Health Equity and Medical Mistrust: A Mixed-Methods Analysis of Medical and Social Determinants Among Transgender Women of Colour in the TURNNT Cohort Study

Abstract

Medical mistrust as a construct often places the onus of blame for adverse health outcomes on individuals rather than on social structures. In this study, we aimed to determine if medical mistreatment and access to transgender care were potential determinants of medical mistrust. We used longitudinal survey data from 193 transgender women of colour living in New York City. We measured medical mistrust using the Group-Based Medical Mistrust (GBMM) scale. Additionally, we analysed and coded open-ended survey data from participants regarding their trust towards medical institutions to identify potential determinants of medical mistrust. From the quantitative analysis, we found that individuals who experienced mistreatment in healthcare and those who reported poor access to transgender care had higher GBMM scores. Qualitative findings suggested that negative experiences within the healthcare system and historical trauma were key factors contributing to mistrust in medical institutions. Addressing medical mistrust should not occur at the individual level, but rather at the structural level. Potential interventions include improving access to gender affirming care and training health professionals.

Introduction

Medical mistrust is defined as mistrust towards medical institutions and personnel, and it may act as a potential cause for lower utilisation of medical services and research participation (Corbie-Smith et al. 2002). Several studies have found that medical mistrust is prevalent among people with multiple marginalised identities, such as transgender women of colour, and has deleterious impacts on health (Ho et al. 2022; Kcomt et al. 2020). This can take the form of low vaccine willingness and uptake (Thompson et al. 2021) or low engagement in routine healthcare (Brenick et al. 2017). Researchers have found that higher levels of perceived medical mistrust are predictive of worse HIV-related care and prevention outcomes (Bogart et al. 2016; Brincks et al. 2019; D’Avanzo et al. 2019). The connection between medical mistrust and health outcomes may be due to the fact that those with high levels of medical mistrust anticipate negative experiences and may not want to engage in care; according to one study that uses data from the 2015 U.S. Transgender Survey, 23% of transgender individuals avoided healthcare due to having mistrust towards medical institutions and anticipating discrimination (Kcomt et al. 2020).

However, researchers have criticised medical mistrust as a construct for being vaguely defined and often placing the onus of responsibility on the individual or the community rather than the structural sources of mistrust (Jaiswal and Halkitis 2019). In this study, we focus on group-based medical mistrust, or the perception of medical institutions acting discriminatorily towards racial groups and/or sexuality and gender minority groups (Cox et al. 2023). Some researchers argue that real and historical structural injustices cause medical mistrust itself (Benkert et al. 2019; Jaiswal, LoSchiavo, and Perlman 2020). As Jaiswal aptly states, “we must rethink this conceptualisation [of medical mistrust], and instead locate mistrust as a phenomenon created by and existing within a system that creates, sustains and reinforces racism, classism, homophobia and transphobia, and stigma” (Jaiswal and Halkitis 2019).

It has been hypothesised that multiple, intersecting, historical forms of oppression may give rise to mistrust towards the medical institution among transgender women of colour. One such source of medical mistrust is the Tuskegee Syphilis Study, a study that unethically used Black Americans as subjects to understand the effects of untreated syphilis, even after penicillin treatment became available (Scharff et al. 2010). Specific to HIV care and prevention, there exists the conspiracy belief that HIV was “created” to destroy Black communities (Bogart et al. 2019). This belief may hinder engagement with medical institutions that can facilitate HIV care. The HIV-related conspiracy may have historical roots as it relates to real discriminatory health practices and human rights violations (Heller 2015). However, the causal roots of medical mistrust need not be historical, as marginalised people today continue to face discrimination by medical personnel. Personal experiences, especially in the healthcare setting, can shape one’s trust in medicine (Hall et al. 2021; Sevelius et al. 2016; Shukla et al. 2025). Understanding the causes of medical mistrust, which may include factors beyond the individual, is necessary to identify interventions to reduce it.

