicting levels of perceived respectful treatment by health care providers (OLS regression) and the odds of perceived disrespectful treatment by receptionists and clerks (logistic regression). Baseline models evaluate the effects of race/gender categories after adjusting for demographic and socioeconomic covariates, and subsequent models incorporate potential explanatory mechanisms of English language fluency and health. Our supplemental moderation analyses (available from authors) provide a formal test of intersectionality, and reveal which within-sex race differences and within-race gender differences are statistically significant.

3 Results

3.1 Bivariate analysis

Supplemental Table S1 presents descriptive statistics for all variables used in the analysis, for the full sample and each of the ten race/gender subgroups; statistically significant contrasts are presented in the far-left column. Bivariate results for the two dependent variables are presented in Fig. 1a, b. For the four-item perceived respectful treatment by health care providers scale (Fig. 1a), we did not detect statistically significant within-race gender differences, although we did find significant within-gender race differences (p < 0.05). Among men, whites, Blacks, and Hispanics perceived significantly better treatment from their health care providers than did Asians (3.72, 3.70, and 3.71 versus 3.45, respectively), whereas Hispanics—who perceived the highest levels of respectful treatment—also reported significantly better treatment than AIAN men (3.71 vs. 3.45). Asian and AIAN men did not differ from one another, however. Among women, we found generally similar patterns, such that whites, Blacks, and Hispanics perceived significantly better treatment from health care providers relative to their Asian counterparts (3.69, 3.67 and 3.75 vs. 3.46, respectively). Hispanic women, like their male counterparts, perceived the most respectful treatment, with their levels significantly higher than those reported by Asian women.

Fig. 1
figure 1

a Mean score, overall respectful treatment from health care providers (range: 1 = never to 5 = always). b Percentage who perceive that are not always treated with respect by receptionists and clerks at health care centers

Full size image

Results for perceived respectful treatment from clerks and receptionists followed generally similar patterns. A significantly higher proportion of Asian men and women (28 and 35 percent, respectively) perceived poor treatment, relative to their white, Black, and Hispanic counterparts. Although we found no gender differences in reports of perceived disrespectful treatment from health care providers, we find that women are consistently more likely than men to report perceived disrespectful treatment from receptionists and clerks.

Supplemental Table S1 shows no significant subgroup differences for self-rated physical health, although white women are significantly more likely than both white men and women of other racial and ethnic groups to report frequent anxiety. Asian women report the lowest rates of hopelessness, whereas AIAN men report the highest rates of feeling worthless. We detect stark race and ethnic differences in language; more than 85 percent of Hispanic and Asian patients speak a language other than English at home, whereas fewer than five percent of whites and Blacks, and roughly a quarter of AIAN patients did so. AIAN women are significantly more likely than their male counterparts to speak a language other than English at home (28 vs 20 percent, p < 0.05). We find substantial subgroup differences regarding socioeconomic status indicators, consistent with prior studies of gender and racial/ethnic stratification in the United States. Black and Hispanic patients (both men and women) are most likely to live beneath the federal poverty line and least likely to have a high school diploma, although white and Black women are significantly more likely than their male counterparts to have graduated high school.

3.2 Multivariable analysis

Table 1 presents OLS regression results for perceived treatment by health care providers, and Table 2 presents logistic regression results for perceived treatment by clerks and receptionists. The baseline model adjusts for demographic and socioeconomic factors, Model 2 incorporates language fluency, and Model 3 includes health.

Table 1 OLS Regression models predicting overall patient perceptions of respectful treatment from health center providers, HCPS 2014 (N = 5385)
Full size table
Table 2 Logistic regression models predicting patient perceptions of “not always” respectful treatment by clerks and receptionists, HCPS 2014 (N = 5385)
Full size table

3.2.1 Respect from health center care providers

Table 1 shows that patients in the two smallest racial group categories—Asian and AIAN persons—report significantly less respectful treatment from health care providers, relative to whites. We detect no gender differences, regardless of race/ethnicity. Supplemental moderation analyses reveal that these racial differences do not differ significantly in magnitude for men versus women, indicating the primacy of race/ethnicity rather than intersections of race/ethnicity and gender. Further, the two largest racial groups—Blacks and Hispanics—do not differ significantly from whites in their reports of respectful treatment from providers, with similar patterns evidenced for men and women. These disadvantages documented for Asian and Native American patients did not attenuate and increased slightly after adjusting for the potential explanatory mechanisms of language and health.

