Dissertations and Theses

Date of Degree


Document Type


Degree Name

Doctor of Public Health (DPH)


Epidemiology and Biostatistics


Elizabeth Kelvin

Committee Members

Katarzyna Wyka

Christian Grov

Kathrine Meyers

Denis Nash

Subject Categories

Community Health and Preventive Medicine | Epidemiology | Health Services Research | Immune System Diseases | Public Health


HIV, Pre-Exposure Prophylaxis, multi-level modeling, MSM


Background: Men who have sex with men (MSM) remain the group most heavily affected by HIV in the United States (US), with MSM of color further disproportionately represented. In July 2012, the US Food and Drug Administration (FDA) approved the use of tenofovir disoproxil fumarate plus emtricitabine (TDF-FTC, Truvada) for use as pre-exposure prophylaxis (PrEP) in HIV-negative adults who are at high-risk of HIV infection. Despite its effectiveness, PrEP prescriptions are reaching only a small proportion of those who could benefit from the drug and prescription rates vary both by race and geographic region. The goal of this dissertation was to explore the complex relationship among multilevel predictors of PrEP use in a geographically diverse sample of MSM in the US.

Methods: Data from a cross-sectional survey of MSM residing in all 50 US states was used to test a series of hypotheses regarding PrEP use. Using the framework of the social-ecological model and multi-level logistic and linear regression we explored if (a) variables at the individual- (i.e. demographics, identity), social- (MP factors and social behaviors), and state-levels (state equality, HIV prevalence, socio-economic factors) influenced the outcome (Aim 1 and 2: PrEP use and Aim 3: Patient Trust in Physician) among MSM with a focus on the equality of the state (LBGTQ+ and racial equality); we further extended this hypothesis to examine if our state equality measures modified the association between individual identity-level variables (sexual and racial/ethnic identity) and our outcome (PrEP use or PTP scores), and (b) if the association between state-level (LGBTQ+ and racial equality) and individual-level (sexual and race/ethnic identity) variables and PrEP use was mediated through patient trust in physician (PTP). We examined these hypotheses among all participants, and then conducted a sub-analysis looking at the first hypothesis, among only MSM of color, as this group currently has the highest HIV incidence in the US and we felt warranted a closer look in an attempt to tease out reasons for this high risk.

Results: Aim 1 included 4,165 HIV-negative MSM of whom 13.4% (N = 560) were taking PrEP at the time of the survey. In Aim 2, there were 1,465 HIV-negative MSM of color, of whom 13.6% (N = 199) were taking PrEP. In the analyses for both aims, we found that age, several main partner (MP) factors, number of sexual partners, STI history, and the LGBTQ+ equality of the state where the participant resides were all significantly associated with higher odds of using PrEP, while sexual identity and other main partner (MP) factors were associated with lower use. In aim 1 alone, higher education and a mid-level HIV prevalence were also associated with higher PrEP use. Lastly, only in aim 2, race/ethnicity was shown to be associated with PrEP use, with Hispanic and Asian MSM being less likely to use PrEP.

In aim 3, of the 2,750 HIV-negative participants who reported having a primary care provider, there were several individual- and social-level factors associated with PTP scores. Geographic region, sexual identity, and individual drug use were shown to be associated with lower PTP scores, while participants who reported PrEP use were shown to have higher PTP scores. The interaction analysis suggested that racial/ethnic identity influenced PTP differently depending on the racism level at the state-level. After stratifying on the racism/racial equality variable of the state, the results suggest, that with the exception of black participants, compared to whites, all other races had higher PTP scores in high racial equality states (less racism) than in low racial equality states (more racism).

Conclusions: This is one of the first studies that explored several multi-level predictors of PrEP use, and given our large sample, it also offered a unique opportunity for us to examine the influence of state-level factors on PrEP use, as both a main effect and as an effect modifier. The results of our work suggests that individual-, social-, and state-level variables, specifically the LGBTQ+ equality of the state of residence, are associated with PrEP use. The findings of this study suggest that policies that increase inequality may have an impact on HIV prevention interventions.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.