In this mixed-methods study, we sought to understand the determinants of medical mistrust among transgender women of colour in New York City. Home to one of the largest transgender populations in the USA, New York City is a national leader in trans-specific healthcare, making it a critical setting for this analysis. Understanding these factors may allow us to focus on the structural determinants rather than individual and community dispositions. Using mixed methods, we aimed to answer the following three questions:

• How do factors such as access to transgender care and prior healthcare experience affect medical mistrust? For this, we used quantitative data analysis and looked at medical mistrust scores.

• How do transgender women of colour think about the causes of medical mistrust, in their own words? To answer this question, we reviewed and coded qualitative data regarding perceptions of medical mistrust.

• How do medical mistrust scores align, or not align, with perceptions of medical mistrust? We created a joint display to explore this question.

Materials and Methods

Study Population

We used longitudinal data from 193 transgender women of colour from the “Trying to Understand Neighborhoods and Networks Among Transgender Women of Color” (TURNNT) Cohort Study (Callander et al. 2020; Furuya et al. 2024). We recruited this group of transgender women of colour using social media marketing, print advertising, event-based recruitment, referrals from organisational outreach and partnership, and snowball sampling via referrals from enrolled participants. The research team recruited participants from August 31, 2020 until November 04, 2022. Eligible participants identified as transgender women of colour, were aged between 18 and 55 at enrolment, and lived in the New York City metropolitan area (including northern New Jersey, Long Island, and the lower Hudson Valley). Participants spoke either English or Spanish. Participants were compensated with pre-paid gift cards as follows: $50 per interview at enrolment (wave 1), $75 per interview at the 6-month stage (wave 2), and $100 per interview at the 12-month stage (wave 3).

Ethics Statement

At the initial enrolment interview, the research team reviewed with interested individuals a consent form describing the study background, requirements, timeline, confidentiality procedures, and participant rights. Participants gave consent voluntarily by agreeing to the terms of the study and signing the consent form. Participants were also asked to sign an optional release of information form to allow for the sharing of sexually transmitted infection (STI) and HIV testing information with the study team. The study was approved by the Institutional Review Board at Columbia University Irving Medical Center (IRB-AAAS8164).

Mixed-Methods Approach

To identify potential determinants of medical mistrust, we used a mixed-methods approach. Specifically, we integrated quantitative survey data and qualitative open-ended responses to provide a deeper understanding of the issue. Researchers can use mixed-methods approaches to provide complete and corroborated results; if the results from the quantitative and qualitative approaches are similar (method triangulation), then convergence has been achieved, and we can tell a more complementary story (Houghton and Paniagua-Avila 2023; Creswell and Plano Clark 2007). Additionally, using a mixed-methods approach can be thought of as placing participants at the centre of the research process, in that their experiences and knowledge to inform the research process, building trust and collaboration (Guan et al. 2025). In this study, we used a concurrent design, where information from both the qualitative and the quantitative was collected at the same time and integrated during analysis (Fetters et al. 2013).

Quantitative Analysis

Medical Mistreatment

To capture the construct of negative healthcare experiences, we asked participants at baseline, “Thinking back over the past six months, how often did you experience mistreatment in health care settings?” Participants could reply: “Never”, “Sometimes”, “Most of the time”, and “All or almost all of the time”. Given that only three participants replied using the latter two responses, we decided to dichotomise the variable. Those who replied, “Never” were relabelled as “No, I did not experience mistreatment within the past 6 months”, and those who replied “Sometimes”, “Most of the time”, and “All or almost of the time” were relabelled as “Yes, I experienced mistreatment within the past 6 months”.

Perceived Access to Transgender Care

To measure perceived access to transgender care, we asked participants at baseline to rate how well they can access transgender care, as done in prior work (Merriman et al. 2025). We asked participants to respond to the following question: “How would you rate your access to transgender-related care in the past six months? Please think about ‘access’ as your ability to attend these services based on their location, affordability, and so on.” Participants could respond to this question with the following options: “Non-existent”, “Poor”, “Ok”, “Good”, and “Great”. Only two participants responded with “Non-existent”; we grouped these individuals with those who responded “Poor”.