In the baseline model (Model 1), both Asian men and women report perceived respect scores that are approximately 0.3 points lower than white men (b = -0.26 and -0.27, p < 0.001, respectively) and AIAN men and women report scores that are slightly more than 0.1 point lower (b = -0.15 and -0.14, p < 0.01, respectively). These effects remain statistically significant and increase in magnitude by six (AIAN men) to 22 (Asian women) percent after language is controlled; effect sizes also increase in magnitude, albeit very slightly, after health is adjusted in Model 3. In the fully adjusted model (Model 3), Asian and AIAN patients report perceived respectful treatments scores that are about 0.3 points (0.5 SDs) and 0.16 points (0.25 SDs) lower than that of white men, respectively.

Although the purported mediators, language and health, did not account for observed race/ethnic differences in perceived respectful treatment, several other covariates are significant predictors. Poverty status is inversely and monotonically related to perceived respectful treatment, such that persons beneath (< 100%) or at (100 to 200%) the FPL report respect scores that are significantly lower than persons whose household income is above 200% FPL (b = -0.074, p < 0.001 and b = -0.030, p < 0.05, respectively) in the fully adjusted model. Age is inversely associated with perceived treatment, such that each successive age group/cohort reports more respectful treatment. We also detect urban and rural differences; persons seeking care at urban health centers report significantly poorer quality treatment relative to those in rural areas.

3.2.2 Respect from heath center clerks and receptionists

Table 2 shows that perceived respectful treatment from health center clerks and receptionists is a gendered experience. Women report significantly higher odds of perceived disrespectful treatment relative to men in every race/ethnic group in the baseline model. While Asian and Native American men report significantly higher odds of disrespectful treatment relative to white men, these effects are much smaller in magnitude than those detected for their female counterparts. However, supplemental moderation analyses reveal that the gender effect does not differ significantly on the basis of race/ethnicity.

In the baseline model, White and Black women are 1.5 times as likely and Hispanic women 1.37 times as likely as white men to report disrespectful treatment from clerks and receptionists, although Black, Hispanic, and white men do not differ from one another. Native American women are about 2.5 as likely as white men to perceive disrespectful treatment. Asian women are most likely to perceive disrespectful treatment (OR = 4.33, p < 0.001); Asian men also report elevated odds relative to white men. The magnitude of this effect is smaller than for their female counterparts (OR = 2.89, p < 0.001), although this difference is not statistically significant. After language is adjusted the odds ratio for Asian women decreases substantially (from 4.33 to 3.84) yet remains statistically significant, whereas the effect for Hispanic women is no longer statistically significant. Controlling for language does not appreciably attenuate the Asian-white male disparity, and adjusting for health (Model 3) does not substantially alter the significance or magnitude of other disparities detected.

The effects of several covariates were similar to those in the models predicting perceptions of treatment from health care providers. Persons with poor health perceived less satisfactory treatment with health care center staff, although we found no significant effects of mental health symptoms. Each successive age group is less likely to perceive disrespectful treatment. Poverty status was not related to reports of perceived respectful treatment, although persons with a high school diploma or less had lower odds of reporting perceived disrespectful treatment.

4 Discussion

We explored how race and gender intersect to affect patients’ reports of perceived disrespectful treatment from health care providers and staff in low-income health care settings. Overall, we found marked race/ethnicity and gender differences in our study outcomes yet were surprised to find no statistical evidence of multiplicative race/ethnicity and gender differences. Three main findings emerged from our research. First, Asian American patients are most likely to report disrespectful treatment from health care providers and staff, with women especially vulnerable to the latter. Second, women of all race/ethnic groups are more likely to perceive disrespectful treatment from health center staff, relative to men. Third, Black or Hispanic patients do not fare worse than white patients regarding perceived treatment by health care providers in low-income health care settings, a finding that diverges from an extensive literature on racism in health care in the general population. We discuss the implications of our findings, situating our interpretations within the literature on race and gender disparities in health care encounters.

4.1 Asian patients perceive disrespectful encounters

Asian men and women reported the lowest levels of perceived respectful treatment from health care providers, and the highest odds of infrequent perceived respectful treatment from center receptionists and clerks. Asian women’s perceived disrespectful encounters with staff were partly accounted for by language, yet effects persisted after controlling for language fluency, self-rated health, and other socioeconomic and demographic covariates. Our findings are broadly consistent with studies documenting interpersonal and institutional discrimination against Asian Americans in the United States [48], a pattern that intensified during the COVID-19 pandemic [49]. However, few studies have examined Asian men’s and women’s experiences in health care settings in general, and none to our knowledge focus specifically on FQHCs. One recent study reported that 10 percent of Asian participants in the California Health Interview Survey reported lifetime experiences of health care discrimination, and that these encounters were linked with elevated risk of psychological distress [50]. However, they did not compare Asians with other ethnic groups, or explore the distinctive case of low-income settings.

