One in seven people offered PrEP in African study start it immediately

One in seven people (14%) offered pre-exposure prophylaxis (PrEP) in a large community study of enhanced HIV testing and treatment in Kenya and Uganda started PrEP the day they were offered it, and 18% within a month, the 9th International AIDS Society Conference on HIV Science (IAS 2017) in Paris heard last month.

Of those who were assessed by an HIV risk score algorithm to be at high risk of HIV, only 11% started PrEP within 30 days. However, people who fell outside the risk score criteria were also given the chance to request PrEP themselves, and although there were fewer in this group, nearly 40% started PrEP within 30 days, meaning that more people who self-referred ended up taking PrEP than those who were told they were at risk.

PrEP was offered within SEARCH, a large community-randomised study of enhanced HIV testing and treatment in Kenya and Uganda. Its primary purpose is to see if its enhanced programme will increase the proportion of people with HIV who are diagnosed and virally undetectable on treatment and thereby reduce HIV incidence in the intervention communities.

SEARCH randomised 32 communities of about 10,000 people each to receive either standard-of-care HIV testing and treatment, or an enhanced package. This package includes a multi-disease prevention campaign that tests people for hypertension and diabetes as well as HIV, and immediate HIV treatment for those who test HIV positive, regardless of CD4 count. Testing occurs at a two-week health fair that is held in each of the 16 intervention communities, and is also offered through door-to-door testing for those who do not attend the health fairs.

SEARCH started in summer 2013 but after the World Health Organization recommended PrEP for all populations at risk of HIV in November 2015, it started offering PrEP to people who tested HIV negative but were determined to be at high risk of HIV. Their ‘risk score’ is determined by an algorithm that includes factors such as age, gender, marital status, education, occupation, circumcision in men, alcohol use and whether people are polygamous. However, presenter James Ayieko explained that the algorithm was a dynamic one and was being continually adapted as new factors associated with HIV infection came to light during the course of the study, or turned out not to be associated with it.

“It was developed primarily as a research tool to assess HIV infection risk,” he explained, “rather than to set criteria for PrEP. It’s therefore quite conservative and almost designed to minimise the number of people we offer PrEP to: those we do offer it to are genuinely at very high risk.” It was therefore decided to also allow people to self-refer and discuss whether they needed PrEP even if their risk seemed lower than that suggested by the algorithm.

Ayieko presented results from five of the 16 intervention communities, three in Kenya and two in Uganda. During the period of study 21,212 people (out of a possible maximum population of 50,000) tested negative for HIV. Of these 2991 (14%) were determined to be in need of PrEP by their risk score and were offered PrEP during their post-test counselling. A further 1073 (5%), having heard of PrEP during the community education campaigns, decided they might be at risk and came forward to discuss it even though their risk score was lower than the threshold for PrEP (some people with high risk scores were also self-referrals, but were included within the ‘high risk’ category). Altogether 4064 people, nearly one in five, were assessed as potential PrEP users.

Of these, 739 (18%) actually started taking PrEP within 30 days of being offered it and 571 (14%) started PrEP the day they discussed it. A much higher proportion of people who self-referred started PrEP within a month than people with high risk scores – 39 vs 11%. Although the vast majority of people started PrEP within 30 days of discussing it, a few individuals took up to six months before deciding to start.  

Only 30% of people assessed by the researchers as being at high risk of HIV infection actually considered themselves at high risk, whereas half of those who self-referred did.

The researchers wanted to find out more about the 'early adopters' of PrEP, those who started it within 30 days. Looking firstly at those who were assessed as being at high risk, early adopters were 53% more likely to be male, 57% more likely to have no education other than primary, 66% more likely to perceive themselves to be at risk of HIV, 92% more likely to be polygamous, and 289% (nearly four times) more likely to have an HIV-positive spouse. They were five times more likely to have been assessed at a community health fair (rather than during household testing).

Among those who were not considered at risk but nonetheless started PrEP, early adopters were older (people aged 46-55 were 130% more likely to start than under-25s), 47% more likely to have had no education than primary, 160% more likely to have an HIV-positive spouse, and, in contrast to those assessed as being at high risk, 86% more likely to have been tested at home rather than at a community health fair. They were actually not significantly more likely to perceive themselves as being at current risk than those who decided not to start; this does not mean starters were not at risk, but that there was no correlation between self-perception of current HIV risk and deciding to start PrEP.

Ayieko commented that it was disappointing both that only a minority who were actually at risk perceived themselves to be at risk or started PrEP. In an accompanying poster presentation, factors identified as barriers to PrEP were identified through interviews with 42 people who decided not to start PrEP and 63 other community members. Out of a large number of barriers, one factor that stood out for the researchers was that interviewees would have preferred to be able to take PrEP on demand rather than daily. Women, in particular, spontaneously introduced the idea of preferring periodic injections, which is what they are more used to in the context of contraception. Another barrier mentioned by young people was the assumption that parental consent would be needed for them to start.

Nonetheless, Ayieko said, the study does show that population-based  PrEP delivery and uptake is feasible and acceptable in “real-life” rural African settings.