Medical Mistrust

We used the Group-Based Medical Mistrust (GBMM) scale to capture the construct of medical Mistrust (Shelton et al., 2010). At the 12-month follow-up, we administered two versions of the 12-item survey for each participant, one capturing medical mistrust among those of the same racial/ethnic group (GBMM-Race) and the other capturing medical mistrust among those of the same gender identity (GBMM-Gender). We used outcome data from 12-month follow-up to have temporality in the analysis and to reduce the possibility of reverse causation. The exact questions can be seen in supplemental online Tables S1 and S2. Participants responded to each question using a 4-item Likert Scale (“Strongly Disagree”, “Disagree”, “Agree”, and “Strongly Agree”). We converted each response to a numerical value (“Strongly Disagree”=1, “Disagree”=2, “Agree”=3, and “Strongly Agree”=4); we reverse-coded questions 2, 8, 10, and 11 so that higher scores corresponded to a higher sense of medical mistrust. For each individual, we calculated their mean GBMM-Race and GBMM-Gender score. Individuals who did not respond to half or more of the questions were coded as missing. We also created a composite GBMM score (GBMM-Composite) that takes the average of the GBMM-Race and GBMM-Gender scores. While these scales have previously been validated among Black and Latina women and other minoritised individuals (Martinez et al. 2022; Shelton et al. 2010; Thompson et al. 2004), the scale has not yet been validated among transgender women of colour. Therefore, we calculated the Cronbach’s alpha and conducted confirmatory factor analysis to verify the reliability of our measures.

Covariates

Based on the literature and our theoretical causal model, we anticipated several variables that might act as confounders in the relationship between our exposures and the outcome of interest. We used baseline self-report measures to capture the following characteristics: annual income (“No income,” “$1 to $9,999,” “$10,000 to $29,999,” “>$30,000,” and “Missing”), education (“Less than High School,” “High School Graduate or Equivalent,” and “More than High School”), age category (“18–24”, “25–34”, “35–44”, and “45 and older”), and incarceration history (“Ever incarcerated,” and “Never incarcerated”). Controlling for these potential confounders would allow us to get a less biased effect of medical mistreatment and perceived access to transgender care on GBMM scores.

Statistical Analysis

Descriptive analyses were performed to characterise the socio-demographic composition of the study sample. We conducted a two-sample t-test to compare the mean GBMM scores between those who experienced a negative healthcare experience and those who did not. We conducted an ANOVA to see if there were any differences in GBMM scores by access to transgender care ratings. In addition, we assessed the relationship between our exposures and outcome, controlling for confounders; to do this, we created modified Poisson regression models with robust standard errors. Using these regression models allowed us to estimate the adjusted prevalence ratio even when the outcome is common (Zou 2004). We controlled for age, income, education, and incarceration history. All quantitative analyses were conducted using R/R Studio (Version 2023.03.1, +446).

Qualitative Analysis

After providing survey responses for the GBMM scale, participants were invited to add commentary regarding medical mistrust. We asked, “Is there anything you would like to add or share about the above statements?” We compiled all responses and analysed them using Framework Method, which allows us to systematically organise, code, and interpret written comments (Gale et al. 2013). First, the study team read through all the responses and developed their individual codes. We then compiled a preliminary codebook that included detailed definitions for each code. One coder then went through and coded all the transcripts, and another coder verified these codes. We reviewed as a team and reached consensus on any disputed coding. We developed themes based on the saliency and frequency of codes, and we selected quotes to highlight the significant themes.

Integrated Analysis

We used a joint display to triangulate quantitative and qualitative measures of medical mistrust.

Joint display allowed us to integrate quantitative and qualitative data to produce insights above and beyond results from individual analyses; in other words, the integrated analysis is greater than the sum of the individual analyses (Guetterman et al. 2015). Specifically, for each qualitative theme of medical mistrust, we classified the speaker of the coded text by medical mistrust score (using a median cutoff of the GBMM-Composite score). Through narrative discussion, we described how the quantitative and qualitative results converged or diverged from each other.