With the data at hand, we cannot ascertain the reason for Asian patients’ higher levels of perceived disrespectful treatment. Our supplemental analysis of the specific items making up the four-item outcome measure showed that Asian Americans reported the poorest scores on all dimensions, especially the belief that their doctors spent enough time with them. We believe these negative experiences may partly be a function of perceived loss of face that can occur in interactional contexts that involve help-seeking [51]. Face loss, or the perceived loss of prestige, honor, dignity, and respect within interpersonal contexts is an important phenomenon in many Asian cultures [51]. For example, one hospital-based study in China revealed that physicians’ efforts to “save face” in the context of medical decision making prevented them from seeking information from other practitioners, ultimately compromising the quality of care delivered to patients [52]. Our results suggest another way that saving face may bear on patient well-being. Among Asians in the U.S., perceptions of disrespectful treatment and face loss may be intensified in low-income health care settings, in which patients may feel demeaned by providers and staff or embarrassed by the use of public insurance or charity care [53]. In 2014, Asian Americans had the second lowest poverty rate of any ethnic group (12 percent), slightly behind non-Hispanic whites (10 percent), although there is wide variation based on national origin, such that Hmong persons have poverty rates more than twice that of persons of Filipino origin [54].

Asian Americans are stereotyped as a “model minority” that has attained high levels of education and economic security in the United States [23]. Patients who challenge health care providers’ and staff members’ assumptions about Asian American achievement may be treated in a way they perceive to be disrespectful. Further, practitioners in low-income settings may be more accustomed to treating Black and Hispanic patients, who make up a larger share of FQHCs patient populations than do Asians [27].

Enhancing cultural sensitivity training in medical schools and continuing education programs for practitioners and staff at health centers could help to identify aspects of treatment perceived as disrespectful, with particular attention to specific cultural phenomena like “saving face.” Structural changes, such as ensuring that all health care centers have access to interpreters representing a range of languages and dialects also may help to ensure patient comprehension and enhanced communication. Effective interventions should take into consideration perceptual biases among both U.S. and foreign-trained physicians, as the latter are especially likely to be employed in low-income health settings [55]. Foreign-trained physicians account for nearly half of all providers in areas where annual per capita income is beneath $15,000, and roughly one-third of health care providers in communities with mean household income beneath the FPL. While foreign-trained doctors provide critical services in “healthcare deserts,” they also may carry different perceptual biases than their U.S.-born counterparts, thus distinctive educational resources may be necessary.

4.2 Women patients’ encounters with staff members

Counter to our expectations, we found that men and women did not differ with respect to their perceptions of interactions with health care providers. In both bivariate and multivariable analyses, we found no statistically significant within-race gender differences in perceptions that one’s health care provider listened to them and treated them respectfully. However, women were significantly more likely than their men to report that clerks and receptionists did not always treat them in a respectful manner, and this gender gap did not differ significantly across racial/ethnic groups. These gaps did not diminish after mental and physical health were controlled. The HCPS does not obtain information on the nature of interactions between patients and health center staff, so we cannot ascertain the mechanism(s) contributing to these gender gaps. However, given the pivotal role of staff as gatekeepers in health care systems, our results are noteworthy and warrant future exploration.

Most research on microaggressions in health care focuses on the patient-clinician encounter, neglecting other interactions in which patients might perceive disrespectful treatment, such as encounters with desk or telephone clerks, receptionists, or medical assistants [6, 56]. However, these gatekeepers play a crucial role in shaping the quality of a health care encounter, including the amount of time a patient spends waiting, and the dignity with which they are treated when checking in, scheduling appointments, and completing paperwork. Studies based on time diary data show that women (and especially ethnic minority women) spend more time than men waiting yet less time receiving care in health settings [57]. Waiting has been characterized as a violation of respect and dignity as it reflects a power imbalance between the parties waiting and being waited for. Longer wait times may be particularly distressing to lower-income women who face more work and family time constraints than middle-class women, including lost hourly wages [58].

Qualitative studies show that health center staff members’ verbal and nonverbal communication also are related to patient satisfaction with care [56]. Focus group interviews have found that low-income and ethnic minority patients in the southern U.S. were especially likely to report that staff spoke to them in a curt or condescending tone [56]. Patients also reported that staff members’ nonverbal behaviors, like avoiding eye contact, not smiling, and multitasking or shuffling papers when speaking to them were perceived as disrespectful gestures. Although Tajeu et al. [56] did not identify the gender of participants reporting these concerns, we suspect that women would be more sensitive to such gestures, given studies documenting women’s sensitivity to verbal and nonverbal dynamics in interpersonal exchanges [59]. Demeaning or disrespectful encounters with staff may discourage women, especially low-income and ethnic minority women, from seeking care in the future, or following up with provider recommendations. These encounters, in turn, may contribute to widening race, class, or gender disparities in health.

