Correlates of Medical Mistrust Among Minority Women at Risk for HIV and Their Networks

Abstract

Background

Medical mistrust is a barrier to engaging in HIV prevention and treatment, including testing and adherence to antiretroviral therapy. Research often focuses on how race and experiences of discrimination relate to medical mistrust, overlooking the role that other characteristics may play (e.g., history of physical abuse, diagnosis of mental illness). Furthermore, studies are often restricted to samples of men who have sex with men and findings may not generalize to other at-risk groups.

Aims

The current study explores a range of demographic, cognitive, behavioral, and social network correlates of medical mistrust.

Method

This study employed an egocentric network design among a racially diverse sample of at-risk women and women in their social networks (n = 165).

Results

Results from multivariable linear regressions stratified by race (Black vs. others) indicate that medical mistrust is associated with both individual-level and network-level characteristics. Across both groups, age and experiences of racial discrimination were associated with higher medical mistrust. Having a regular sex partner and having a higher proportion of network members who are family was significantly associated with medical mistrust among non-Black women.

Discussion

Individual-level and network-level variables were significantly associated with medical mistrust. Therefore, interventions that attempt to mitigate medical mistrust as a barrier to HIV prevention and treatment should consider how mistrust may be related to characteristics of individuals and broader contexts.

Conclusion

Health interventions may benefit from conceiving of medical mistrust as a complex, rational response to cumulative discriminatory life experiences and a reflection of the networks within which individuals are embedded.

Keywords

Black/African American women discrimination HIV social networks

Racial/ethnic minority women in the United States are disproportionately affected by HIV (Centers for Disease Control and Prevention [CDC], 2019b). The rate of HIV infection among Black/African American women (BAAW) is 15 times the rate of White women (24.2 vs. 1.6 per 100,000 in 2017; CDC, 2019a) and the rate for Latinx women is three times the rate of their White counterparts (4.9 vs. 1.6 per 100,000; CDC, 2019a). For people of color, one barrier to HIV prevention, testing, and treatment is medical mistrust (MM; Ball et al., 2013; Bogart et al., 2016, 2019). Research on MM often explores its historical roots and downstream consequences (Brandon et al., 2005), while little has focused on more proximate determinants. Furthermore, MM is studied as an individual trait, overlooking how it might be shaped by broader social contexts (see Bogart et al., 2016, 2019, for notable exceptions). Finally, when MM is studied as a barrier to HIV prevention and treatment, it is often investigated among men who have sex with men (MSM; Bogart et al., 2011; Cahill et al., 2017). Considering these gaps, this study examines the demographic, psychosocial, behavioral, and social network correlates of MM among a racially diverse sample of women at risk for HIV, and women in their social networks. Williams and Mohammed (2013) offer a framework for orienting this work. Within their “Racism and Health” framework, they argue that upstream factors such as race, socioeconomic position, and gender, can operate through proximal pathways and lead to behavioral and psychological responses that impact health (p. 1157; Williams & Mohammed, 2013). Our work sheds light on the complexities of MM. Findings can inform interventions that mitigate its effects on HIV prevention and treatment.

MM is recognized as a barrier to care but still not fully understood (Benkert et al., 2019). One reason is because there is little consensus on how to define and measure it. One conceptualization is that trust is a bipolar concept such that mistrust indicates the opposite of trust (M. A. Hall et al., 2001). We, however, view MM as a distinct construct that captures active distrust in health care providers and/or institutions (Benkert et al., 2019). Given the centrality of race in shaping daily life and for understanding health disparities in the United States (Williams & Mohammed, 2013), we conceive of MM as a group-based concept (Thompson et al., 2004). With race at the fore, MM can be understood as a rational response to historical trauma and daily experiences of discrimination (Benkert et al., 2019), and something that is transmitted and reinforced through broad societal mechanisms (e.g., structural racism), personal social networks (e.g., vicarious racism), and everyday interactions between individuals (e.g., individual racism; Jaiswal & Halkitis, 2019).

Existing research on MM has largely focused on its historical causes and its consequences. Historical atrocities (e.g., Tuskegee Syphilis Study) have been cited as driving forces of MM (Dula, 1994). Cultural transmission is one way that MM spreads within communities and across generations, such that individuals learn these attitudes through the experiences of other members of their group (Benkert et al., 2019; Terrell & Terrell, 1981). Among people who are BAA, direct experiences of discrimination (Lillie-Blanton et al., 2000) and even vicarious experiences of discrimination (e.g., learning of others’ experiences) in a medical setting are associated with increased MM (Powell et al., 2019). Johnson et al. (2004) found that doctors are less likely to take a patient-centered approach, more likely to be verbally dominant when interacting with BAA patients, and less likely to treat pain among BAA patients (Cintron & Morrison, 2006; Hoffman et al., 2016). Other minority women also have reason to mistrust medical institutions. For example, Latinx women who were enrolled in unethical medical experiments carried out by the U.S. government in the 1940s (Rodriguez & García, 2013) or have more recently experienced forced sterilization (Kendall & Albert, 2015). Regardless of the mechanisms of transmission, research shows that MM has far reaching effects for many health behaviors, including those related to HIV.

Understanding MM is important for the entire HIV prevention and care continuum. There is evidence that MM lowers willingness to use preexposure prophylaxis (PrEP; Eaton et al., 2014), is associated with reduced condom use (Ross et al., 2006), is a barrier to regular HIV testing (Tun et al., 2012), and results in poorer adherence to antiretroviral therapy (ART; Bogart et al., 2011). While most evidence shows that MM is a barrier, some studies report mixed findings or no effect. For example, endorsement in conspiracy beliefs did not necessarily reduce ART adherence as compared to those who did not endorse conspiracy beliefs (Clark et al., 2008). Similarly, a qualitative exploration of MM found that people living with HIV who held conspiratorial beliefs about HIV, were, nevertheless, able to prioritize their health and seek care (Jaiswal et al., 2018). Outside the context of HIV, a study examining the effect of MM on three types of preventive health screening found that MM delayed blood pressure screening but did not delay routine check-ups or cholesterol screenings (Powell et al., 2019). These mixed findings suggest that researchers do not yet fully understand the conditions in which MM is a barrier and for whom.

There is at least one other study that considers more proximate causes and psychosocial correlates of MM (Hammond, 2010). This study found that age and experiences of discrimination were positively correlated with MM. A limitation of that study and many others in this area is that samples are restricted to MSM. Given the disproportionate disease burden experienced by MSM, this focus is sensible and strategic, but it leaves many other communities underrepresented in innovations to prevent new infections. Notable exceptions that focus on minority women have studied outcomes related to breast cancer screening (Molina et al., 2015; Shelton et al., 2011; Thompson et al., 2004) or cervical cancer screening (M. B. Hall et al., 2018). Studies with a focus on minority women and HIV do not necessarily include MM in their analyses (Bogart et al., 2015). The current study fills a gap by shifting the focus to a diverse sample of at-risk women and other women in their networks, because MM might operate differently among these communities. Unique factors include experiences of violence (Gielen et al., 2001) or having and caring for children, which both disproportionately affect women. By assessing the correlates of MM in the context of HIV testing, interventions can be tailored specifically to the needs of minority women.