Results

In our sample of transgender women of colour, 24.4% of the participants identified as Black, 47.7% identified as Latina, 23.3% identified as Multiracial, and 4.6% identified as Asian, Middle Eastern/North African, or other. The mean age at baseline was 38 years. The demographic characteristics of the study population are shown in Table 1. Among participants, 28.0% reported having experienced healthcare mistreatment at baseline. In terms of access to transgender care, 7.8% of participants rated their access as “Poor”, 17.6% as “Ok”, 37.3% as “Good”, and 35.8% as “Great”.

Table 1.

Characteristics of the study population.

	Quantitative Sample (n=193)		Qualitative Sample (n=57)
	N	%	N	%
Age at Baseline
18–24	14	7.3	2	3.5
25–34	55	28.5	17	29.8
35–44	72	37.3	22	38.6
45–55	49	25.4	16	28.1
Missing	3	1.5	0	0
Race/Ethnicity
Black	47	24.4	13	22.8
Latina	92	47.7	21	36.8
Asian	5	2.6	2	3.5
Middle Eastern/North African	2	1.0	1	1.8
Multiracial	45	23.3	20	35.1
Other	2	1.0	0	0
Education
Less than High School	63	32.6	21	36.8
High School	54	28.0	14	24.5
More than High School	76	39.4	22	38.6
Income Status
$0	27	14.0	5	8.8
$1 to $9,999	75	38.9	20	35.1
$10,000 to $29,999	44	22.8	15	26.3
$30,000 to $49,999	14	7.3	4	7.0
$50,000 or more	12	6.2	7	12.3
Missing	21	10.9	6	10.5
Incarceration History
No, never	139	72.0	39	68.4
Yes	53	27.5	17	29.8
Missing	1	0.5	1	1.8
HIV Status
Living with HIV	99	51.3	28	49.1
Not Living with HIV	90	46.6	26	45.6
Missing	4	2.1	3	5.3
Experienced Healthcare Mistreatment in Past 6 Months
Yes	54	28.0	17	29.8
No	136	70.4	40	70.2
Missing	3	1.6	0	0
Transgender Care Access Rating
Non-Existent/Poor	15	7.8	3	5.3
Ok	34	17.6	12	21.1
Good	72	37.3	15	26.3
Great	69	35.8	26	45.6
Missing	3	1.6	1	1.8
Mean Gender-Based Medical Mistrust Score [Mean (SD)]	2.43 (0.45)		2.52 (0.51)
Mean Race-Based Medical Mistrust Score [Mean (SD)]	2.47 (0.47)		2.62 (0.51)
Mean Composite Medical Mistrust Score [Mean (SD)]	2.45 (0.47)		2.57 (0.47)

Note: This table describes the demographic, socioeconomic, and sexual health characteristics of the Trying to Understand Neighborhoods and Networks Among Transgender Women of Color (TURNNT) cohort.

Quantitative Results

The mean (SD) GBMM-Race score was 2.42 (0.46), the mean (SD) GBMM-Gender score was 2.47 (0.47), and the mean (SD) GBMM-Composite score was 2.45 (0.47). We found that the Cronbach’s alpha for GBMM-Race was 0.91, and for GBMM-Gender it was 0.89. When we conducted confirmatory factor analysis with the two factors, we found that the Comparative Fit Index was 0.969 and the Tucker-Lewis Index was 0.966; these are greater than 0.95, suggesting good model fit. We found that the correlation between GBMM-Gender and GBMM-Race was 0.95. Results from the structural equation modelling are shown in online supplemental figure Figure 1.

