Medical Mistrust May Mediate the Relationship Between Perceived
Medical Mistrust May Mediate the Relationship Between Perceived Discrimination and Adherence in Latino Men Laura M. Bogart 1, Frank H. Galvan 2, Glenn Wagner 3, Lori Mizuno 4, & David J. Klein 1 1 Children’s Hospital Boston/Harvard Medical School 2 Charles Drew University of Medicine and Science 3 RAND Corporation 4 Bienestar Human Services, Inc.
Background: Racial/Ethnic Disparities and Adherence • Many Latinos living with HIV have poor adherence to antiretroviral therapy In the 14 -site MACH 14, adjusted electronically monitored 12 -month adherence was 65% for Latinos vs. 71% for Whites (Simoni et al. , 2010) o In the 4 -city MACS cohort, Latinos were 2. 16 times as likely than Whites to not report 100% adherence in multivariate analyses [95% confidence interval (CI) = 1. 47 -3. 18] (Lee et al. , 2009) o
Background: Discrimination and Health • Latino men who have sex with men (MSM) living with HIV may experience discrimination due to multiple stigmatized characteristics o Due to, e. g. , ethnicity, HIV-sero-status, sexual orientation • Discrimination is related to poor health behaviors (Pascoe & Richman, 2009) o Racial discrimination is related to lower adherence over time among African American MSM living with HIV (Bogart et al. , 2010)
Adherence Percentage by Racial Discrimination Among 152 African American MSM with HIV Bogart et al. 2010
Background: Discrimination and Health • Discrimination may foster mistrust of public health institutions, including suspicion of medical information and treatments o This can lead to reluctance to adhere to health care provider recommendations Nonadherence Discrimination Mistrust
Research Goal • To examine associations of discrimination and medical mistrust with HIV treatment nonadherence among Latino MSM living with HIV o Medical mistrust was hypothesized to mediate the relationship between perceived discrimination and nonadherence
Methods: Participants • 208 HIV-positive Latino men on antiretroviral treatment • Recruited and surveyed by staff at Bienestar o o AIDS service organization with 9 locations in S. CA Provides culturally relevant services for Latinos living with HIV
Methods: Procedure • Surveyed using an audio computer-assisted interview (ACASI) in either Spanish or English • Electronic adherence data (MEMS) downloaded one -month post-baseline • Incentives: $30 baseline, $30 one-month follow-up
Methods: Multiple Discrimination Scale • Participants reported experiences with 10 different discrimination events in the past year • All 10 discrimination events asked separately for ethnicity, HIV-serostatus, & sexual orientation (30 questions total) o e. g. , “In the past year, were you physically assaulted or beaten up because someone thought that you were gay? ” (yes/no) • Summed discrimination events by subscale o o o HIV-serostatus ( =. 87) Ethnicity ( =. 86) Sexual orientation ( =. 86)
Methods: Medical Mistrust • 8 -item Medical Mistrust scale ( =. 79) (La. Veist et al. , 2000) o Sample item o Response options § Patients have sometimes been deceived or misled at hospitals § 1, Strongly Disagree; 2, Disagree; 3, Agree; 4, Strongly Agree
Methods: Self-Reported Adherence • Many patients find it difficult to take all their HIV medications exactly as prescribed. How many doses of your HIV medication did you miss in the last 7 days? ____ doses (Golin et al. , 2002; recommended in Simoni et al. , 2006) o o Recoded as “any” doses missed in last 7 days Associated with detectable viral load, r =. 14, p <. 05
Results: Socio-Demographic Characteristics • • Mean age (SD) = 45 (9) years 34% <$5, 000 annual income 36% working full/part time 86% stable housing o Owned/rented home or in subsidized housing vs. homeless/marginally housed • 81% MSM o Of those, 7% (n = 11) identified as heterosexual
Results: Socio-Demographic Characteristics (cont’d) • 82% speak Spanish “always” or “almost always” • 50% undocumented residency status • 49% lived in US >20 years
Results: Any Discrimination in Past Year % *Sexual orientation discrimination for MSM only
Results: Medical Mistrust • Medical Mistrust Scale: o o 86% agreed with at least one mistrust belief item Average (on 1, disagree, to 4, agree) = 2. 1 (SD = 0. 7)
Results: Adherence • Self-reported adherence: o 23% reported missing a dose in the 7 days prior to baseline o o o Mean = 89% of prescribed doses taken (SD = 17%) Range: 0% - 100% of doses taken 71% were adherent ≥ 90% of the time • Electronically monitored adherence in month since baseline:
Results: Mediation Analysis • Bivariate regressions for all three pathways • Bootstrapping used to calculate confidence interval for indirect effect of perceived discrimination on nonadherence (Preacher & Hayes, 2008) • Multivariate analysis (sexual orientation as covariate) Self-reported nonadherence Discrimination Mistrust
Results: Bivariate Tests • Discrimination was related to nonadherence Discrimination Type Unadjusted OR (CI) Ethnicity 0. 85 (0. 74 – 0. 97)* HIV-serostatus 0. 83 (0. 72 – 0. 95)** Sexual orientation 0. 84 (0. 74 – 0. 97)* * p <. 05; ** p <. 01
Results: Bivariate Tests • Discrimination was related to medical mistrust Discrimination Type Ethnicity Unadjusted b (SE) 0. 09 (0. 02)*** HIV-serostatus 0. 06 (0. 02)* Sexual orientation 0. 06 (0. 02)* * p <. 05; ** p <. 01; *** p<. 001
Results: Bivariate Tests • Medical mistrust was related to nonadherence o OR (CI) = 0. 42 (0. 26 – 0. 69), p <. 001
Results: Mediation • Medical mistrust fully mediated the relationship between ethnic discrimination and nonadherence • The relationship between ethnic discrimination and nonadherence was nonsignificant when mistrust was added Ethnic Discrimination OR = 0. 86 (0. 75 -0. 99), p =. 04 OR = 0. 91 (0. 79 -1. 06), p =. 23 Nonadherence 95%CI = -. 14, -. 02 Mistrust
Results: Mediation • Medical mistrust fully mediated the relationship between sexual orientation discrimination and nonadherence • The relationship between sexual orientation discrimination and nonadherence was nonsignificant when mistrust was added Sexual Orientation OR = 0. 860 (0. 750 -0. 996), p =. 04 Discrimination OR = 0. 90 (0. 78 -1. 04), p =. 16 Nonadherence 95%CI = -. 10, -. 01 Mistrust
Results: Mediation • Medical mistrust fully mediated the relationship between HIVserostatus discrimination and nonadherence • The relationship between HIV-serostatus discrimination and nonadherence was nonsignificant when mistrust was added HIV-Serostatus Discrimination OR = 0. 84 (0. 73 -0. 97), p =. 01 OR = 0. 87 (0. 76 -1. 01), p =. 06 Nonadherence 95%CI = -. 10, -. 01 Mistrust
Discussion • Culturally relevant interventions for Latinos should address discrimination and medical mistrust Community-based interventions can address mistrust by openly acknowledging discrimination and reasons for mistrust, and by encouraging adaptive coping with discrimination o Healthcare-based interventions, such as treatment navigation and advocacy, can help to overcome language barriers and mistrust o • Further research should test effects with electronic monitoring over longer time-periods
Acknowledgments • Funding o National Institute of Mental Health R 01 MH 072351 (supplement; Cynthia Grossman, Program Officer) • Bienestar Partners o Gustavo Arguelles, Robert Contreras, Oscar De La O, Victor Martinez, Silvia Valerio • Charles Drew Collaborator o Argelis Ortiz
- Slides: 25