MM is important to consider in the broader social context since cultural transmission is one mechanism by which MM is spread. Social networks (as the primary focus of the social context for an individual) are also important determinants of HIV risk behaviors and care, as they transmit and uphold social norms about condom use (Latkin et al., 2003) or needle-sharing among injection drug users (Latkin et al., 2010). Ties to acquaintances might provide new sources of information (Granovetter, 1973, 1983) while family or friendship ties can provide social support (Gielen et al., 2001). Beyond these roles that network members fulfill, there are structural aspects of networks that are also important. In consideration of network-based interventions around HIV risk reduction, it is important to know the size of a network and how densely connected network members are to each other (Valente, 2012). Having larger networks corresponds to more HIV testing (Huang et al., 2012) and use of PrEP (Kuhns et al., 2017). Finally, research suggests that MM at the network-level has effects on individual health outcomes. While Clark et al. (2008) found that an individual’s conspiratorial beliefs were not necessarily a barrier to care, being embedded within a network where conspiracy beliefs are common might be. Bogart et al. (2016) found that if an individual’s network held HIV conspiracy beliefs, the individual was less likely to adhere to their ART regimen.

In sum, the literature in this area provides evidence of the far-reaching consequences of MM. It does not, however, capture the ways in which MM relates to other demographic, psychosocial, or social network processes. Evidence is lacking among minority women at risk for HIV. Before interventions can mitigate the downstream effects of MM relative to HIV prevention and treatment, we must first understand what else is related to MM. The objective of this study is to describe how characteristics of individuals and social networks are related to MM in a racially diverse sample of at-risk women and other women in their networks.

Method

Sample

Using convenience sampling and peer-referral sampling techniques (Heckathorn, 1997; Kimbrough et al., 2009), participants were recruited from a small Western city between 2016 and 2018. There were two phases of recruitment. Index participants were recruited through agency based, street based, and online efforts, including clinics and social service organizations, universities and community colleges, churches, community events, bars and restaurants, and various online platforms. To be eligible as an index participant, individuals had to be cis- or transfemale, aged 18 years or older, self-identify as African American/Black or Hispanic/Latinx, living in the metropolitan area, willing to refer others to the study, and at risk for HIV, defined as (1) negative or unknown HIV status and (2) self-report any HIV risk behavior (defined as multiple sex partners in past 6 months, injecting drugs, transactional sex, or sex with a high-risk partner [defined as partner injects drugs, partner engages in transactional sex, partner uses noninjection drugs, partner has concurrent partners, partner is a MSM]). If eligible, participants completed a 2-hour, interviewer-administered survey to assess their attitudes, behaviors, and social networks. Index respondents were given five coupons linked to their study ID and asked to refer network members to participate. The only eligibility criterion for referred participants was that they were linked to an index participant by presenting a coupon or knowing her name. As a result, the full sample (N = 219) comprises BAAW and Latinx women and people in their networks. Index and referred participants were compensated $45 for participation and index participants were also eligible to receive an additional $5 compensation for each referral, up to five. Participants provided written informed consent and study protocols were approved by the institutional review board at the University of Nevada, Reno.

Measures

Individual-level measures were all self-reported and included demographics, psychological characteristics (e.g., perceptions of HIV risk, stigmatizing attitudes about HIV), HIV risk and prevention behaviors, and previous history with regard to experiences of discrimination, physical and sexual abuse, engagement in health services, and diagnoses of mental health disorders and sexually transmitted infections (Table 1). During the social network elicitation portion of the survey, respondents (i.e., egos) could list up to twenty people in their network (i.e., alters) who they had “communicated with in the past 6 months.” Ego answered a series of questions about each alter’s demographics, ego’s relationship to alters, ego’s perceptions about alters’ attitudes and behaviors, and behaviors that respondent and alters engaged in together (e.g., unprotected sex, drug use). From this social network elicitation, ego-level structural and functional elements of these networks were calculated. Structural measures of networks were generated from the open-source software package used for network data collection (EgoWeb 2.0, computer software). Network size was measured by the number of alters listed by ego. Structural measures indicating how connected or disconnected ego’s network members are to each other include density, dyads, and components (Wasserman & Faust, 1994). Alters are nested within egos, and the alter-level data were aggregated to the ego-level as proportions that range from zero to one such that higher values indicate a greater proportion of ego’s network with a given attribute (e.g., female). Proportion-based network-level variables included demographics, functional role (e.g., family member, community leader), and ego’s perceptions of network members’ personal history, attitudes, and behaviors. Finally, three measures captured the average level of trust, conflict, and risk for contracting HIV from ego’s listed alters, on a scale from 1 to 10 and then aggregated to the ego-level via an average score.

Table 1.

Detailed List and Description of Measures.

Single variable measures	Level of measurement and response options
Sex/gender	Nominal; female, transfemale, male, transmale, gender-queer/nonbinary; recoded as binary; women = 1 (cisgender and transgender); men = 0 (excluded)
Race/ethnicity	Nominal, select all that apply: Black or African American, African, Afro-Caribbean, Black (other), Asian, American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, White, Latino or Hispanic, and Other (specify); recoded as binary; any Black/African American = 1
Age	Continuous in years
Income	Continuous in dollar amount from all sources per month
Partner	1 = Yes, 0 = No
Children	Count of number of children
Homeless (past 6 months)	1 = Yes, 0 = No
Arrested (ever)	1 = Yes, 0 = No
Physical abuse (before 18 years)	1 = Yes, 0 = No
Sexual abuse (before 18 years)	1 = Yes, 0 = No
Been forced to have sex (ever)	1 = Yes, 0 = No
Injected drugs (ever)	1 = Yes, 0 = No
Overdosed from heroin (ever)	1 = Yes, 0 = No
Had regular sex partners (past 6 months)	1 = Yes, 0 = No
Had casual sex partners (past 6 months)	1 = Yes, 0 = No
Unprotected sex (past 6 months)	1 = Yes, 0 = No
High risk sex partner (past 6 months)	Sex with any of the following: someone who injects drugs, a man who has sex with men, someone who is HIV positive, someone who was released from jail in the past 6 months; 1 = Yes, 0 = No
Engaged in transactional sex (ever)	1 = Yes, 0 = No
HIV testing frequency	Ordinal; 1 = I have never tested, 2 = Less than every 3–4 years, 3 = About every 3–4 years, 4 = About every 2 years, 5 = 1–2 times per years, 5 = More than 2 times per year
Perceived HIV risk	Ordinal; 0 = Not at all at risk, 1 = A little at risk, 2 = Somewhat at risk, 3 = Very at risk, 4 = Extremely at risk
In drug treatment (ever)	1 = Yes, 0 = No
In 12-step such as AA/NA (ever)	1 = Yes, 0 = No
Dx with mental illness (ever)	1 = Yes, 0 = No
Dx with STI (ever)	1 = Yes, 0 = No
Dx hepatitis C (ever)	1 = Yes, 0 = No
Dx hepatitis B (ever)	1 = Yes, 0 = No
Scale measures	No. of items	Cronbach’s α in current study	Citation
Group-Based Medical Mistrust Scale	12	.89	Thompson et al. (2004)
Multiple discrimination scale–race	10	.80	Bogart et al. (2013)
HIV knowledge	18		Carey and Schroder (2002)
HIV stigma	17	.85	Earnshaw et al. (2012)
HIV testing barriers	10	.80	Earnshaw et al. (2012)

Note. AA/NA = Alcoholics Anonymous/Narcotics Anonymous; Dx = diagnosis; STI = sexually transmitted infection.