We identified a significant association between medical mistreatment and GBMM scores, with individuals who experienced mistreatment in healthcare reporting higher GBMM-Race, GBMM-Gender, and GBMM-Composite scores compared to those who did not (Figure 1). We found that individuals who experienced mistreatment in healthcare, compared to those who did not, reported higher GBMM-Race (Mean: 2.67 vs. 2.34), GBMM-Gender (Mean: 2.74 vs. 2.37), and GBMM-Composite (Mean: 2.70 vs. 2.36) scores. The mean differences for all GBMM scores were statistically significant at the level of significance of 0.05. We observed an inverse relationship between access to transgender-specific care and race-based medical mistrust. Participants who reported poorer access to transgender care had significantly higher GBMM-Race scores, whereas those with better access exhibited lower GBMM-Race scores (Figure 2). Compared to those who reported great access to transgender care, participants who reported poor access to transgender care had significantly higher GBMM-Race scores (Mean: 2.70 vs. 2.40), GBMM-Gender scores (Mean: 2.74 vs. 2.46), and GBMM-Composite scores (Mean: 2.72 vs. 2.42). The ANOVA result for GBMM-Race was statistically significant, but not for GBMM-Gender and GBMM-Composite.

Figure 1. Boxplot of Medical Mistrust Scores by Medical Mistreatment.

Note: Boxplot of Medical Mistrust Scores by Medical Mistreatment. We plotted the distribution of race-based medical mistrust scores (GBMM-Race), gender-based medical mistrust scores (GBMM-Gender), and the combined medical mistrust score (GBMM-Composite) measured at the 12-month follow-up visit, stratified by whether or not an individual experienced medical mistreatment at baseline. Sample sizes for each box plot are listed.

Figure 2. Boxplot of Group-Based Medical Mistrust Scores by Transgender Care Access Rating.

Note: Boxplot of Group-Based Medical Mistrust Scores by Transgender Care Access Rating. We plotted the distribution of race-based medical mistrust scores (GBMM-Race), gender-based medical mistrust scores (GBMM-Gender), and the combined medical mistrust score (GBMM-Composite) measured at the 12-month follow-up visit, stratified by how they rated their access to transgender care at baseline. Sample sizes for each box plot are listed.

The results from the multivariable regression models can be found in Table 2. All observed associations remained statistically significant after adjusting for potential confounders, including age, income, education, and incarceration history. Participants who reported experiencing medical mistreatment had GBMM-Race scores that were 0.27 points higher (95% CI: [0.12, 0.42]), GBMM-Gender scores that were 0.31 points higher (95% CI: [0.16, 0.47]), and GBMM-Composite scores that were 0.29 points higher (95% CI: [0.15, 0.43]) compared to those who did not report mistreatment. Additionally, individuals who reported good access transgender-specific care had GBMM-Race scores that were 0.32 lower (95% CI: [0.07, 0.57]), GBMM-Gender scores that were 0.35 lower (95% CI: [0.09, 0.61]), and GBMM-Composite scores that were 0.33 lower (95% CI: [0.09, 0.57]) compared to those with poor access, after adjusting for confounders.

Table 2.

Effects of Medical Mistreatment and Transgender Care Access on Medical Mistrust.

Adjusted Effect of Medical Mistreatment on Medical Mistrust
Scores	Beta [95% CI]		N
	No	Yes
GBMM-Race	Ref.	0.27 [0.12, 0.42]	180
GBMM-Gender	Ref.	0.31 [0.16, 0.47]	182
GBMM-Composite	Ref.	0.29 [0.15, 0.43]	178

Adjusted Effect of Transgender Care Access Rating on Medical Mistrust
Scores	Beta [95% CI]				N
	Poor	Ok	Good	Great
GBMM-Race	Ref.	−0.22 [−0.50, 0.06]	−0.32 [−0.57, −0.07]	−0.32 [−0.55, −0.05]	181
GBMM-Gender	Ref.	−0.31 [−0.61, −0.02]	−0.35 [−0.61, −0.09]	−0.30 [−0.56, −0.04]	183
GBMM-Composite	Ref.	−0.25 [−0.52, 0.03]	−0.33 [−0.57, −0.09]	−0.31 [−0.55, −0.06]	178

Note: Using multivariable regression models, we estimated the effect of baseline medical mistreatment and transgender care access on race-based medical mistrust scores (GBMM-Race), gender-based medical mistrust scores (GBMM-Gender), and the combined medical mistrust score (GBMM-Composite) measured at the 12-month follow-up visit, adjusting for baseline age, income, education, and incarceration history.