Improving the quality of interactions between staff and patients is an important yet challenging goal. Facilities serving low-income patients are historically under-resourced, so staff behaviors and attitudes may be a consequence of insufficient training and low staffing levels which overburden workers—difficulties that have been exacerbated by the COVID-19 pandemic [60]. Other practices like tying health care provider reimbursement and financial incentives to patient satisfaction also may help to mitigate women patients’ feelings of being disrespected [61]. Researchers increasingly recognize the limits of microlevel interventions like implicit bias training, and call for more systemic, organizational-level interventions to effect meaningful and long-term change [62].

4.3 White, Black, and Hispanic patients do not differ in reports of disrespectful treatment

One of the most surprising findings is that Black and Hispanic patients do not differ significantly from whites with respect to perceived disrespectful treatment from healthcare providers in FQHCs. In stark contrast, Asian and Native American patients report significantly lower levels of perceived respectful treatment, with similar race gaps for both genders. We also found little evidence that Black and Hispanic patients perceived less respectful treatment from health center staff, except for Black women perceiving significantly poorer treatment from staff. These findings diverge from a vast literature documenting perceived racial bias experienced by Black and Hispanic patients [63]. However, few studies have focused explicitly on low-income health care settings, in which health care providers may be sensitive to ensuring respectful treatment of Black and Hispanic patients. Federal data show that Blacks and Hispanic patients are over-represented and white and Asian patients underrepresented among the patient populations seeking care at FQHCs [27]. While Blacks and Hispanics account for 13 and 18 percent of the U.S. population overall, they account for 22 and 36 percent of patients receiving care at FQHCs.

One puzzling finding was that Black women but not men were more likely to report disrespectful treatment from health center staff. This may reflect gender differences in women’s sensitivity to verbal and nonverbal communication [59]. It also may reflect Black men’s motivation to feel and represent themselves as respected, especially when interacting with clerks, receptionists, and other staff members—positions overwhelmingly held by women. Research on Black male respectability underscores that Black men may actively comport themselves in ways to avoid discrimination and racial profiling [64]. However, patients should not bear the burden of altering their behavior to evade disrespectful treatment from staff. Staff training could emphasize how gender and race intersect to shape interactions in health care settings, especially those serving historically marginalized communities. Effective training programs could target components of the cultural health capital model, as articulated by Shim [65], including “cultural skills, verbal and nonverbal competencies, and …interactional styles” to facilitate respectful treatment of patients.

5 Limitations and future directions

Our analyses have several limitations and raise questions to be pursued in future work. First, the HCPS does not assess a patient’s attribution for their perceived disrespectful treatment, although other studies find ethnic minority patients generally attribute discriminatory treatment to their race [56]. Future studies should explore the attributions patients make for disrespectful treatment, as these perceptions may shape their subsequent health-seeking behavior. Second, our data did not include detailed indicators of the interactions patients experienced in the health care setting; future studies should assess factors including wait time, verbal and nonverbal communication, and other behaviors that the patient deemed disrespectful, as this information may inform health center practices.

Third, we could not ascertain the nativity or immigration status of HCPS participants; immigration status affects both access to and the experience of receiving healthcare [66]. Fourth, future studies should explore the intersection of gender, race/ethnicity and sexual orientation given LGBTQ persons’ histories of discrimination and marginalization in the U.S. healthcare system. Fifth, our analyses explained less than 5 percent of the variance in our outcome measures, suggesting that other mechanisms should be considered in future studies of perceived disrespectful treatment in low-income health settings. Finally, our study is based on one time point (2014) and it is unclear whether these different patterns would emerge in other  time periods. Historical events in the early 2020s including the rise of anti-Asian sentiment in the wake of the COVID-19 pandemic, and greater sensitivity to systemic racism following the George Floyd murder and the rise of the Black Lives Matter (BLM) movement may create a context in which Asians perceive even higher levels of disrespect, whereas Black patients experience improvements in their quality of experience. Future studies should explore how racial and gender disparities in perceived respectful treatment change over time.

Despite these limitations, our study documented how race and gender shape perceptions of disrespectful treatment by health care providers and staff at low-income health settings and revealed the vulnerability of low-income Asian patients—an often-overlooked population due to pervasive perceptions of Asians as the “model minority” [23]. We also identified a stark gender disparity in perceived treatment by gatekeepers, such that women across racial and ethnic groups were more likely than men to perceive disrespectful treatment, yet less support for intersectional effects. Our results underscore the urgency of well-designed training programs for health care professionals and staff, and for implementing structural changes in care delivery that advance the goal of improving health equity.