Analytic Strategy

The analytic sample was restricted to female-identified participants, inclusive of four transgender women (n = 165). Analysis was conducted with SAS Version 9.4 (SAS Institute Cary, NC). First, we described the Group-Based Medical Mistrust Scale (GBMMS; Thompson et al., 2004, Table 2) overall and by individual item. Using a series of between group t tests, we evaluated differences in GBMMS scores between two racial/ethnic subgroups. Due to the small number of women in some racial/ethnic categories, we conducted analyses comparing BAA-identified women to all others. The Supplemental Appendix includes tables that present more granular analysis of racial/ethnic categories; conclusions derived from these analyses were largely consistent with the simpler categorization, so we present the two-category analysis for simplicity. We also looked at the correlations between GBMMS and the mean level of trust that egos reported having in their network members.

Table 2.

Descriptives of Group-Based Medical Mistrust Scale (GBMMS): Mean, Standard Deviation, and Racial/Ethnic^a Differences.

GBMMS item	Overall^b (n = 165)	Black¹ (n = 103)	Non-Black² (n = 62)
Average GBMMS score^c	2.64 (0.78)	2.74 (0.8)²	2.49 (0.71)¹
Individual items^d
1. Doctors and health care workers sometimes hide information from patients who belong to your racial group.	2.85 (1.27)	3.00 (1.31)	2.61 (1.16)
2. Doctors have the best interests of people of your racial group in mind.^e	3.28 (1.08)	3.15 (1.08)	3.48 (1.07)
3. People of your racial group should not confide in doctors and health care workers because it will be used against them.	2.12 (0.89)	2.17 (0.9)	2.03 (0.87)
4. People of your racial group should be suspicious of information from doctors and health care workers.	2.51 (1.13)	2.62 (1.16)	2.32 (1.07)
5. People of your racial group cannot trust doctors and health care workers.	2.20 (1.04)	2.23 (1.02)	2.15 (1.08)
6. People of your racial group should be suspicious of modern medicine.	2.59 (1.26)	2.64 (1.3)	2.52 (1.21)
7. Doctors and health care workers treat people of your racial group like “guinea pigs.”	2.58 (1.19)	2.71 (1.19)	2.37 (1.18)
8. People of your racial group receive the same medical care from doctors and health care workers as people from other groups.^e	3.05 (1.2)	3.05 (1.21)	3.05 (1.21)
9. Doctors and health care workers do not take the medical complaints of people of your racial group seriously.	2.85 (1.11)	2.97 (1.09)	2.66 (1.13)
10. People of your racial group are treated the same as people of other groups by doctors and health care workers.^e	3.07 (1.17)	2.95 (1.2)	3.27 (1.09)
11. In most hospitals, people of different racial groups receive the same kind of care.	3.15 (1.11)	3.04 (1.13)	3.34 (1.05)
12. You have personally been treated poorly or unfairly by doctors or health care workers because of your race.^e	2.55 (1.24)	2.67 (1.29)	2.35 (1.15)

Race/ethnicity defined as follows: Black = Black or multiracial with Black; Non-Black = non-Hispanic White, Hispanic/Latino, Asian, American Indian/Alaskan Native, Native Hawaiian/Pacific Islander, multiracial (excluding multiracial Black). ^bSignificant differences (between group t test; p < .05) between racial/ethnic subgroups indicated with superscripts: ¹Black, ²Non-Black. ^cAverage score based on items answered; 6 respondents answered 11/12 items. ^dItem answer choices: Strongly disagree (1), Disagree (2), Neither agree nor disagree (3), Agree (4), Strongly agree (5). ^eIndicates item that was reverse-coded for scale.

Second, we conducted bivariate and multivariable analyses to investigate correlates of MM, first examining associations in the overall sample (n = 165) and subsequently examining associations in racial/ethnic subgroups (results from the BAA vs. other subgroups are presented here, with tables available in the Supplemental Appendix). We used linear regression to estimate the unadjusted coefficients of each variable’s association with MM. Next, we evaluated correlations between individual- and network-level variables that were significant at p < .10. Research designs that rely on one-step referral processes may affect the independence of responses between index participants and referred participants (Allison, 1999). Prior to running multivariable models, an intraclass correlation coefficient (ICC) was calculated to investigate the possibility of correlated outcomes. The ICC was small and not statistically significant (ICC < .0001, p = 1.0); as a result, cases were treated as independent in the final models. Finally, variables that were significant (p < .05) or approaching significance (p < .10) in the bivariate analysis were entered into multivariable linear regression models. A final model was selected using manual backward-selection, retaining variables with p < .10. In a sensitivity analysis to evaluate the possible effects of missing data, the multivariable models were rerun using multiple imputation, a technique that models the properties of variability and uncertainty around the missing data (Enders, 2010). Findings from imputed models were not substantively different and, as a result, we present the original models without imputation.

Results

Descriptive statistics of GBMMS overall and individual items are presented in Table 2. Of the 165 women respondents in this sample, 72 were BAA (43.6%), 23 were Hispanics (13.9%), 22 were non-Hispanic Whites (13.3%), and eight respondents (4.6%) were another racial/ethnic group. One in four (24.2%, n = 40) respondents were multiracial, including some respondents who identified as Black and some other race. When dichotomized, 103 women indicated any Black (including multiracial) and 62 did not. The average MM score for BAA respondents (2.74, on a Scale of 1 to 5) was higher than for other racial/ethnic groups (2.49, p < .05). As a check, we also looked at the correlations between GBMMS and the mean level of trust that egos reported having in their network members. There was a slight but significant negative relationship between GBMMS and an aggregated measure of trust within ego’s relationships (Pearson R correlation coefficient = −.22, p < .01), indicating that MM is somewhat but not strongly related to interpersonal trust at the network level.

Individual-level correlates of MM are described in Table 3, first by the overall sample and then split by BAAW versus else. For the overall sample, eight individual-level factors were positively and significantly (p < .05) associated with MM: being BAA, older age, experiencing physical abuse before 18 years, experiencing sexual abuse before 18 years, ever having been forced to have sex, experiences of discrimination, perceived barriers to HIV testing, and ever having engaged in transactional sex. One additional variable approached significance (p < .10): having ever overdosed from heroin. In the subgroup analysis of BAAW, age, discrimination, and engaging in transactional sex were significant. Among non-Black women, only discrimination was associated with MM although having a regular sex partner and previous engagement in a 12-step program were approaching significance (p < .10), which was not the case for BAAW.

Table 3.

Individual-Level Variables: Sample Descriptives and Unadjusted Linear Regression Coefficients of Medical Mistrust on Individual-Level Variables.