Qualitative findings

We found that the characteristics of those who responded and those who did not were similar, though those who responded tended to have higher medical mistrust scores (Table 1). Several key themes arose, including personal experiences, recognition of structural determinants, personal advocacy, and the need for training for care providers.

Personal Experiences Inform Levels of Medical Trust

Several study participants stated that their levels of trust towards the medical institution were based on their own personal experiences. Some respondents reported having negative experiences when interfacing directly with healthcare providers. Misgendering and using dead-names were frequently mentioned in these cases, as one participant stated, “People of gender have to fight for the name you want to be called instead of being called by your government name…. That’s really annoying.” Another respondent recounted a time when the healthcare system failed to provide gender-affirming care:

“My orchiectomy consultation had the hospital somehow register me under my deadname even though I set up the appointment under my legal name. That caused me to be deadnamed in notes and have a 2-hour conflict with the hospital information system department to correct the mistake. That has now slowed down my transition process because I’m not seeing that doctor and need to find a new one. Additionally, when I was institutionalised for inpatient mental healthcare, I was misgendered and asked inappropriate questions about my gender by healthcare providers, workers, and other patients.”

Another participant had experience working in the health care setting as a scribe. Here, she recounts an experience where she witnessed explicit discrimination firsthand:

“I remember one specific time with a [patient]. [The care provider] misgendered them. Behind the scenes when we got back to a disclosed area, she said, ‘I don’t know why “he” doesn’t go somewhere else that knows about these things such as Callen-Lorde¹ or somewhere that is LGBT.’”

Transgender women of colour encounter discrimination beyond racism and transphobia, in that they face unique, intersecting forms of discrimination. One participant stated, “I experienced mistreatment by ambulance workers at the beginning of COVID-19 due to my race and gender.”

Historically, there has been a stereotype that all transgender women of colour are sex workers. One participant recounted how this stereotype permeated her quality of care:

“I’ve experienced a doctor that judged me because of my sexual orientation and classified me as a sex worker and treated me as such instead of as a person and when anything came up, it was always revolving around sex, when it wasn’t revolving around my pain it was never about my wellbeing. I felt so unlike a person, I felt ugh it was disgusting for them to treat me like that over and over again and to never listen to what I had to say unless it had to do with sex or she was looking at my body, so I had to get a new doctor to be treated like a person. She made me feel like I was a guinea pig all the time, so I had to leave her because I couldn’t eat or sleep knowing I had to go back and see her every month to get my medicine. It messed me up.”

Other participants reported having a positive experience. One participant reported having an empowering experience throughout their process of transitioning:

“I honestly feel that healthcare workers and the medical providers that I have experience with, I feel that they are great. I feel they did their job to the best of their knowledge, and I’ve never really had a bad experience in the healthcare system because of my race or gender or anything like that. My doctors and my nurses were very respectful of my pronouns when I changed it, so I really liked that.”

Yet others responded that the kind of care they receive depended on doctors, nurses, and health care systems. As one participant stated, “Based on my personal experience I have been treated fairly, I can not speak for others as I have heard they’ve had bad experiences.” These experiences may affect one’s level of mistrust towards the medical institution.

Recognition of Historical and Structural Determinants

Many of the women in the study stated that their levels of mistrust towards the medical institution went beyond mistrusting individuals and attributed their mistrust to historical medical institutional violations of marginalised groups.

“The US Healthcare system is a for-profit industry with a history of abuse and mistreatment directed towards minorities. So, while I personally trust my doctor, I think a healthy amount of scepticism is a good thing especially if you are a trans woman of colour.”

Other participants recognised that institutional and medical knowledge surrounding transgender care for women of colour was still nascent.

“I think it’s a systemic issue; it’s less about individual providers but they do the things that they do because of a trickle-down effect. They’re just following the book, it’s a problem with the system as a whole. Many tests are developed based on white and cis bodies.”

Similarly, another participant noted that this research gap needs to be addressed. Specifically, researchers often focus on negative and not on the positive health.

“There needs to be a study for the study. We also have to study people like myself who may be on Medicaid or public assistance but don’t experience trauma, but don’t experience these instances that actually happen. There needs to be a little something more. People who are doing well as well people who are struggling and then we will have a study.”

