An Examination of Factors Associated with Healthcare Discrimination in Latina Immigrants: The Role of Healthcare Relationships and Language

Abstract

Background

Understanding factors that are associated with perceived discrimination in Latina immigrants may provide opportunities to improve care for this growing population.

Objective

To examine the prevalence of discrimination experiences in urban Latina immigrants and identify socio-cultural and healthcare factors that predict discrimination experiences.

Design

Cross-sectional survey of 166 Latina immigrants.

Measurements

Socio-cultural: region of origin, primary language, and education. Healthcare factors: insurance, place of care, patient-provider communication, trust in provider, and satisfaction with care. Multivariable logistic regression was used to examine factors that predicted discrimination.

Results

42% had at least one discrimination experience. Communication with providers was the factor most strongly associated with reporting having a discrimination experience while controlling for other variables (p<. 01). Women with good communication with their provider were 71% less likely to report discrimination.

Conclusion

Better communication with providers may reduce Latinas’ perceptions of discrimination and thereby improve healthcare access and use of services.

INTRODUCTION

Compared to other racial and ethnic groups, Latina immigrants have lower utilization of preventive screening services such as mammograms.^1–4 This lower use of preventive services may be due in part to barriers in access to care (e.g., insurance).^1,2,4 Beyond structural barriers to care, negative interactions with the healthcare system may prevent Latinas from seeking and using healthcare services.^{5, 6} In a cross-sectional survey, Smyser and Ciske⁷ found that 21% of Latinos living in Southeast Seattle experienced at least one incident of discrimination in a healthcare setting that resulted in their reluctance to seek out needed healthcare. In a follow-up study, the investigators found that 70% of Latinos who had experienced prior discrimination in a health care setting reported that they subsequently delayed seeking needed healthcare in contrast to 48% of Latinos without prior discriminatory experiences. Thus, perceived discrimination may be a barrier to Latina immigrants’ receipt of optimal care. While several studies have examined the association between discrimination and healthcare seeking variables, the measurement of discrimination has not always been specific to health care interactions. Most previous research focused on Latinos from mid-western states or Mexican Americans, making them less comparable to Latinos in other areas in the US.^8–17 Healthcare experiences of Latinos may vary by region and ethnic subgroup; Finch and colleagues reported differences in perceived discrimination among immigrant Mexican Americans.¹⁸ While discrimination experiences have been noted to impact patients’ perceptions of quality care, few studies have examined factors that are associated with an increased likelihood of having a discrimination experience in Latina immigrants.¹⁹

Limited data are available regarding Latinos in the mid-Atlantic region. Latinos in this region differ from many other studied groups in that they are largely uninsured and are from Central and South America.^{20, 21} Prior studies found that discrimination was associated with lower healthcare satisfaction, which in turn influenced mammography adherence. The goal of the current investigation was to identify socio-cultural and process-of-care factors that were correlated with perceived healthcare discrimination in Latina immigrants. A better understanding of factors that are associated with perceived healthcare discrimination may help identify potentially mutable targets for future interventions; thus improving the healthcare experiences of Latinas and thereby positively influencing their access and utilization of preventive health services.

METHODS

Data Source and Setting

Data for this study were collected as part of a general survey of healthcare perceptions of Latina immigrants in the Washington, DC metro area.²⁰ Because this investigation was part of a larger initiative related to breast cancer screening, women who were 40 years and older were eligible for participation. Additionally, women who either attended one of three primary care clinics or who were listeners of a Latino radio health program were invited to participate in the study. The three clinics were part of the Latin American Cancer Research Coalition (LACRC) where most patients are Latino (60% to 90%).²⁰

Procedures

Clinic staff reviewed medical charts and appointment logs to screen women for eligibility prior to scheduled appointments. Women receiving other services at the clinic such as English language classes or social services were also invited to participate by a bicultural and bilingual research assistant who approached them in the waiting area.

The study was also announced on a Latino health radio broadcast over three days to recruit women who were non-clinic attendees. Interested women contacted LACRC offices and were screened by bilingual research staff who confirmed eligibility and scheduled an interview. Written consent was obtained for all participants prior to the interview, which was read aloud to women in their language of preference; 99.4% of the interviews were conducted in Spanish. All participants received a $15 grocery store voucher for their participation. Study procedures were approved by the Georgetown University IRB.

