High uptake of HIV self-testing by female sex workers in African countries

A year ago there was virtually no evidence on the acceptability and feasibility of HIV self-testing in female sex workers, but a series of presentations from Zimbabwe, Zambia, Uganda and Kenya at the 9th International AIDS Society Conference on HIV Science (IAS 2017) in Paris, France, this week suggest that self-testing has potential in improving the uptake of testing among sex workers in Africa.

In follow-up questionnaires, 100% said the test was easy to use, 100% trusted the result they had been given and 98% were comfortable learning the result without someone else present.

Self-testing might have advantages for sex workers of privacy and flexibility, allowing women to test themselves at a time of their choosing, without needing to fit in with the hours of clinical services, or to interact with healthcare workers who may have stigmatising attitudes. Its convenience may also make regular, repeat testing more feasible.

Policy makers have great interest in self-testing because they hope it may be a more acceptable option for groups of people who have limited engagement with existing HIV testing services. The proportion of people living with HIV who remain undiagnosed is particularly high in some groups and sub-populations, including female sex workers.

Sue Mavedzenge of RTI International reported on work she has done in Zimbabwe. Female sex workers are disproportionately affected by HIV (prevalence over 50%, compared to 14% in the general population), but their engagement with health services does not reflect their heightened risk. Of all sex workers living with HIV, only 64% are aware of their status, 43% are taking HIV treatment, and 34% are virally suppressed. Increasing HIV testing is the necessary first step to better engagement with both prevention and treatment services.

When self-testing was given as an option to around 600 women seeking HIV testing in a dedicated sex worker clinic, 54% chose the self-testing option. Self-testing doesn’t need to take place at home, and in fact 96% of those using the self-testing kit did so in a private room at the clinic. This gives the possibility of seeking staff assistance in case of difficulty or of dealing with a reactive result – which 30% of those testing did have.

In follow-up questionnaires, 100% said the test was easy to use, 100% trusted the result they had been given and 98% were comfortable learning the result without someone else present. Substantial proportions expressed a willingness to pay for testing kits (for example, 36% said they would be willing to pay under USD$1 and 27% would be willing to pay between USD$2 and USD$3).

Of those testing positive, 99% had connected with follow-up services within two weeks of the test.

What’s the best way to distribute self-test kits to sex workers? Among the Zimbabwean respondents, there were preferences for distribution through clinics (62%), pharmacies (18%), peer networks (14%) and workplaces (13%). However this was a sample of sex workers recruited through clinics and there was also recognition that not all sex workers are engaged with clinics.

Distribution models in Uganda and Zambia

Two randomised controlled trials, conducted in Uganda and Zambia, explored distribution methods in more detail. While the studies were done separately they had virtually identical methods.

Study staff recruited peer educators, who were current or former female sex workers working in sex worker organisations. The peer educators then recruited up to eight study participants each; to be eligible, participants had to have exchanged sex for money or goods in the past month and not be diagnosed with HIV.

Once a group of participants had been recruited they were randomised, as a group, to one of three study arms:

• Direct distribution – the peer educator provides a self-testing kit to participants, both at the beginning of the study and three months later.
• Coupon – the peer educator provides a coupon which participants can take to a drug store or health clinic and exchange for a self-testing kit, both at the beginning of the study and three months later.
• Standard of care – the peer educator provides information about existing HIV testing services.

In each country, just under 1000 female sex workers were recruited as study participants. While women who had tested for HIV in the past three months were excluded, testing rates were relatively high before the intervention was provided – in each country, around two-thirds had tested within the past year. It is possible that the researchers could have demonstrated a greater impact of self-testing if they had recruited women with less familiarity with HIV testing.

When the studies completed after four months, testing rates were very high in all study arms. In Uganda, 100% of women in the direct distribution arm had taken a test in the previous four months, compared to 97% of women in the coupon arm and 87% of those in the standard of care arm. In Zambia, the comparable figures were 95, 84 and 89% respectively. While the differences in Uganda were statistically significant, they were not in Zambia.

Based on the Ugandan data, Katrina Ortblad concluded that either model of self-testing achieved higher uptake than the standard of care, but that direct distribution by peer educators performed better than the coupon system.

However, in both studies, linkage to care appeared to be a little poorer in the self-testing arms, although these differences were not statistically significant.

Although intimate partner violence (IPV) is common in both these study populations, there were only five cases of IPV that were linked to self-testing recorded in the two studies.

Text message promotion to female sex workers and to truckers

A randomised controlled trial in Kenya tested the impact of a very simple text message intervention on the uptake of self-testing by female sex workers – and by truck drivers, many of whom may be their clients.

The setting was the North Star Alliance network of healthcare clinics on major transport routes in Kenya. Using the clinics’ database of electronic patient records, the researchers sent text messages (in English and Kiswahili) to people who had not recently tested and were either female sex workers or male truckers. In the intervention arm, the message informed clients about the availability of self-testing at North Star Alliance clinics. In the standard of care arm, the message simply encouraged clients to test for HIV.

Among sex workers, 6.1% in the control group and 15.9% in the intervention group took an HIV test in the two months after text messages were sent. This represents a statistically significant three-fold increased odds of taking a test.

While the male truckers were much less likely to test, the intervention also increased their odds of taking a test three-fold. Among truckers, 1.3% in the control group and 4.1% in the intervention group took an HIV test.

The right test?

Each of these studies employed the OraQuick HIV Self-Test, an oral-fluid based assay which this week became the first self-testing device to be prequalified by the World Health Organization. However, during questions and answers, Barry Kosloff of the London School of Hygiene and Tropical Medicine asked whether this is the ideal testing device to promote to individuals at high risk of having a recently acquired HIV infection – such as female sex workers.

The window period of OraQuick is relatively long, meaning that recent infections could be missed. The development and licensing of self-test kits that are more sensitive to recent HIV infection may be necessary if self-testing is truly to meet sex workers’ needs.