Individual level	All women (n = 165)			Black women (n = 103)			Non-Black women (n = 62)
Individual level	% (n) or M (SD)	Unadjusted coefficient	p	% (n) or M (SD)	Unadjusted coefficient	p	% (n) or M (SD)	Unadjusted coefficient	p
Demographics
Black (including multiracial)	62.4% (103)	0.25	<.05*	100.0% (103)	—	—	0% (0)
Race/ethnicity
Black	43.6% (72)	—	—	69.9% (72)	—	—
White	13.3% (22)			0(0)
Hispanic	13.9% (23)			0% (0)			37.1% (23)
Other	4.9% (8)			0% (0)			35.5% (22)
Multiracial	24.2% (40)			30.1% (31)			27.4% (17)
Gender^a
Female	97.6 (161)	—	—	98.1% (101)	—	—	96.8% (60)
Transfemale	2.4% (4)	—	—	1.9% (2)	—	—	3.2% (2)
Male
Age (years)	36.35 (12.92)	0.01	.02*	35.23 (12.52)	0.01	.04*	38.21 (13.45)	0.01	0.08
Years of education	13.06 (2.23)	0.01	.84	13.21 (2.01)	−0.03	.44	12.82 (2.54)	0.03	.35
No. of children	1.70 (1.90)	0.00	.98	1.70 (1.93)	−0.02	.61	1.69 (1.86)	0.04	.47
Monthly income (in dollars)	989.35 (917.64)	0.00	.61	975.63 (870.19)	0.0	.79	1011.48 (996.35)	0.00	.23
Has partner	53.7% (88)	0.07	.52	57.8% (59)	−0.05	.75	46.8% (29)	0.27	.14
Personal history
Homeless (past 6 months)	52.1% (86)	0.17	.16	49.5% (51)	0.21	.19	56.5% (35)	0.16	.38
Arrested (ever)	68.5% (113)	0.09	.47	68.9% (71)	0.02	.91	67.7% (42)	0.21	.28
Physical abuse before 18 years	49.4% (77)	0.25	.04*	48.5% (47)	0.28	.09	50.9% (30)	0.23	.21
Sexual abuse before 18 years	58.3% (91)	0.29	.02*	61.9% (60)	0.23	.17	52.5% (31)	0.31	.09
Been forced to have sex (ever)	54.7% (87)	0.29	.02*	56.6% (56)	0.24	.15	51.7% (31)	0.35	.05
Experiences of discrimination	2.84 (2.47)	0.13	<.0001*	3.07 (2.49)	0.13	<.0001*	2.45 (2.41)	0.11	.002*
Psychological/cognitive
HIV knowledge	0.69 (0.19)	−0.28	.37	0.70 (0.19)	−0.25	.55	0.66 (0.2)	−0.49	.29
HIV stigma	2.03 (0.52)	−0.06	.62	2.05 (0.53)	−0.06	.67	2 (0.52)	−0.08	.67
Perceived risk for contracting HIV	1.08 (1.04)	0.05	.41	1.02 (1.1)	0.05	.50	1.18 (0.92)	0.07	.47
Perceived barriers to HIV testing	2.75 (0.76)	0.18	.02*	2.71 (0.72)	0.20	.07	2.82 (0.83)	0.19	.08
HIV risk and prevention behaviors
Injected drugs (ever)	26.7% (44)	0.06	.64	23.3% (24)	0.18	.34	32.3% (20)	−0.03	.88
Overdosed from heroin (ever)	6.1% (10)	0.48	.06	9.9% (10)	0.40	.14	0% (0)	—	—
Had regular sex partners (past 6 months)	79.1% (121)	0.21	.19	81.8% (81)	0.04	.87	74.1% (40)	0.38	.09
Had casual sex partners (past 6 months)	52.9% (81)	0.03	.83	57.6% (57)	0.06	.71	44.4% (24)	−0.12	.55
Unprotected sex (past 6 months)	73.9% (122)	0.13	.33	80.6% (83)	0.16	.34	62.9% (39)	0.05	.80
High-risk sex partner (past 6 months)	40.6% (58)	0.09	.52	44% (40)	0.12	.56	34.6% (18)	−0.12	.58
Engaged in transactional sex (ever)	34.0% (55)	0.28	.03*	38% (38)	0.48	.004*	54% (100)	−0.18	.37
HIV testing frequency
Never	18.3% (30)	0.05	.18	16.7% (17)	0.07	.15	21% (13)	−0.01	.84
Less than every 3−4 years	17.1% (28)			12.8% (13)			24.2% (15)
About every 3−4 years	6.1% (10)			6.9% (7)			4.8% (3)
About every 2 years	12.8% (21)			13.7% (14)			11.3% (7)
1−2 times per year	36% (59)			37.3% (38)			33.9% (21)
More than 2 times per year	9.8% (16)			12.8% (13)			4.8% (3)
Health and health services
In drug treatment (ever)	34.2% (56)	−0.17	.18	29.1% (30)	−0.26	.14	42.6% (26)	0.03	.87
In 12-step such as AA/NA (ever)	43.0% (71)	0.01	.92	34.0% (35)	−0.11	.53	58.1% (36)	0.35	.06
Dx with mental illness (ever)	53.9% (89)	0.09	.46	55.3% (57)	−0.02	.88	51.6% (32)	0.26	.15
Dx STI (ever)	37.8% (62)	0.04	.78	43.7% (45)	0.01	.96	27.9% (17)	−0.03	.90
Dx hepatitis C (ever)	5.5% (9)	−0.19	.48	4.9% (5)	−0.18	.63	6.6% (4)	−0.16	.66
Dx hepatitis B (ever)	1.8% (3)	−0.12	.78	2.9% (3)	−0.21	.65	0% (0)	—	—

Note. AA/NA = Alcoholics Anonymous/ Narcotics Anonymous; Dx = diagnosis; STI = sexually transmitted infection.

p < .05.

Table 4 summarizes network-level characteristics and their association with MM. Average network size was 14.3 alters (the possible number of alters was capped at 20). Average network density was 0.36, indicating that of all possible connections between listed alters, 36% were present on average. Proportion variables can be interpreted as percent of ego’s network with a given attribute. Therefore, an average respondent’s network was 50% female and 42% BAA. Five network-level variables were significantly associated with MM (p < .05) in the overall sample. Proportion of network who are friends and mean network trust were both negatively associated, such that networks with fewer friends and lower levels of trust correspond to higher MM. Conversely, having a high proportion of network members who are BAA, engage in transactional sex, or are stigmatizing toward HIV corresponded to higher MM. One additional variable approaching significance (p < .10) was considered in multivariable models: conflict with network members. Additional analysis presented in Table 4 investigated correlates among stratified samples of BAAW and women of other racial/ethnic identities. In the analysis of BAAW, network proportion friend and mean network trust were negatively and significantly associated with MM and several network variables were approaching significance. Density was negatively associated with MM and network-level conflict, transactional sex, and HIV stigma were positively associated (p < .10). For non-Black women, having more community leaders in the network was positively associated with MM (p < .05), as was having a network with more HIV-positive diagnoses members (p < .10).

Table 4.

