Informing Strategies for Effective HIV Treatment and Prevention

Rachel Silverman | 2018

Advisor: Grace C. John-Stewart

Research Area(s): Infectious Diseases, Public Health Practice


HIV continues to challenge individual and public health throughout the world. Ensuring HIV-infected and high-risk individuals are aware of their infection status, engage in care, and receive effective treatment and prevention tools as early as possible is key to improving health outcomes, disrupting transmission, and reducing the burden of disease globally. To best combat the HIV epidemic, it is crucial to determine if guidelines and practices are effective, understand barriers to successful implementation of recommendations, and ascertain how to optimize interventions. One barrier to effective HIV treatment is pre-antiretroviral treatment (ART) drug resistance (PDR). PDR to first line regimens has shown to contribute to treatment failure, which is especially problematic in setting where resistance testing is unavailable and alternative regimens are limited. Understanding PDR prevalence and associated risk-factors can inform policy to best manage care and ensure treatment effectiveness. Additionally, despite efforts to treat infected individuals early in the course of disease progression, many seek treatment only when they are symptomatic at later stages of infection. Investigating the extent to which individuals seek treatment late in disease progression and factors associated with early mortality can help programs target high-risk individuals, address contributing factors, and improve health outcomes. While interventions can successfully improve diagnosis rates, treatment uptake, and prevention, it is important to evaluate effective strategies to ensure they are optimized, especially when resources are limited. To investigate the prevalence and risk factors of PDR in a low resource setting in Kenya, we conducted a cross-sectional analysis of ART-eligible HIV-infected participants in 2013-2014. We found an overall PDR-prevalence of 10%, and the highest prevalence (22%) among ARV-naïve women 18-24 years old. Given our observed PDR-prevalence, resistance-testing and/or alternative ARVs may be warranted in high HIV prevalence settings, with special attention to young women. We also conducted a longitudinal survival analysis to estimate incidence and identify risk factors of early mortality following ART initiation at similarly managed rural and urban clinics in Kenya. We found that the rural clinic had a 2-fold greater risk of mortality than the urban clinic (unadjusted hazard ratio 2.20; 95% CI 1.29-3.76; P = 0.004), despite a lower average baseline CD4 count among the urban cohort. Mortality risk associated with low CD4 count, low BMI, and PDR was also greater in the rural setting. Other statistically significant (p<0.05) predictors of mortality included male gender, older age, fewer years of education, unemployment, low body mass index, low CD4 count, and PDR. Efforts to engage patients, especially men, earlier in HIV infection remain critical, and interventions to improve BMI and target less educated patients could improve health outcomes, especially in rural areas with high infectious disease burdens. PDR may influence short-term mortality and further studies to optimize management will be important in settings with increasing PDR. Additionally, we evaluated an effective partner services (PS) program for bacterial sexually transmitted diseases (STDs) that includes integrated HIV-objectives (HIV-testing, pre-exposure prophylaxis promotion, and re-linkage to care) in a higher resource setting, Washington state (WA). We assessed STD PS costs, including time, and identified modifiable areas to improve efficiency at three WA health jurisdictions (King, Pierce, and Spokane counties) using activity-based micro-costing that included interviews, observational time-studies with disease intervention specialists (DIS) and other program staff, and individual case time tracking. We also collected program expenditures and analyzed surveillance and service delivery data to determine costs of program objectives. In King, Pierce, and Spokane, respectively, DIS spent approximately 6.5, 6.4, and 28.8 hours per syphilis case and 1.5, 1.6, and 2.9 hours per GC/CT case. Difficult-to-reach heterosexuals (including many reported methamphetamine drug users) comprised 68% of syphilis cases in Spokane, which increased the time required for case-finding, whereas most cases in King (88%) and Pierce (82%) were among easier to reach men who have sex with men. In all jurisdictions, time-consuming DIS activities included provider contact (9%) data entry (20%), record searches to locate cases/partners (19%), and contacting cases/partners (32%). Time spent on expedited partner therapy (EPT) for STDs and HIV-related objectives was minimal (30 seconds-5 minutes per interview). In 2016, DIS interviewed 363, 97, and 104 syphilis and 1,929, 1,948, and 280 GC/CT cases in King, Pierce, and Spokane, respectively. Cost-per-interview ranged from $516-$2,155 for syphilis, $215-$472 for GC, and $161-$533 for CT. We found that STD PS resource needs depend on epidemic characteristics and program models. Electronic reporting and medical records access could improve efficiency. Integrating HIV-objectives minimally impacted STD PS specific program costs. These data are necessary for program planning, budget impact analysis, and estimating the cost-effectiveness of PS. Overall, the work presented in this dissertation contributes to our understanding of barriers and strategies for successful HIV treatment and prevention and can be used to improve health outcomes and combat the HIV epidemic in Kenya and the United States.