Measures

The outcome variable, healthcare discrimination, was assessed with the Race-Based Experiences scale. The scale queries whether respondents perceive that they had been treated unfairly when accessing healthcare based on their ethnicity using six questions (yes vs. no; e.g. treated with disrespect, treated as if they were not smart, etc.). For the analysis, discrimination was categorized as a dichotomous variable: any discrimination experience (yes to any discrimination item) versus none.²²

Predictor Variables

Three process-of-care items were adapted from measures that have been previously validated in both Spanish and English speaking samples and relate to patients’ general perceptions of their relationship with their providers.²³ Patient-provider communication. Women rated their information received from based on two items: doctor’s explanations of her health problems or treatment and whether or not she left her provider’s office with unanswered questions (four point likert scale; alpha =.70). Trust in providers (one item assessing overall trust in providers) and overall satisfaction with their relationship ranged from 1 to 10. Healthcare-related variables were insurance status (yes vs. no), length of the patient-provider relationship (or at their current place of care), whether respondent attended an LACRC clinic (yes vs. no) and source of health information: Medical entities (e.g., physicians, nurses, hospitals, etc.) versus other (e.g., family, friends, etc.). Demographic variables included region of origin (Central America, South America, or Mexico), education (High School vs. > High School), language spoken at home (Spanish vs. English + Spanish), living arrangement (married/living with a partner or not married/not living with a partner), age, and reported race.

Statistical Analysis

Univariate and bivariate analyses (i.e. Chi-square and t-tests) were performed prior to multivariate modeling. Factors significantly associated with the discrimination outcome from the bivariate analyses were included in the multivariate analysis. Multivariate logistic regression analyses were used to assess the extent to which perceived discrimination was influenced by demographic and other variables. A forward stepwise modeling strategy was employed; variables with significance level of .05 were entered and kept in the model. Odds ratios (OR) with 95% confidence intervals (CI) were used to determine the magnitude of associations between predictors and the outcome. We also report the Akaike Information Criteria (AIC) to compare the three regression models. SPSS version 16 was used for all statistical analyses.²⁴

RESULTS

Sample Characteristics

Participants’ ages ranged from 40 to 77 years (m=51 years; SD = 8.9). Most women were from South America (57%), monolingual (75%), and were either married or living as married (57%). Fifty-one percent had completed high school education or less. Most women (62%) reported their doctor/hospital as a source of medical information (See Table 1).

Table 1.

Sample Characteristics by Rates of Discrimination Experience (N=166)

		Experienced Discrimination
	Total	Yes	No	P-value
	N (%)	%	%
Education (Highest Completed)
≤ High School	86 (51)	32.6	67.4	.02
≥ Some College or More	80 (48)	51.2	48.8
Age				.46
≥ 49	80 (48)	49.3	47.4
≥50	86 (52)	50.7	52.6
Race
White	37 (22)	29.7	70.3	.26
Black	5 (3)	60.0	40.0
Indian	55 (33)	40.0	60.0
Other/No race reported	69 (42)	47.8	52.2
Region of Birth
South America	95 (57)	45.3	54.7	.47
Mexico	19 (11)	31.6	68.4
Central America	52 (31)	38.5	61.5
Language at Home
Monolingual (Spanish Only)	124 (75)	36.3	63.7	.02
Other (English and Spanish)	42 (25)	57.1	42.9
Marital/Partnered Status
Married/Living as Married	95 (57)	42.1	57.9	.87
Single (Divorced/Widowed)	71 (43)	40.8	59.2
Insurance
Yes	71 (43)	50.7	49.3	.04
No	93 (56)	34.4	65.6
Source of Medical Information
Doctor/Hospital	103 (62)	36.9	63.1	.13
Other (Family, Media, etc.)	63 (38)	49.2	50.8
Length of Medical Relationship
≤ 2 years	81 (49)	34.6	65.4	.08
> 2 years	79 (48)	48.1	51.9
LACRC Healthcare Provider*
Yes	97(58.4)	32.0	68.0	.00
No	69(41.6)	55.1	44.9

^*LACRC - Latina American Cancer Research Coalition

Description of Reported Discrimination Experiences

Forty-two percent of the sample reported experiencing discrimination because of their ethnicity when seeking healthcare. The number of reported discrimination experiences ranged from zero to six in the sample (m=1.3; SD=1.9). Table 2 shows the frequency of the specific types of discrimination experiences reported by the women. The most common type of discrimination experienced because of their ethnicity was that providers did not listen to them (27%) followed closely by feeling that their providers treated them discourteously (25%).

Table 2.

Type of Healthcare Discrimination Experienced by Latina Immigrants (N=166)

Number and Percent of Women Having the following type of discrimination when seeking healthcare because of their ethnicity (n=166)	N	%
The doctor did not listen to you?	45	27
Been treated with disrespect?	30	18
Received poor service?	26	16
Staff acted as if they were better than you?	39	24
Treated as if you were not smart	36	22
Discourteous treatment	41	25
Total discrimination experience	70	42

Results of Bivariate Analyses

The results of bivariate analysis in Table 1 showed that women who completed at least some college or higher education reported that they experienced more discrimination than women who completed high school or less (51%; p=. 01). Interestingly, over half of the women (51%; p=. 03) who had insurance also reported that they experienced discrimination. Non-LACRC patients reported a higher percentage of discrimination experiences than did LACRC patients (55% vs. 32%; p=. 003).