Network-Level Variables: Sample Descriptives and Unadjusted Linear Regression Coefficients of Medical Mistrust on Network-Level Variables.

Network level	All women (n = 165)			Black (n = 103)			Non-Black (n = 62)
Network level	% (n) or M (SD)	Unadjusted coefficient	p	% (n) or M (SD)	Unadjusted coefficient	p	% (n) or M (SD)	Unadjusted coefficient	p
Structural
Size	14.27 (6.06)	−0.01	.34	13.24 (6.2)	−0.01	.69	15.97 (5.46)	−0.01	.71
Density	0.36 (0.29)	−0.16	.44	0.38 (0.3)	−0.47	.08	0.33 (0.27)	0.38	.26
Components	1.39 (0.91)	−0.08	.23	1.34 (0.85)	−0.05	.57	1.48 (1.00)	−0.09	.31
Dyads	9.75 (50.32)	0.00	.65	11.52 (62.31)	0.00	.76	6.81 (17.34)	−0.01	.23
Network member demographics
Proportion female	0.50 (0.22)	−0.01	.97	0.52 (0.24)	−0.21	.55	0.47 (0.19)	0.29	.55
Proportion Black	0.42 (0.37)	0.34	.04*	0.62 (0.31)	0.24	.36	0.07 (0.13)	0.09	.90
Proportion live same city	0.64 (0.34)	−0.25	.16	0.64 (0.35)	−0.14	.53	0.65 (0.33)	−0.45	.11
Proportion ego has children with	0.02 (0.09)	0.28	.67	0.03 (0.11)	0.45	.55	0.01 (0.05)	−1.71	.31
Network member role
Relationship
Proportion family	0.34 (0.27)	0.11	.64	0.35 (0.27)	−0.23	.44	0.31 (0.26)	0.61	.08
Proportion friend^a	0.71 (0.34)	−0.44	.01*	0.71 (0.35)	−0.57	.01*	0.73 (0.32)	−0.16	.56
Proportion acquaintance	0.11 (0.21)	0.29	.31	0.10 (0.20)	0.57	.16	0.12 (0.23)	0.00	.99
Role^b
Proportion core group	0.37 (0.27)	−0.17	.44	0.40 (0.28)	−0.28	.33	0.32 (0.26)	−0.15	.67
Proportion opinion leader	0.39 (0.34)	−0.10	.58	0.39 (0.35)	−0.12	.60	0.39 (0.31)	−0.06	.84
Proportion community leader	0.20 (0.25)	0.02	.93	0.22 (0.25)	−0.52	.10	0.17 (0.26)	0.79	.03*
Ego/alter risk behavior
Proportion sex partner (ever)	0.16 (0.20)	0.15	.63	0.19 (0.23)	0.15	.68	0.13 (0.15)	−0.25	.68
Proportion paid ego for sex^c	0.10 (0.27)	0.17	.49	0.11 (0.29)	0.23	.46	0.08 (0.25)	−0.05	.90
Proportion used drugs together	0.30 (0.33)	−0.06	.69	0.30 (0.34)	−0.05	.85	0.29 (0.32)	−0.40	.16
Support
Proportion monetary support	0.38 (0.33)	−0.11	.54	0.38 (0.33)	−0.25	.31	0.37 (0.32)	0.10	.72
Proportion emotional support	0.28 (0.27)	0.09	.67	0.28 (0.28)	0.16	.59	0.29 (0.27)	0.45	.17
Proportion health support	0.23 (0.29)	0.26	.24	0.23 (0.28)	−0.02	.95	0.25 (0.30)	0.28	.38
Mean trust	6.69 (2.24)	−0.08	.01*	6.56 (2.43)	−0.10	.003*	6.90 (1.91)	0.00	.99
Mean conflict	2.50 (1.74)	0.06	.08	2.45 (1.60)	0.09	.09	2.58 (1.95)	0.04	.35
Perceptions of network member history/attitudes/behaviors
Proportion incarcerated	0.05 (0.09)	1.01	.13	0.06 (0.14)	1.23	.12	0.04 (0.06)	−0.88	.55
Proportion engaged in transactional sex	0.04 (0.11)	1.18	.03*	0.20 (0.35)	1.02	.08	0.02 (0.05)	1.23	.52
Proportion HIV stigma	0.17 (0.31)	0.46	.02*	0.85 (0.34)	0.44	.06	0.10 (0.22)	0.27	.51
Proportion support ego HIV testing	0.84 (0.34)	−0.07	.70	0.17 (0.35)	−0.31	.20	0.84 (0.35)	0.29	.27
Proportion tested for HIV	0.14 (0.31)	−0.13	.51	0.02 (0.14)	−0.22	.35	0.09 (0.25)	−0.05	.88
Proportion diagnosed with HIV	0.02 (0.11)	−0.56	.31	1.51 (1.33)	−0.74	.20	0.00 (0.01)	15.01	.06
Mean risk for getting HIV from alter	1.42 (1.14)	0.04	.41	0.06 (0.14)	0.01	.86	1.28 (0.70)	0.17	.19

Friend is not mutually exclusive from family or acquaintance. ^bNot mutually exclusive categories. ^cValid proportion, based on whether ego had sex with alter.

p < .05.

Table 5 shows the multivariable results. The correlation between being BAA and proportion of ego’s network that is BAA (Pearson R correlation = .72) was potentially concerning but had a variance inflation factor of 2.3, well within the range of acceptable values (Hair et al., 2013). As a result, we concluded that estimates in subsequent multivariable models would not be seriously distorted. In the full sample, being older, experiencing more discrimination, having a network with a higher proportion of members who are BAA was associated with increased MM (adjusted R² = .18). Among the subset of women who are BAA, the only significant variable retained in the final model was experiences of discrimination, though age and proportion of network members who paid for sex approach significance (ps = .08). Among the subset of women who are not BAA, discrimination, age, having a regular sex partner, and proportion in the network who is a family member were positively associated with MM. Finally, the more granular racial/ethnic subgroup comparisons presented in the appendix—while underpowered to complete a more sophisticated multivariable analysis—illustrated in a bivariate context that there were different correlates of MM for women of different racial/ethnic backgrounds. For example, among Latinx women, a network with more drug partners was marginally and negatively associated with MM (p < .10). But for White women, it was having a geographically local network that was negatively associated (p < .05).

Table 5.

Multivariable Linear Regression Modeling Individual-Level and Network-Level Characteristics on Measure of Group-Based Medical Mistrust (Overall and Stratified by Race).

