Combating HIV/AIDS with evidence: increasing self-testing in Uganda
Context
HIV/AIDS is widespread in Uganda, with 5.7 per cent of the adult population (15–49 years) living with the virus. The UNAIDS 90-90-90 target set in 2014 envisioned that, by 2020, 90 per cent of the people living with HIV would know their HIV status. Couples testing plays a critical role in this vision and was first recommended by the World Health Organization (WHO) and strongly supported by UNAIDS in the 2012 guidelines on couples HIV testing and counselling. Yet, testing rates amongst couples have been persistently low.
According to estimates from the 2011 Uganda AIDS Indicator Survey, only 16 per cent of the couples had been tested together. Most women in Uganda get tested for HIV when they are pregnant, and are provided with antiretroviral drugs for prevention of mother-to-child transmission. However, male partners seldom get themselves tested for HIV when accompanying their partners to antenatal care centres. Given this male testing gap in Uganda (and elsewhere in Sub-Saharan Africa), in 2016, WHO recommended HIV self-testing (HIVST) as a potentially promising approach to improve the uptake of testing amongst partners.
Uganda’s health ministry was keen to gather evidence on the effectiveness and acceptability of HIVST before promoting the approach. In 2015, after learning about the 3ie evidence programme on HIVST, the ministry expressed interest in collaborating on a study to understand how HIV testing services (HTS) delivery approaches could be improved in Uganda. Following that, 3ie supported researchers from Makerere University, Mildmay Uganda and Medical University of South Carolina, amongst others, to provide rigorous and timely evidence to inform the national HTS guidelines.
This evaluation, which began in 2016, examined whether HIV oral self-testing kits delivered to partners through pregnant women attending antenatal care would increase the uptake of HIVST and knowledge of HIV status. The evaluation also aimed to ascertain if this approach could eradicate supply- and demand-side barriers to couples testing.
Cluster randomised controlled trials were conducted amongst pregnant women visiting antenatal care centres at Mpigi Health Centre Level 4, and at Entebbe and Nakaseke hospitals. Clinic days were randomised into two arms, intervention days or control, with 1,618 women involved in the pilot intervention. Participants in both arms received information and education around HIV and were encouraged to bring their male partners to the clinic for testing. Participants in the intervention group received HIVST to deliver to their partners for testing.
When Uganda’s health ministry launched national HIV testing guidelines in January 2017, it provided no recommendations on HIVST but noted that pilot studies, including the 3ie study, were testing HIVST uptake and that the evidence from those studies would inform the scale-up of HIVST and the HTS programme in general. Further, it was remarked that if the study findings were positive, they may inform decisions on HIVST and adjustments to the policy of 2016.
Evidence
The evaluation showed that the provision of HIV oral self-testing kits led to a dramatic increase in partner and couples HIV testing. The intervention was associated with a 38.5 per cent increase in HIV testing as compared to the control group (T = 76, C = 37.5). Moreover, more couples in the intervention group tested together than in the control group.
Findings also demonstrated the effectiveness of HIVST in identifying the numbers of male partners living with HIV. It is well acknowledged that testing is an important first step in HIV treatment, care and prevention. Based on the reports given by women and men, 42 male partners in the intervention group were identified as HIV-positive, while only 11 were identified in the control group. The results also suggested that the intervention was cost effective, as compared to the control group. The per partner tested cost was US$30.30 for the intervention group and USD$31.20 for the control.
However, the identification of HIV-positive male partners through HIVST did not automatically lead to medical clinic visits after HIV diagnoses, suggesting that more interventions are needed to enhance links with HIV care centres. Reasons identified for the gap included unwillingness to commit for a long term, difficulty in accessing the clinic (travel, wait, opportunity costs) and privacy.
Evidence impacts
Type of impact: Change policies or programmes
Decision makers use findings from an evaluation or systematic review to adjust their programming to fix targeting, cash transfer amounts, training modules or other factors that inhibit the policy or programme’s ability to achieve its intended impacts.
This is one of 3ie’s seven types of evidence use. Impact types are based on what we find in the monitoring data for an evaluation or review. Due to the nature of evidence-informed decision-making and action, 3ie looks for verifiable contributions that our evidence makes, not attribution.
Read our complete evidence impact typology and verification approach here.
Close windowPositive evaluation findings, extensive engagement with key health ministry officials and technical working groups, promising evidence from 3ie studies conducted in other Sub-Saharan African countries, and timely release of evaluation results all contributed to evaluation recommendations that formed a part of the 2018 Addendum to the National HIV Testing Services Policy and Implementation Guidelines issued by the ministry. The purpose of the addendum is to supplement the 2016 policy and implementation guidelines. The addendum also aims to provide alternative testing approaches that can improve access to HIV testing for high-priority populations and citizens who might not get tested otherwise.
'After the study results were out, dissemination at national level took place. Then around July 2017 we started working on developing an addendum. The addendum was then going to be operationalised and provide guidance for self-testing, implementation. And guess what? We were very comfortable to start, the models that we have studied already. So you know, [in the] guidelines we stated, different models for distribution at different channels, but because we're very comfortable having proven (the) ANC (antenatal care) concept, it was our very first model in distribution, which is still the biggest channel for distribution.'
Type of impact: Change policies or programmes
Decision makers use findings from an evaluation or systematic review to adjust their programming to fix targeting, cash transfer amounts, training modules or other factors that inhibit the policy or programme’s ability to achieve its intended impacts.
This is one of 3ie’s seven types of evidence use. Impact types are based on what we find in the monitoring data for an evaluation or review. Due to the nature of evidence-informed decision-making and action, 3ie looks for verifiable contributions that our evidence makes, not attribution.
Read our complete evidence impact typology and verification approach here.
Close windowIn 2018, with the guidelines in place, the health ministry initiated health facility-based distribution of oral self-testing kits in 21 priority districts as part of a phased rollout of HIVST in Uganda. In 2019, around 400 antenatal care facilities delivered HIVST kits to their clients.
Suggested citation
International Initiative for Impact Evaluation (3ie), 2020. Combating HIV/AIDS with evidence: increasing self-testing in Uganda [online summary], Evidence Impact Summaries. New Delhi: 3ie.
Related
Preparation meets opportunity: how 3ie’s stakeholder engagement paid off on HIV self-testing
This blog highlights how Uganda’s government engaged with the evidence from 3ie’s HIV self-testing evidence programme.
Evidence impact summaries aim to demonstrate and encourage the use of evidence to inform programming and policymaking. These reflect the information available to 3ie at the time of posting. Since several factors influence policymaking, the summaries highlight contributions of evidence rather than endorsing a policy or decision or claiming that it can be attributed solely to evidence. If you have any suggestions or updates to improve this summary, please write to influence@3ieimpact.org
Last updated on 13th November 2020