Results of Multivariate Analyses

The multivariate logistic regressions that are displayed in Table 3 show that in the first regression model, demographic factors such as education (OR: 2.15; 95% CI, 1.11–4.17; p=. 02) and home language (OR: 2.58; 95% CI, 1.22–5.46; p=. 01) were associated with having a discrimination experience. Women who were more educated and who spoke both Spanish and English were more likely to report having a discrimination experience compared to women with less education or who only spoke Spanish. Insurance status and place of care were added to the second model. Education remained significant but home language was not after controlling for other factors.

Table 3.

Adjusted Odds Ratios (ORs) of Predicting a Discrimination Experience

Variables	Model 1		Model 2		Final model
	OR(95%CI)	P value	OR(95%CI)	P value	OR(95%CI)	P value
Education
>high school	2.15 (1.11–4.17)	.02	2.19 (1.12–4.32)	.02	1.6 (.73–3.55)	.24
≤high school)	Ref		Ref		Ref
Home Language
Spanish/English	2.58 (1.22–5.46)	.01	1.99 (.89–4.39)	.09	2.56 (.1.03–.6.39)	.04
Spanish only	Ref		Ref		Ref
Health Insurance
Yes	--	--	0.61 (0.29–1.24)	.17	.62 (.27–1.4)	.25
No			Ref		Ref
Provider Ethnicity
Non-Latino	--	--	1.85 (.91–3.76)	.09	1.76 (.77–3.97)	.18
Latino (vs. Latino Provider)			Ref		Ref
Healthcare Satisfaction (continuous)	--	--	--	--	.78 (.62–.97)	.03
Trust in Provider (continuous)	--	--	--	--	.89(.71–1.12)	.34
Communication with Providers (continuous)	--	--	--	--	.62(.44–.89)	.00
AIC	204.5		202.2		177.2

In the third and final model, trust in providers, overall satisfaction and communication with providers were added variables. After adding these variables to the model, the effects of education and place of care on discrimination were no longer significant. The effects of home language on discrimination regained significance (p=. 04). Satisfaction with healthcare and communication with providers were statistically significant. Women reporting higher satisfaction reported lower discrimination after controlling for other factors (.78; 95% CI, .62–.97; p=. 03). Additionally, women reporting better communication were less likely to report discrimination after controlling for other factors (OR: .62; 95% CI, .44–.89; p=. 008). Trust in provider, however, was not significant in this model (p=. 34). The AIC’s were computed as 204.5, 202.2 and 177.2, for model 1, model 2 and model 3, respectively, indicating the final model as the best model.

DISCUSSION

Latinas’ perception of discrimination or bias in healthcare has been cited as a barrier to mammography and preventive healthcare services.^25–27 Our data suggest that women’s rating of their communication with providers played a critical role in their perception of discriminatory healthcare experiences. Results suggest that factors relating to the quality of the patient-provider relationship are stronger predictors of the perception of a discriminatory healthcare experience beyond the demographic factors assessed in this sample. These data support other work that found that strong patient-provider relationships that are characterized by good communication during encounters are particularly important in vulnerable populations.^28–31

Interestingly, the specific type of experience most frequently perceived to be discriminatory by the women in the study was that providers did not listen to them because of their ethnicity. Our data suggest that simple strategies of active listening and not talking down to patients may improve interactions between providers and Latina immigrants and reduce perceptions of discrimination. Another approach that may be useful in improving patient-provider communication is to increase providers’ knowledge of cultural values and expectations that may impact their interactions with Latina immigrants. Sensitizing providers to important cultural norms such as ‘personalismo’ which emphasizes interpersonal relations wherein people deal with one another as caring, compassionate persons rather than as impersonal players of specific roles may help improve communication.^34–37 Failure to display or recognize this cultural norm with Latina patients may give the impression that providers are not listening to them. The cultural value of ‘respeto’ emphasizes high regard for authority, which may cause Latina immigrants to be less likely to ask questions than US-born women out of concern for being rude or challenging the authority of their physician.^32,33

Cultural competency training can help providers become more knowledgeable about cultural norms and preferences and thereby increase patient satisfaction with care. For example, Chen and colleagues used a trained in-person interpreter to serve as a cultural educator for clinicians. Over a two-month period, clinicians were introduced to Latino cultural values and practices that could have an impact on communication and physicians’ history-taking. They were also trained to optimize the use of interpreters in improving communication and were taught a few Spanish expressions to help establish rapport with Spanish-speaking patients. Results from the Chen study show that properly designed educational programs for clinicians that increase cultural competency can also increase patient satisfaction with healthcare encounters.³⁸

The fact that Latina immigrants from LACRC clinics reported significantly less experiences of discrimination compared to non-LACRC clinics suggests that services that are tailored for Latino can play a major role in reducing health disparities. However, it is noteworthy that around a third of the patients from LACRC clinics reported discrimination, which calls for the necessity of conducting interventions to improve patient-provider relationships even within the clinics from the LACRC network.