Characteristic	All women (n = 165)				Black women (n = 103)				Non-Black women (n = 62)
	Full model (n = 110)		Reduced model (n = 165)		Full model (n = 72)		Reduced model (n = 100)		Full model (n = 50)		Reduced model (n = 54)
	Adjusted coefficient	p	Adjusted coefficient	p	Adjusted coefficient	p	Adjusted coefficient	p	Adjusted coefficient	p	Adjusted coefficient	p
Intercept	1.51	.01*	0.35	.00*	1.87	.01*	1.91	<.0001*	.42	.45	.82	.02*
Individual-level variables
Black (including multiracial)	−0.13	0.54	—	—	—	—	—	—	—	—	—	—
Age (years)	0.01	.04	0.01	.01*	.02	.06	.01	.08	.02	.03*	.02	.02*
Physical abuse before 18 years	0.10	.53	—	—	.13	.50	—	—	—	—	—	—
Sexual abuse before 18 years	−0.04	.80	—	—	—	—	—	—	.25	.18	—	—
Been forced to have sex (ever)	0.00	.99	—	—	—	—	—	—	−.26	.24	—	—
Experiences of discrimination	0.04	.23	0.03	.04*	.06	.14	.12	<.001*	.10	.03	.10	.01*
Perceived barriers to HIV testing	0.15	.13	0.09	.08	.12	.41	—	—	.07	.58	—	—
Overdosed from heroin (ever)	0.40	.16	—	—	—	—	—	—	—	—	—	—
Had regular sex partners (past 6 months)	—	—	—	—	—	—	—	—	.76	<.01*	.72	<.001*
Engaged in transactional sex (ever)	0.09	.60	—	—	.38	.11	—	—	.02	.94	—	—
In 12-step such as AA/NA (ever)	—	—	—	—					.16	.44	—	—
Network-level variables
Density	—	—	—	—	−.59	.09	—	—	—	—	—	—
Proportion Black	0.56	<0.05*	0.19	.02*	—	—	—	—	—	—	—	—
Proportion family	—	—	—	—	—	—	—	—	.70	.12	.98	.01*
Proportion friend^a	-0.19	0.43	—	—	−.28	.39	—	—	—	—	—	—
Proportion community leader	—	—	—	—	—	—	—	—	.21	.60	—	—
Proportion paid ego for sex^b	−0.32	0.32	—	—	−.70	.10	.28	.08	—	—	—	—
Mean trust	−0.031	.45	—	—	−.01	.80	—	—	—	—	—	—
Mean conflict	0.060	.19	0.04	.07	.03	.66	—	—	—	—	—	—
Proportion HIV stigma	0.46	.07	0.23	.05	.39	.16	—	—	—	—	—	—
Proportion diagnosed with HIV	—	—	—	—	—	—	—	—	7.00	.35	—	—
Model fit (adjusted R²)	.21		.18		.19		.20		.36		.35

Note. AA/NA = Alcoholics Anonymous/Narcotics Anonymous.

Friend is not mutually exclusive from family. ^bValid proportion, based on whether ego had sex with alter.

p < .05.

Discussion

We sought to understand what demographic, psychosocial, behavioral, and social network factors are associated with MM in a racially diverse sample of women. We found that group-based MM was highest among BAAW compared to women of other races and ethnicities. These findings align with previous research that shows racial differences in group-based MM while also lending some credence to the claims of M. A. Hall et al. (2001), which suggest that processes of interpersonal trust may operate independently of institutional mistrust.

In the full sample we identified positive and significant associations for age, experiences of discrimination, and proportion of network who is BAA. In the models that stratified by race, higher proportion of BAA in one’s network was not retained, suggesting that the variation in MM that was due to racial composition of the network was accounted for when stratifying by race. Furthermore, for the subgroup analysis and final models of BAAW, only discrimination was associated with MM. These findings speak to the “Racism and Health” framework offered by Williams and Mohammed (2013). Our findings suggest that it was not race, per se, that is associated with MM. Rather, experiencing discrimination due to one’s race and the accumulation of these experiences over a lifetime may be more proximate causes of MM. Furthermore, the composition of our broader social networks—members’ demographic characteristics and the roles they occupy—is associated with MM. This aligns with other research in which MM can be viewed as a rational response to both direct and vicarious experiences (Benkert et al., 2019).

These findings shed light on factors to consider when developing an HIV-based intervention for BAAW to mitigate MM’s effect on HIV prevention and treatment. As other research has suggested, providers should assess attitudes that contribute to MM and listen to narratives of patients’ experiences when developing a HIV prevention or treatment plan (Bogart et al., 2010; Gaston & Alleyne-Green, 2013). BAAW in our sample experienced more discrimination than their non-BAAW peers and much of their networks were also BAA. MM might be reinforced by network members who have similar perspectives and experiences. While there are many approaches to network-based interventions (Valente, 2012), some are more feasible than others. While it would not be reasonable or ethical to coach women to reconstruct their entire social network from scratch, it may be helpful to work with women to identify people within their networks who they could turn to with important health-related matters. For example, overall trust in one’s network was negatively associated with MM, so selectively drawing on more trusted members could be an adaptive technique. Considering these network-level findings, HIV interventions may benefit if providers also assess patients’ social networks to understand broader influences on HIV prevention and care. These findings and their implications, however, should be interpreted in light of their limitations.

While this study demonstrates individual- and network-level factors related to MM, there are some limitations worth noting. First, given the cross-sectional design of this study, we cannot make any claims about causality or direction of influence. For example, it is not clear if having a higher proportion of friends in one’s network decreases levels of MM or, rather, if being a mistrusting person means one is less likely to have friends. Second, given our relatively small sample size, it is possible that some subtle differences were not detected or not statistically significant in the multivariable model because our analysis was underpowered. Third, many network-level variables were collected from the perspective of the ego and not necessarily representative of the alter’s actual attitudes. Nevertheless, perceptions can be important determinants in and of themselves (Green et al., 2014). Finally, this sample was recruited in a small Western city, and may not be representative of BAAW and their networks in larger, more urban settings. However, given the dearth of research representing the experiences of BAAW in nonurban, noncoastal settings, we consider this a strength of the current research. Future research could investigate these findings in a larger and more geographically diverse sample of BAAW at risk for HIV and how receptive BAAW are to a network-based intervention around MM in the context of the HIV testing and prevention.

Conclusion

The current study set out to address several gaps in the MM literature. Specifically, while much is known about the prevalence and consequences of MM, relatively few studies have investigated characteristics associated with MM as an outcome. Furthermore, many studies about MM focus on MSM, which are limited in their applications to other at-risk populations. Finally, studies often treat MM as an individual trait that is influenced by other individual-level factors such as race and experiences of discrimination. To address these gaps, we explored individual- and network-level characteristics that were associated with MM among a racially diverse sample of women. We found that being older, having more experiences of racial discrimination, a history of physical abuse, having a regular sex partner, and having a higher proportion of one’s network that is BAA, are all significantly associated with higher MM. When developing interventions to mitigate the downstream effects of MM, practitioners should consider what other characteristics about individuals and their social contexts might be working against public health efforts and which ones could be harnessed and leveraged to create positive change.

Supplemental Material

sj-pdf-1-heb-10.1177_1090198120986783 – Supplemental material for Correlates of Medical Mistrust Among Minority Women at Risk for HIV and Their Networks

Supplemental material, sj-pdf-1-heb-10.1177_1090198120986783 for Correlates of Medical Mistrust Among Minority Women at Risk for HIV and Their Networks by Laura M. Johnson, Harold D. Green, Brandon Koch, Robert Harding, Jamila K. Stockman and Karla D. Wagner in Health Education & Behavior

Footnotes

Acknowledgements

We are grateful for the team whose efforts made this work possible, including Nicole Auldridge, Ashley Dawkins, Elvira Diaz, Hannah Nagy, Caitlin O’Leary, Krysti Smith, George Wendt, Ryan Wong, and other team members who contributed to the recruitment, data collection, data management, and project management required for this study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was part of a larger project, “Capitalizing On Networks to Decrease Obstacles to Testing” (CONnect the DOTs), supported by the National Institute on Drug Abuse of the National Institutes of Health under Award Number R01DA038185. Additional funding sources include K01DA031031 and San Diego Center for AIDS Research Grant P30AI036214.

