Defense Date


Document Type


Degree Name

Doctor of Philosophy


Healthcare Policy & Research

First Advisor

Dr. April D. Kimmel

Second Advisor

Dr. Bassam Dahman

Third Advisor

Dr. Daniel Nixon

Fourth Advisor

Dr. Tilahun Adera


In Eastern and Southern Africa, hard-to-reach populations (e.g., long distance truck drivers and female sex workers), defined as populations that are difficult to interact or engage with due to their unique behaviors and characteristics, are disproportionately affected by the HIV epidemic and are at high-risk of acquiring and transmitting HIV. Further, these populations have substantially low uptake of HIV testing services, and those that have been diagnosed with HIV and on antiretroviral therapy experience high loss-to-follow-up from treatment programs.

Hard-to-reach populations face unique barriers in accessing and utilizing routine HIV care such as provider stigmatization towards sex workers and highly mobile nature of their occupations. Innovative and targeted strategies, which may be resource-intensive, are required to improve their engagement and retention in care. Evidence on cost-effective strategies to improve HIV testing uptake and to reduce loss to follow-up from HIV treatment programs in hard-to-reach populations in Eastern and Southern Africa remains limited.

This dissertation is comprised of three papers examining the cost-effectiveness of HIV testing and loss to follow-up strategies among hard-to-reach populations in Eastern and Southern Africa, using female sex workers and long-distance truck drivers as case study populations and Kenya as a case study setting. In paper one, I conducted a trial-based cost-effective analysis of offering the choice to HIV self-test compared to provider-administered HIV testing among long-distance truck drivers in Kenya. Paper two extended the analysis for paper one by examining the cost-effectiveness of a broad range of alternative HIV testing strategies among hard-to-reach populations in Eastern and Southern Africa using a lifetime Markov model. Seven strategies were examined: i) No testing, ii) voluntary counseling and testing, iii) provider-initiated and -administered testing, delivery of: iv) self-testing kits, v) self-testing coupons, and vi) HIV testing referral cards in the community using peer-educators, and vii) offering a choice of self-testing at the health facility. In paper three, I applied the same Markov model from paper two to examine strategies to prevent loss to follow-up among female sex workers on antiretroviral therapy in Eastern and Southern Africa. Strategies included: 1) No intervention; 2) Home ART delivery using community-health workers; 3) Home ART delivery using community-health workers plus monthly nutrition supplement; 4) physical and phone-tracing of patients that miss an appointment plus transport refund to the health facility; 5) physical and phone-tracing with free medical care for opportunistic infections; 6) free medical care for opportunistic infections with transport refund to the health facility and free breakfast. Data for paper one came from a randomized controlled trial (n=150, intervention; n=155, control), while data for paper two and three came from peer-reviewed and grey literature. All costs were reported in 2017 international dollars in paper one and 2017 US dollars for paper two and three.

Findings from these studies suggest that investing resources in strategies that offer choices in HIV testing approaches such self-testing at the health facility or in communities using peer educators would improve HIV testing uptake and reaching out to patients on treatment in their communities to deliver them ART drugs may improve retention in ART programs in Eastern and Southern Africa. In paper one, I found that offering a choice of HIV self-testing at the clinic was cost-effective compared to only the provider-administered HIV testing with an incremental cost-effectiveness ratio (ICER) equal to $163. In paper two, delivery of HIV self-testing kits in the community using peer educators was cost-effective (ICER < $600) in both truck drivers and female sex worker sub-populations. Finally, in paper three, delivery of antiretroviral therapy drugs to female sex workers in the community was cost-effective (ICER < $500).


© Deo Mujwara

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission