Zambian Peer Educators for HIV Self-Testing: a randomized study of testing provision to female sex workers in Zambia

Publication Details

Chanda, MM, Ortblad, KF, Mwale, M, Chongo, S, Kanchele, C, Kamungoma, N, Fullem, A, Bärnighausen, T, and Oldenburg, CE, 2018, The Zambian Peer Educators for HIV Self-Testing (ZEST) Study: a randomized controlled trial of HIV self-test provision for female sex workers in Zambia, 3ie Grantee Final Report. New Delhi:  International Initiative for Impact Evaluation (3ie)


Link to Source
Author
Michael M. Chanda, Katrina F. Ortblad, Magdalene Mwale, Steven Chongo, Catherine Kanchele, Nyambe Kamungoma, Andrew Fullem, Till Bärnighausen, Catherine E. Oldenburg
Institutional affiliations
None specified
Grant-holding institution
None specified
Country
Zambia
Region
Sub-Saharan Africa (includes East and West Africa)
Sector
Health Nutrition and Population
Subsector
HIV and AIDS
Gender analysis
 
Subsector
HIV and AIDS
Gender analysis
 
Equity Focus
None specified
Evaluation design
Randomised Control Trials (RCT), Mixed Methods
Status
3ie Final Grantee Report
3ie Funding Window
HIV self-testing thematic window

Synopsis

In three border towns, an established peer educator delivers the HIV self-testing intervention to female sex workers (FSWs) to increase their awareness of status and linkage to care for those who test positive.

Context

HIV prevalence among FSWs is 13.5 times higher than that of women aged 15-49 years in the general population globally, and 5 to 15 times higher in other southern African countries. However, HIV testing coverage in the country remains below national and international targets among FSW.

Research questions

  • Does the distribution of HIV self-test (HIVST) kits by peer educators increase HIV testing uptake for female sex workersin three Zambian border towns?
  • Does the distribution of HIVST kits by peer educators lead to an increase in knowledge of HIV status for FSWs?
  • Does the distribution of HIVST kits by peer educators reduce time to linkage to care for FSWs?
  • Does the distribution of HIVST kits at fixed distribution points increase HIV testing uptake for these FSWs?
  • Does the distribution of HIVST kits at fixed distribution points lead to an increase in knowledge of HIV status for FSWs?
  • Does the distribution of HIVST  kits at fixed distribution points reduce time to linkage to care for FSWs?

Methodology

Intervention design

The pilot intervention consisted of the distribution of HIVST for FSWs in three of the country’s busiest border towns, Livingstone, Chirundu, and Kapiri Mposhi through two distinct mechanisms: 1) fixed distribution points (i.e. drug stores or participating healthcare facilities) following referral with a coupon from peer educators, and 2) direct distribution from peer educators.

Peer educators recruit FSWs and provide four intervention visits. The visits include HIV risk reduction counselling, condom distribution, and information on where to get HIV testing. Peer educators were current or former sex workers who were recruited by sex work organizations operating in each of the study communities.  In the delivery arm, peer educators distributed two HIVST kits: one at the first peer educator visit, and a second one three months after the first peer educator visit. Peer educators were trained on use of the oral HIVST and shared this information with participants. To preserve participant confidentiality, there was no HIV status requirement for distribution of the second HIVST kit.

In the fixed distribution point arm, peer educators distributed coupons which participants could use to collect an HIVST kit at one of several participating distribution sites, which were health clinics or pharmacies. Existing staff were briefly trained on study procedures and the use of the HIVST. Participants were required to bring the coupon to the distribution site, which was exchanged for a single HIVST. While the coupon did not contain any information about the study, staff members at the distribution sites were aware that the study was specifically for sex workers. As with the delivery arm, peer educators distributed one coupon at the first peer educator visit and a second three months after the first peer educator visit. The content of the test and instructions provided to participants were identical. As with the delivery arm, there was no HIV status requirement for distribution of the second coupon.

Peer educators in all arms provided information about existing HIV testing services and information about where to get a confirmatory test and link to care if they tested positive. While peer educators were available should participants have questions or need support, participants tested for HIV at a time and place of their own choosing and were not required to disclose their status to anyone.

Theory of change

The authors theorised that the distribution of HIVST kits via peer educators would lead to improved status knowledge by reducing barriers to HIV testing such as stigma or hours of clinic operation. Enacted or perceived sex work stigma from healthcare providers and from the community may be addressed by HIVST, by allowing individuals to test for HIV in private without fear of being seen in the clinic and without fear of judgment from providers. This would lead to improved uptake of HIV testing, which would lead to improved knowledge of status and ultimately reduce time to linkage to care. However, it is also possible that a community-based intervention such as HIVST could be unsuccessful if individuals are concerned about others discovering their HIV status.

As a whole, the intervention is expected to increase awareness of status and linkage to care for those who test positive. This will ultimately contribute to overall improved health outcomes and reduced HIV transmission in these communities.

Evaluation design

Using a cluster randomised trial design, a total of 160 peer educator and participant groups were randomised to one of the three study arms: 1) direct HIV self-test distribution to FSW, 2) referral via coupon to HIVST distribution points in the town (drug stores and health centres), or 3) referral to fixed HIV testing points in clinics in town. All analyses were intention-to-treat. The primary outcome of interest was the proportion of participants reporting testing for HIV in the previous one month as measured at the one-month visit. To estimate risk ratios, the authors used a mixed-effects generalised linear model with a Poisson distribution, log link, and robust error term (Zou 2004), with a fixed effect for randomisation arm and study site and a random effect for peer educator group. Secondary analyses were done with dichotomous variables, including past one-month testing at the four-month visit, correct knowledge of HIV status, linkage to care and use of ART were estimated with an identical model. While, the use of the HIVST kit was compared between the two intervention arms (delivery and coupon).

Main findings

  • FSWs used HIVST when made available directly via delivery from a peer educator or when required to collect them at existing health system delivery points (e.g. clinics and pharmacies)
  • There was no difference in HIV testing among participants randomized to standard testing in existing facilities (88.5%), peer delivery (94.9%), or facility collection (84.4%)
  • Among participants reporting an HIV positive self-test, linkage to HIV-related care and ART initiation were non-significantly lower in the HIV self-testing arms (71.6 in delivery and 76.6% in fixed site) compared to the standard of care arm (85.7%%), although there was limited power to detect differences
  • HIV self-testing appears safe, acceptable, and acceptable for female sex workers in the country’s transit towns

Cost efficiency analysis

The cost per participant in the standard of care arm was $40.39. In the delivery arm, the cost per participant was $53.70, and the cost per participant in the coupon arm was $52.83.

Specific findings for policy and practice

HIV self-testing was accessible and was highly used by participants, but it did not increase HIV testing relative to referral to standard HIV testing services. Although linkage to care and ART initiation were lower in the HIV self-testing arms compared to the standard of care arm, both linkage and ART initiation increased over time. Individuals in the coupon arm were less likely to test for HIV at one month compared to those in the direct delivery arm but this difference was gone by four months, indicating that there may be some short-term barriers to HIV self-testing that reduce over time. This indicates that the delivery model may matter in the short term, but once individuals have more time to adjust to the new technology, delivery of HIV self-testing via existing health systems infrastructure may be sufficient for implementation.

Additional publications

Presentations

Media coverage

Scroll to Top