ORCID iD

Laura M. Johnson

Supplemental Material

Supplemental material for this article is available online at .

References

Allison

P. D.

(1999). Multiple regression: A primer. Pine Forge Press.

Ball

Lawson

Alim

(2013). Medical mistrust, conspiracy beliefs & HIV-related behavior among African Americans. Journal of Psychological and Behavioral Science, 1(1), 1–7.

Benkert

Cuevas

Thompson

H. S.

Dove-Meadows

Knuckles

(2019). Ubiquitous yet unclear: A systematic review of medical mistrust. Behavioral Medicine, 45(2), 86–101. https://doi.org/10.1080/08964289.2019.1588220

Bogart

L. M.

Derose

K. P.

Kanouse

D. E.

Grifin

B. A.

Haas

A. C.

Williams

M. V.

(2015). Correlates of HIV testing among African American and Latino church congregants: The role of HIV stigmatizing attitudes and discussions about HIV. Journal of Urban Health, 92(1), 93–107. https://doi.org/10.1007/s11524-014-9927-y

Bogart

L. M.

Galvan

F. H.

Wagner

G. J.

Klein

D. J.

(2011). Longitudinal association of HIV conspiracy beliefs with sexual risk among black males living with HIV. AIDS and Behavior, 15(6), 1180–1186. https://doi.org/10.1007/s10461-010-9796-7

Bogart

L. M.

Landrine

Galvan

F. H.

Wagner

G. J.

Klein

D. J.

(2013). Perceived discrimination and physical health among HIV-positive Black and Latino men who have sex with men. AIDS and Behavior, 17(4), 1431–1441. https://doi.org/10.1007/s10461-012-0397-5

Bogart

L. M.

Ransome

Allen

Higgins-Biddle

Ojikutu

B. O.

(2019). HIV-related medical mistrust, HIV testing, and HIV risk in the national survey on HIV in the Black community. Behavioral Medicine, 45(2), 134–142. https://doi.org/10.1080/08964289.2019.1585324

Bogart

L. M.

Wagner

Galvan

F. H.

Banks

(2010). Conspiracy beliefs about HIV are related to antiretroviral treatment nonadherence among African American men with HIV. Journal of Acquired Immune Deficiency Syndromes, 53(5), 648–655. https://doi.org/10.1097/QAI.0b013e3181c57dbc

Bogart

L. M.

Wagner

G. J.

Green

H. D.

Mutchler

M. G.

Klein

D. J.

McDavitt

Lawrence

S. J.

Hilliard

C. L.

(2016). Medical mistrust among social network members may contribute to antiretroviral treatment nonadherence in African Americans living with HIV. Social Science & Medicine, 164, 133–140. https://doi.org/10.1016/j.socscimed.2016.03.028

10.

Brandon

D. T.

Isaac

L. A.

LaVeist

T. A.

(2005). The legacy of Tuskegee and trust in medical care: Is Tuskegee responsible for race differences in mistrust of medical care? Journal of the National Medical Association, 97(7), 951–956.

11.

Cahill

Taylor

S. W.

Elsesser

S. A.

Mena

Hickson

Mayer

K. H.

(2017). Stigma, medical mistrust, and perceived racism may affect PrEP awareness and uptake in black compared to white gay and bisexual men in Jackson, Mississippi and Boston, Massachusetts. AIDS Care, 29(11), 1351–1358. https://doi.org/10.1080/09540121.2017.1300633

12.

Carey

M. P.

Schroder

K. E.

(2002). Development and psychometric evaluation of the brief HIV Knowledge Questionnaire. AIDS Education and Prevention, 14(2), 172–182. https://doi.org/10.1521/aeap.14.2.172.23902

13.

Centers for Disease Control and Prevention. (2019a). Estimated HIV incidence and prevalence in the United States 2010–2016. HIV Surveillance Supplemental Report, 24(1). https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-supplemental-report-vol-24-1.pdf

14.

Centers for Disease Control and Prevention. (2019b). HIV and women. https://www.cdc.gov/hiv/group/gender/women/index.html

15.

Cintron

Morrison

R. S.

(2006). Pain and ethnicity in the United States: A systematic review. Journal of Palliative Medicine, 9(6), 1454–1473.

16.

Clark

Mayben

J. K.

Hartman

Kallen

M. A.

Giordano

T. P.

(2008). Conspiracy beliefs about HIV infection are common but not associated with delayed diagnosis or adherence to care. AIDS Patient Care and STDs, 22(9), 753–759. https://doi.org/10.1089/apc.2007.0249

17.

Dula

(1994). African American suspicion of the healthcare system is justified: What do we do about it? Cambridge Quarterly of Healthcare Ethics, 3(3), 347–357. https://doi.org/10.1017/S0963180100005168

18.

Earnshaw

V. A.

Smith

L. R.

Chaudoir

S. R.

Lee

I.-C.

Copenhaver

M. M.

(2012). Stereotypes about people living with HIV: Implications for perceptions of HIV risk and testing frequency among at-risk populations. AIDS Education and Prevention, 24(6), 574–581. https://doi.org/10.1521/aeap.2012.24.6.574

19.

Eaton

L. A.

Driffin

D. D.

Smith

Conway-Washington

White

Cherry

(2014). Psychosocial factors related to willingness to use pre-exposure prophylaxis for HIV prevention among Black men who have sex with men attending a community event. Sexual Health, 11(3), 244–251. https://doi.org/10.1071/SH14022

20.

Enders

C. T.

(2010). Applied missing data analysis. Guilford.

21.

Gaston

G. B.

Alleyne-Green

(2013). The impact of African Americans’ beliefs about HIV medical care on treatment adherence: A systematic review and recommendations for interventions. AIDS and Behavior, 17(1), 31–40. https://doi.org/10.1007/s10461-012-0323-x

22.

Gielen

A. C.

McDonnell

K. A.

A. W.

O’Campo

Faden

(2001). Quality of life among women living with HIV: The importance violence, social support, and self care behaviors. Social Science & Medicine, 52(2), 315–322. https://doi.org/10.1016/S0277-9536(00)00135-0

23.

Granovetter

(1973). The strength of weak ties. American Journal of Sociology, 78, 1360–1380. https://doi.org/10.1086/225469

24.

Granovetter

(1983). The strength of weak ties: A network theory revisited. Sociological Theory, 1, 201–233. https://doi.org/10.2307/202051

25.

Green

H. D.

Hoover

M. A.

Wagner

G. J.

Ryan

G. W.

Ssegujja

(2014). Measuring agreement between egos and alters: Understanding informant accuracy in personal network studies. Field Methods, 26(2), 126–140. https://doi.org/10.1177/1525822X13492676

26.