Empowered Responses

Because of the limitations of the care that they received, many participants reported that they have had to take their own action to navigate the healthcare landscape in order to receive quality transgender care. According to one participant, “I feel like you have to take control of your own health and make sure your medical care with that provider is taken seriously.”

Some participants took the extra effort to educate their care providers.

“I make my doctor do more intensive studying on trans health and my doctor’s office has come to understand the trans community a lot better because there are a few people who had to speak up for themselves. Me personally, I’m just the type of person that if you don’t know what you’re talking about when it comes to my healthcare, then you don’t need to be here. That’s what I said to my doctor, therapist, psychiatrist and that is the reason they have done more for the trans community than most doctors’ offices have done.

Potential Interventions

Participants suggested ways to improve trust in medical institutions. Many recommendations had to do with improving the quality of transgender care and reducing care provider bias. As one participant stated, “I just feel that sometimes the medical profession discriminates, they have stigma too, it’s good to teach them or have workshops so they can get proper training on different races or ethnicities.” Another participant believed that transgender care needed to be holistic and go beyond biomedicine.

“Many doctors think that hormones is all a trans person needs. They pump you up with hormones, and don’t care about the other parts of being trans. The most important thing is to be accepted by society and our family.”

Integrated Results

Online supplemental Table S3 shows the relationship between the quantitative measures and qualitative measures of medical mistrust, stratified by qualitative theme. Under the personal experiences and medical mistrust theme, we found that those with higher composite medical mistrust scores tended to have experienced discrimination firsthand in the medical setting, whereas those with low medical mistrust scores tended to report positive first-hand experiences, with one exception (see Row 16). Additionally, almost all those who recognised historical and structural sources of medical mistrust had high medical mistrust scores, and the one person who scored low spoke of implicit bias, which can be argued to be more of an individual-level source of mistrust.

Participants described empowered responses regardless of medical mistrust score. Finally, in terms of potential intervention to address medical mistrust, those with a higher medical mistrust score offered solutions that were focused on clinicians getting more training in gender-affirming care; those with a lower medical mistrust score recognised the structural barriers to receiving gender-affirming care. This might indicate that those with higher medical mistrust scores may focus on more direct, proximal interventions compared to those with lower medical mistrust scores, who may focus on upstream factors.

In terms of the converging and diverging quantitative and qualitative data, the findings converged in relation to experiences of discrimination, aligning with high scores of medical distrust, and positive experiences aligning with lower medical mistrust scores. The qualitative and quantitative results also converged in terms of historical and structural determinants and offered potential solutions in that the qualitative responses differed between those with high and low scores. This convergence confirms that the quantitative measure was a good measure of mistrust in this population. In contrast, participants had similar empowered responses regardless of medical mistrust score, suggesting that this population had other sources of empowerment.

Discussion

Often, the key question around medical mistrust is: how do we get marginalised individuals to trust medical institutions? However, in this study we wanted to ask the question that Allen et al. asked: namely, are medical institutions trustworthy? (Allen et al. 2022) Throughout this study, we framed medical mistrust not as an individual issue but as a structural concern. We found that, for transgender women of colour, medical mistrust was rooted in personal and historical experiences of discrimination with medical institutions. Using an integrated approach, we found that the quantitative and the qualitative data mostly converge to this conclusion.

From the quantitative analysis, we found that transgender women of colour who experienced medical mistreatment and had less access to transgender care were more likely to have more medical mistrust. From the qualitative analysis, participants recounted personal experiences of discrimination and historical events that gave rise to mistrust towards medical institutions; as a response, some participants took it upon themselves to educate their health providers. Our results align with those of other researchers. In their concept analysis, Shukla et al. (2025) found that causes of medical mistrust include historical trauma and personal experiences. Hall et al.’s (2021) qualitative study investigating the role of stigma and medical mistrust among US people living with HIV diagnosed with COVID-19 found that those who had negative experiences while receiving HIV care had more mistrust that hindered COVID-19 testing. Other researchers have similarly found that medical mistrust is associated with experiences of transphobia and racism (Sevelius et al. 2016).

