Digital Interventions for HIV Self-Testing: The feasibility of HIV self-testing in South Africa and the introduction of digital interventions to measure and optimise programmatic impact

Abstract

HIV self-testing(HIVST) allows a person to test for HIV independent a healthcare facility, and HIVST has been championed as an innovative method to improve testing and linkage to care towards the UNAID goal of 95-95-95 by 2030. The move from facilities may remove barriers around stigma and access, however, introduce challenges regarding counselling, reporting results and linkage to care. Digital health tools are used to optimise many disease verticals, and digital interventions for HIVST are being developed to address the challenges associated with HIVST.

This dissertation investigate the use of HIVSTs and whether digital interventions could be introduced to optimise programmes, by examining four research questions surrounding the usability of HIVST, the usability of digital interventions for HIVST, whether they can improve health impact, and whether they should be scaled up.

Chapter 2 investigated the first research question by evaluating the usability of seven HIVST kits in South Africans and usability was high, with a usability index of 91%. Chapter 3 conducted a cross-sectional study on the usability, sensitivity and specificity of four HIVSTs and the average usability index was high(97%), as were the sensitivity(98.2%) and specificity (99.8%) suggesting HIVSTs are suitable for the South African market.

For the second research question, chapter 4 investigated whether digital interventions could compliment traditional HIV testing services, before considering their use with HIVSTs. The SmartLink app provided HIV results and linkage to care support, and youth showed a 20% increase in linkage to care. Chapter 5 investigated the impact of COVID-19 on technology uptake, and revealed the lockdown forced 90% to use more technology, and two-thirds of participants used their phones for health information. Chapter 6 investigated the usability, acceptability and feasibility of the AspectTM HIVST app in a controlled setting, and almost all participants(98.7%) found the app easy to use. Chapter 7 investigated the real-life use of an app, Ithaka to support HIVST by self-reporting results outside a clinical environment. Participants felt Ithaka was useful and easy to use, and almost one-quarter of participants self-reported results.

For the third research question, chapter 8 investigated an updated Ithaka app to see if a digital intervention for HIVST could improve health impact. The tablet-based version was introduced into a facility-based HIVST programme, and increased testing by 25% in a pre-post evaluation.

Chapters 9 summarised findings from previous chapters to investigate the final research question on whether digital interventions for HIVST should be scaled-up and noted that although evidence is promising, most studies were pilots with small samples that lacked standardised indicators needed to prove impact at scale. Chapter 10 looked at informed consent and although the South African guidelines suggest consent forms should be below grade 8 reading level, only one-third met this threshold. Chapter 11 suggested a path to scale-up, standardisation and regulation by introducing the WHO Prequalification Programme(WHOPQ). Arguments for included WHOPQ’s strong history with HIV, trends in digital health and improved data interoperability, while arguments against included a stronger need to regulate AI/ML-enabled SaMDs first and existing barriers to WHOPQ