Comparing Preferences for Depression and Diabetes Treatment among Adults of Different Racial and Ethnic Groups Who Reported Discrimination in Health Care [Internet]

Excerpt

Background: Racial/ethnic differences in the treatment of depression and diabetes may be explained by differences in patient preferences and the influence of Black and Hispanic/Latino patients' past experiences of discrimination.

Aim 1: Develop and administer a national survey to measure treatment preferences, how preferences vary by race/ethnicity, and how preferences are associated with experiences of prior health care discrimination.

Aim 2: Interview survey participants to understand how prior health care discrimination influences treatment preferences and receipt of preferred treatment.

Aim 3: Interview clinician and health care administration stakeholders to (1) assess how providers typically elicit and incorporate information about patient treatment preferences and prior discrimination into treatment plans and (2) determine the potential utility of collecting more structured data to improve the tailoring of treatment plans.

Methods: This sequential, explanatory, mixed-methods study identified national rates of health care discrimination and elicited patient preferences for treatment of depression and diabetes among White, Black, and Hispanic/Latino individuals with depression or diabetes. Themes inferred from quantitative data were explored and contextualized through in-depth interviews with patients and providers.

In aim 1, we developed and administered a nationally representative survey among Black (n = 505), Hispanic/Latino (n = 504), and White (n = 503) community-dwelling adults (N = 1512) with depression, type 2 diabetes, or both. The purpose of this survey was to assess racial/ethnic differences in patient preferences for treatment of depression and diabetes, to measure rates of health care discrimination by race/ethnicity and gender, and to assess whether treatment preferences differed for those with prior experiences of discrimination. The survey instrument was developed through a community-based participatory process and, upon completion, was administered in 2 major parts: a discrete choice experiment (DCE) module to elicit treatment preferences (treatment type, provider language, trustworthiness, time to provider location, and cost) and standardized survey questions about past experiences of discrimination in health care, mental health status, quality of health care, and other sociodemographic characteristics. We analyzed data using conditional logit regression models to estimate racial/ethnic variation in treatment preferences and to understand whether preferences were associated with prior health care discrimination.

In aim 2, we conducted semistructured, qualitative follow-up interviews with 40 individuals who participated in the aim 1 survey (21 with depression [8 Black, 4 Hispanic/Latino, 9 White] and 19 with diabetes [9 Black, 5 Hispanic/Latino, 4 White, 1 multiracial]) to understand respondents' reported preferences, how discrimination shaped those preferences, and whether and how prior health care discrimination interfered with preferences elicitation or obtaining preferred treatment.

In aim 3, we interviewed 20 clinician stakeholders to understand how providers elicit patients' treatment preferences and asked about past health care discrimination and their openness to using routinely collected data on preferences and health care discrimination to improve treatment planning.

For aims 2 and 3, we transcribed and analyzed interviews using a thematic analysis approach in the Dedoose application. We triangulated findings across all aims to suggest enhanced shared decision-making (SDM) guidelines for patients from marginalized backgrounds.

Results: In aim 1, Black and Hispanic/Latino respondents were significantly more likely to face health care discrimination compared with White respondents in both diagnosis groups. Among the entire group of individuals with depression, Black and Hispanic/Latino respondents did not have a significant preference for 1 treatment modality (medication vs talk therapy), but the subgroup of respondents reporting past health care discrimination had a greater preference for medication vs talk therapy. Among those with type 2 diabetes, past experiences of health care discrimination were associated with respondents having preferences for behavioral modification vs medication (Black and White respondents only).

In aim 2, few participants with depression reported being asked outright about treatment preferences but were typically open to depression treatments that differed from their preferred ones if suggested by a trustworthy provider. Experiences with discrimination in health care led to difficulties in establishing trust and SDM. Analyses of participants with diabetes yielded similar themes.

In aim 3, clinicians reported varied strategies for eliciting patient preferences and no systematic approaches to starting conversations about past health care discrimination. Providers saw potential value in more systematically eliciting treatment preferences and asking about past discrimination but were concerned about the feasibility of data collection and designing appropriate system-level responses to past health care discrimination.

Conclusions: Black and Hispanic/Latino respondents with depression did not have a strong preference between treatment modalities for depression or diabetes, but past health care discrimination was associated with preferring medication over talk therapy for depression and with preferring behavioral modification over medication-only treatment for diabetes. Qualitative results suggest that SDM within the context of a trusted provider relationship can help better elicit and shape treatment preferences and may be key for patient engagement and retention. Providers' acknowledgment of the potential value of eliciting patient preferences and discrimination experiences suggests that the DCE and survey instrument developed in this project have the potential to identify gaps and opportunities to build patient-provider trust and improve treatment plans for marginalized patients.

Limitations: Surveys and interviews were conducted with White, Black, and Hispanic/Latino respondents in the United States only, limiting generalizability to other groups. In aim 1, unobserved factors that were not identified in the survey likely underlay discrimination, and the results showing an association between prior discrimination and preferences for treatment of depression and diabetes cannot be interpreted causally. In aim 2, participants were mostly female, and, while equally balanced by race/ethnicity, no Black men with depression participated in interviews. Aim 3 clinician interviews were limited to a single safety-net institution in New England.