Graduation Year

2024

Document Type

Dissertation

Degree

D.P.H.

Degree Name

Doctor of Public Health (Dr.PH.)

Degree Granting Department

Epidemiology and Biostatistics

Major Professor

Janice Zgibor RPh, PhD, CPH, FACE

Co-Major Professor

Roneé Wilson, PhD, MPH, CPH

Committee Member

Henian Chen, MD, PhD

Committee Member

Stephanie Marhefka-Day, PhD

Committee Member

Judith Rijnhart, PhD

Committee Member

Emma Spencer, PhD, MPH

Keywords

epidemiology, observational studies, survival analysis, mediation analysis

Abstract

Virologic suppression is an essential part to a successful regimen for Persons with HIV (PWH) and the 95-95-95 initiative. Yet, PWH experiencing episodes of low-level viremia (LLV) may be more prone to experience virological failure (VF). Currently, studies debate the clinical relevance of LLV and its effect on virological outcomes, but none have explored its interconnectedness with drug resistant mutations (DRM) and antiretroviral treatment (ART) adherence – all integral elements in predicting virological outcomes.

The goal of the proposed study was to contribute to the knowledge gap regarding the interconnectivity of drug resistant mutations, low-level viremia, adherence, and virological failure by conducting a study with all four elements and a mediation analysis; both items not yet capitalized in this research area.

The study used 2015 to 2022 individual-level data from the Florida Department of Health HIV surveillance databases and programs to conduct retrospective cohort studies using survival and casual mediation methods to explore the associations between DRMs, LLV, adherence, and VF.

The results from the survival analysis suggest LLV is protective against VF and that adherence may be an effect modifier, however, the estimates when including adherence were unreliable due to the large confidence intervals from the estimates. The analyses exploring LLV as a potential mediator on the casual pathway between pre-treatment DRMs and VF suggest LLV is not a mediator. Additionally, the data showed about 40% of PWH in the state of Florida do not have a baseline resistance test upon HIV diagnosis.

The results warrant additional research to unravel the protective relationship shown between LLV and VF found in this project. These results indicate the need for casual mediation analyses in this area of research and advanced statistical methods that may account for mediation, moderation, time-varying variables, and time-dependent coefficients. Furthermore, the considerable gap in baseline resistance testing should be noted and public health policy makers and clinicians should push for increased baseline testing.

Download

Included in

Epidemiology Commons

Share

COinS