In their open-ended responses, participants in this study suggested that change needs to happen to increase trust. First, there is a need to improve medical treatment for transgender women of colour. Concretely, this could mean helping healthcare providers to have more cultural and medical competence to care for transgender patients (White Hughto et al. 2015). One possible manifestation of this is gender-affirming care, which is a specific kind of care for transgender individuals that addresses their health needs in a way that affirms their gender identity (AAMC 2022). Gender-affirming care extends to every part of the medical care experience and can include the use of preferred pronouns, the availability of gender-affirming hormone therapy, and having providers who are also transgender themselves (Loo et al. 2021). Gender-affirming care not only decreases medical mistrust but also can directly improve the mental health of transgender patients (Jarrett et al. 2021). However, there is a long road ahead. Medical schools, schools of nursing s, and residency programmes often do not provide sufficient training to students on how to provide gender-affirming care, and many health providers do not feel confident in providing gender-affirming care (Bhatt, Cannella, and Gentile 2022; Milionis and Koukkou 2023; Vasudevan et al. 2022). Therefore, we advocate for structural change to these programmes to include training on how to provide care that is affirming and non-stigmatising. Second, as one participant suggested, there needs to be greater recognition that the experiences among transgender women of colour are diverse. Medical professionals and academic researchers alike need to recognise that not all transgender women of colour have the same experiences; research that explicitly recognises this and explores this may be beneficial to the public health literature.

Medical mistrust is a public health issue. Researchers have cited medical mistrust as a potential barrier to care among transgender women of colour, particularly for HIV care and prevention outcomes. For example, D’Avanzo et al. found that US transgender women of colour who have higher levels of medical mistrust also had less awareness of PrEP (D’Avanzo et al. 2019). However, other researchers have found a null relationship. Bogart et al. (2019) found that, while Black people in the USA may have higher mistrust towards HIV care, this does not correlate with worse prevention behaviour. In fact, El-Krab et al. (2023) found in their systematic review of 17 studies that there is contradicting evidence connecting medical mistrust and HIV care and prevention outcomes; some suggest a significant negative association, while others were null. This situation is further complicated by the fact that there may be confounders, such as discrimination, that are associated with medical mistrust and HIV care and prevention outcomes.

Future research should continue to examine the determinants of medical mistrust. We encourage researchers to approach medical mistrust not as an individual-level issue, but rather as a structural concern. Applying a historical and critical lens to medical mistrust is important to identify the sources of medical mistrust. Finally, we need more research to identify interventions to reduce medical mistrust among transgender women of colour. Again, we frame this using a structural perspective; this means asking the question: How do we make medical institutions trustworthy?

Limitations

We note several limitations in our study. In terms of our qualitative analysis, we only analysed data for those who provided any response. Those who provided qualitative data in this study may differ from those who did not. When we compared their demographic characteristics, however, we found that they were similar. Nevertheless, they may differ by some unmeasured factor. In our quantitative analysis, we controlled for income, education, age, and incarceration history, but there may be other unmeasured confounders that may bias our results.

Additionally, the TURNNT cohort used convenience sampling methods to form the cohort and, which may not be representative of the wider population of transgender women of colour in New York City. While thie may create a problem for external validity, obtaining a representative sampling of transgender women of colour is challenging due to structural barriers and issues with sampling coverage. The primary study objective of this study was to optimise internal validity. We note that the TURNNT cohort is one of the largest study populations of transgender women of colour.

Conclusion

Medical mistrust can undermine HIV care and prevention outcomes among transgender women of colour, contributing to and exacerbating health disparities. However, this mistrust is often rooted in negative experiences with medical institutions. As participants in this study emphasised, training medical providers in gender-affirming care is essential. By addressing medical mistreatment and discrimination, we can reduce medical mistrust, improve health engagement, and advance health equity. More research is needed to identify ways to make medical institutions trustworthy.

Data availability

The data that support the findings of this study may be made available by the corresponding author, upon reasonable request.