Patient-provider communication, measured as the amount and quality of information received from the providers, was significantly associated with discrimination with women reporting better communication being less likely to report discrimination experiences. Improving the Latinas’ ability to obtain necessary information from providers and encouraging them to increase communication with their providers can also enhance patient-provider interactions. In one study on healthcare communication, Latinos were less likely to report autonomy over healthcare decisions when interacting with their doctors compared with Non-Latino whites.³⁹ Several studies have noted that patients who ask more questions receive more information from their providers, suggesting that provider communication can be activated by patients.⁴⁰ Unfortunately, Latinas face several barriers to effective patient-provider interactions. First, if Latinas are in an ethnically discordant patient-provider relationship they may not be able to effectively communicate with providers in English.^6,41–44 Secondly, Latinas’ high regard for providers may also impact their level of comfort asserting themselves in patient-provider interactions.^{35–37, 40,42} Thirdly, interpreters serve an important role in helping patients understand providers; however, having a third party may in fact perpetuate a more passive role in patient-provider communication. Thus, interventions geared to activate patients and consider the role of interpreters may be useful.

Approaches that provide materials to women in their own language with possible questions to ask their providers may also help encourage participation in their communication encounters. Strategies that build upon natural helpers or peer support may also be useful for women with low literacy.^{45, 46} Findings from a study with Latina breast cancer patients suggested that a peer-based educational intervention improved Latina’s communication skills in talking with their providers about treatment.⁶ Other initiatives such as ‘Ask Me 3’, a national campaign available in English and Spanish, can improve patient-provider communication by encouraging patients to ask their provider key questions (e.g., “What is my main problem?” or “What do I need to do?”). These materials may be particularly useful in settings where low-cost interventions are needed.⁴⁷ Interventions that extend beyond the patient-provider encounter are also necessary. Policies and practices by health systems are also needed to support effective patient-provider communication.⁴⁸

Contrary to our expectation, being monolingual (Spanish only), not having health insurance, and having limited education was associated with reduced likelihood of the perception of a discriminatory health care experience. One explanation is that women who spoke English, had insurance, and had more education were better able to recognize discrimination occurrences or had higher expectations from their healthcare interactions than women who only spoke Spanish, were uninsured, and had less education. It is also possible that the low percentage (25%) of women who reported speaking both English and Spanish at home were learning English and still felt more comfortable talking in Spanish in healthcare encounters. While language barriers have been shown to have an adverse impact on health care outcomes (e.g., utilization, status, receipt of preventive services), we did not find this in our study.^{38, 49} However, Chen and colleagues found that Latinos who did not speak English at home were less likely to receive health care services regardless as to whether or not they were comfortable speaking English.⁴⁹ Assessing English language fluency with more extensive measures and also the language used when talking to providers will be important to consider in future investigations.

Strengths and Limitations

To our knowledge, this is the first study to document predictors of perceived discrimination experiences in a largely South American, female immigrant population in the Mid-Atlantic region. We employed a validated self-reported measure of discrimination to assess perception of ethnicity-based discriminatory healthcare experiences. However, most measures of discrimination have been based on self-report and include constructs captured by our scale. Additionally, we were able to adjust for several demographic characteristics that could potentially predict the relationship between perceived discrimination and health care. The high participation rate and the use of trained bilingual/multicultural staff were also strengths of this study.

Because the data in this study were self-reported, we were unable to directly observe interactions between Latinas and their providers. Communication and discrimination ratings are likely linked because the primary manner that discrimination is experienced within the patient-provider relationship. Future studies that are able to control for this will be beneficial. Additionally, 42% of the study participants were recruited from a Latino radio program: Therefore, our sample may not represent experiences of Latinas not connected to this resource. Our results may also not be generalizable to women under 40. We were also unable to assess information regarding the types of institutions Latinas perceived to be more/less discriminatory. This will be important information for future interventions.

CONCLUSION

Our general findings suggest that a substantial number of foreign born Latino women may have negative experiences in the US healthcare system. Additional research that details the type, timing and frequency of these occurrences will be helpful in fully characterizing Latina health care experiences. Consistent relationships between discrimination and use of health care services have been demonstrated, although mostly in African Americans.⁴⁹ As the fastest growing ethnic group, more attention should be paid to the access and utilization of quality health care services among Latinas. As our findings suggest, good patient-provider communication may be helpful in reducing perceptions of discriminatory health care experience and improve access to and utilization of healthcare services.