Hair

J. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

(2013). Multivariate data analysis: Pearson new international edition. Pearson Higher Education.

27.

Hall

M. A.

Dugan

Zheng

Mishra

A. K.

(2001). Trust in physicians and medical institutions: What is it, can it be measured, and does it matter? Milbank Quarterly, 79(4), 613–639. https://doi.org/10.1111/1468-0009.00223

28.

Hall

M. B.

Vos

Bess

J. J.

Reburn

K. L.

Locklear

G. D.

McAlister

Bell

R. A.

(2018). Cervical cancer screening behaviors and perceptions of medical mistrust among rural Black and White women. Journal of Health Care for the Poor and Underserved, 29(4), 1368–1385. https://doi.org/10.1353/hpu.2018.0101

29.

Hammond

W. P.

(2010). Psychosocial correlates of medical mistrust among African American men. American Journal of Community Psychology, 45(1–2), 87–106. https://doi.org/10.1007/s10464-009-9280-6

30.

Heckathorn

D. D.

(1997). Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems, 44(2), 174–199. https://doi.org/10.2307/3096941

31.

Hoffman

K. M.

Trawalter

Axt

J. R.

Oliver

M. N.

(2016). Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proceedings of the National Academy of Sciences of the U S A, 113(16), 4296–4301. https://doi.org/10.1073/pnas.1516047113

32.

Huang

Z. J.

Nehl

E. J.

Zheng

Smith

B. D.

Zhang

McNabb

Wong

F. Y.

(2012). Social network and other correlates of HIV testing: Findings from male sex workers and other MSM in Shanghai, China. AIDS and Behavior, 16(4), 858–871. https://doi.org/10.1007/s10461-011-0119-4

33.

Jaiswal

Halkitis

P. N.

(2019). Towards a more inclusive and dynamic understanding of medical mistrust informed by science. Behavioral Medicine, 45(2), 79–85. https://doi.org/10.1080/08964289.2019.1619511

34.

Jaiswal

Singer

S. N.

Griffin Tomas

Lekas

H.-M.

(2018). Conspiracy beliefs are not necessarily a barrier to engagement in HIV care among urban, low-income people of color living with HIV. Journal of Racial and Ethnic Health Disparities, 5(6), 1192–1201. https://doi.org/10.1007/s40615-018-0466-1

35.

Johnson

R. L.

Roter

Powe

N. R.

Cooper

L. A.

(2004). Patient race/ethnicity and quality of patient–physician communication during medical visits. American Journal of Public Health, 94(12), 2084–2090. https://doi.org/10.2105/AJPH.94.12.2084

36.

Kendall

Albert

(2015). Experiences of coercion to sterilize and forced sterilization among women living with HIV in Latin America. Journal of the International AIDS Society, 18(1). https://doi.org/10.7448/IAS.18.1.19462

37.

Kimbrough

L. W.

Fisher

H. E.

Jones

K. T.

Johnson

Thadiparthi

Dooley

(2009). Accessing social networks with high rates of undiagnosed HIV infection: The social networks demonstration project. American Journal of Public Health, 99(6), 1093–1099. https://doi.org/10.2105/AJPH.2008.139329

38.

Kuhns

L. M.

Hotton

A. L.

Schneider

Garofalo

Fujimoto

(2017). Use of pre-exposure prophylaxis (PrEP) in young men who have sex with men is associated with race, sexual risk behavior and peer network size. AIDS and Behavior, 21(5), 1376–1382. https://doi.org/10.1007/s10461-017-1739-0

39.

Latkin

C. A.

Forman

Knowlton

Sherman

(2003). Norms, social networks, and HIV-related risk behaviors among urban disadvantaged drug users. Social Science & Medicine, 56(3), 465–476. https://doi.org/10.1016/S0277-9536(02)00047-3

40.

Latkin

C. A.

Kuramoto

S. J.

Davey-Rothwell

M. A.

Tobin

K. E.

(2010). Social norms, social networks, and HIV risk behavior among injection drug users. AIDS and Behavior, 14(5), 1159–1168. https://doi.org/10.1007/s10461-009-9576-4

41.

Lillie-Blanton

Brodie

Rowland

Altman

McIntosh

(2000). Race, ethnicity, and the health care system: Public perceptions and experiences. Medical Care Research and Review, 57(1_suppl), 218–235. https://doi.org/10.1177/1077558700057001S10

42.

Molina

Kim

Berrios

Calhoun

E. A.

(2015). Medical mistrust and patient satisfaction with mammography: The mediating effects of perceived self-efficacy among navigated African American women. Health Expectations, 18(6), 2941–2950. https://doi.org/10.1111/hex.12278

43.

Powell

Richmond

Mohottige

Yen

Joslyn

Corbie-Smith

(2019). Medical mistrust, racism, and delays in preventive health screening among African-American men. Behavioral Medicine, 45(2), 102–117. https://doi.org/10.1080/08964289.2019.1585327

44.

Rodriguez

M. A.

García

(2013). First, do no harm: The US sexually transmitted disease experiments in Guatemala. American Journal of Public Health, 103(12), 2122–2126. https://doi.org/10.2105/AJPH.2013.301520

45.

Ross

M. W.

Essien

E. J.

Torres

(2006). Conspiracy beliefs about the origin of HIV/AIDS in four racial/ethnic groups. Journal of Acquired Immune Deficiency Syndrome, 41(3), 342. https://doi.org/10.1097/01.qai.0000209897.59384.52

46.

Shelton

R. C.

Goldman

R. E.

Emmons

K. M.

Sorensen

Allen

J. D.

(2011). An investigation into the social context of low-income, urban Black and Latina women: Implications for adherence to recommended health behaviors. Health Education & Behavior, 38(5), 471–481. https://doi.org/10.1177/1090198110382502

47.

Terrell

(1981). An inventory to measure cultural mistrust among Blacks. The Western Journal of Black Studies, 5(3), 180.

48.

Thompson

H. S.

Valdimarsdottir

H. B.

Winkel

Jandorf

Redd

(2004). The Group-Based Medical Mistrust Scale: Psychometric properties and association with breast cancer screening. Preventive Medicine, 38(2), 209–218. https://doi.org/10.1016/j.ypmed.2003.09.041

49.

Tun

Kellerman

Maimane

Fipaza

Sheehy

Nel

(2012). HIV-related conspiracy beliefs and its relationships with HIV testing and unprotected sex among men who have sex with men in Tshwane (Pretoria), South Africa. AIDS Care, 24(4), 459–467. https://doi.org/10.1080/09540121.2011.617412

50.

Valente

T. W.

(2012). Network Interventions. Science, 337(6090), 49–53. https://doi.org/10.1126/science.1217330

51.

Wasserman

Faust

(1994). Social network analysis: Methods and applications (Vol. 8). Cambridge University Press. https://doi.org/10.1017/CBO9780511815478

52.

Williams

D. R.

Mohammed

S. A.

(2013). Racism and health I: Pathways and scientific evidence. American Behavioral Scientist, 57(8), 1152–1173. https://doi.org/10.1177/0002764213487340

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB