The most recent of a series of national gay men's surveys in Australia shows that over one in seven (14.7%) of all the non-HIV-positive respondents in the 2018 survey is taking PrEP. A smaller survey conducted in 2017, which focused on gay men's opinions of biomedical prevention alone, found that one in five (20.5%) were taking PrEP. It also found that about one in seven (15.1%) reported that they were taking PrEP and experiencing reduced anxiety and more confidence about sex.
This high proportion of gay men on PrEP has enabled Australian researchers to construct two sample 'PrEP cascades', which could be used in other countries and locations to gauge the uptake of PrEP among their key populations at risk of HIV — and some of its impact.
One of the most powerful advocacy tools for measuring progress towards full HIV treatment coverage has been the 'treatment cascade', exemplified by UNAIDS’ 90-90-90 campaign. This measures progress against a standard in which 90% of people with HIV have been diagnosed, 90% of them are on sustained antiretroviral therapy, and 90% of those have undetectable viral loads. Achieving, as a result, 71.9% of a country’s entire HIV-positive population should result not only in far fewer cases of AIDS and deaths, but also, as the more person-centred U=U campaign reminds us, mean that far fewer people are infectious.
Globally, however, we also now have biomedical prevention in the shape of PrEP. It is a prevention measure for HIV-negative people which, unlike, say, the distribution and social marketing of condoms, depends mainly on provision via formal healthcare systems. Given that the scale-up of PrEP could make its own substantial and cost-effective contribution to ending the global HIV epidemic, as this recent Dutch modelling study shows, it is important to measure this contribution and devise a ‘PrEP cascade’ to measure its distribution and effectiveness.
Professor Martin Holt of the University of New South Wales and colleagues have now devised two PrEP cascades based on the data provided by two different surveys that could be applied to other countries.
The first cascade has only three steps that measure eligibility, awareness and use. The figures in it have been derived from a multi-year survey that has enabled the researchers to chart how PrEP use has increased over time.
The second cascade is from a smaller study and only, so far, makes use of one year’s worth of figures. However, it contains seven steps. These add willingness to use PrEP, discussing PrEP with a doctor, regular HIV and STI testing and whether PrEP has resulted in decreased anxiety about HIV and greater sexual pleasure. Holt says this step was added because 90-90-90 has been criticised for not including quality of life and mental health in its metric, and indeed this is sometimes added as a fourth 90.
The first survey is the Gay Community Periodic Survey (GCPS), which has been conducted in Adelaide, Canberra, Melbourne, Perth, Queensland (Brisbane, Cairns and the Gold Coast) and Sydney since 1996. Tasmania was added in 2014, which was also when it started to collect figures on PrEP. Participants are recruited at gay venues and festivals and increasingly online. It recruits annually in Sydney, Melbourne and Queensland and biennially elsewhere. The latest figures are from 2018.
The second survey is from the PrEPARE Project, a biennial online-only survey specifically assessing gay men’s attitudes to biomedical prevention. It has been conducted in 2011, 2013, 2015 and 2017, but only the figures from 2017 were used in the cascade.
The first step in both cascades is eligibility for PrEP, although this is defined slightly differently in the two surveys. Eligibility criteria were:
- methamphetamine (crystal meth) use in the last six months,
- any STI diagnosis (other than HIV) in the last year,
- condomless anal sex in the last six months with a non-virally suppressed HIV-positive regular partner, and
- condomless anal sex with a casual partner in the last six months.
However, for the last criterion the GCPS restricted this to receptive condomless anal sex, while the PrEPARE Project included receptive or insertive sex. These criteria are also slightly different from the national PrEP guidelines, which specify condomless sex in the last three, not six, months and only syphilis, rectal gonorrhoea or rectal chlamydia in the last three months as STIs.
Briefly, in terms of participant characteristics, there were 39,670 cumulative non-HIV-positive participants in the GCPS from 2014 (many will have participated in several different years), with 8638 in 2017 and 7878 in 2018. Across the years, their average age was 35, with 70% born in Australia, 51% university educated, and 64% in full-time employment.
In 2017 in the PrEPARE Project there were 1038 non-HIV-positive participants. Their average age was 36, more were born in Australia (80%) than in GCPS, slightly fewer had a university degree (45%) and the same proportion as in GCPS were employed. They reported more condomless sex, sexual partners and HIV testing than the men in the other survey.
The Gay Community Periodic Survey's three-step cascade for 2018 was as follows:
- Eligible for PrEP: 37%
- Aware of PrEP: 32%
- Using PrEP: 15%
Eligibility increased from 28% of participants in 2014 to 36% in 2017 and 37% in 2018, reflecting more gay men fitting into the condomless sex and the STI criteria.
Awareness increased rapidly from 30% of those eligible (8% of all respondents) in 2014 to 80% of those eligible in 2017 and 87% in 2018 (32% of all respondents).
The proportion actually using PrEP in 2014 was very low – only 3.7% of those eligible, or 0.3% of all respondents. This rose rapidly to 37% of those eligible (11% of all respondents) in 2017, and 45% of those eligible (15% of all respondents) in 2018.
The PrEPARE Project's seven-step cascade for 2017 was as follows:
- Eligible for PrEP: 54%
- Aware of PrEP: 53%
- Willing to use PrEP: 36%
- Discussed PrEP with a doctor: 26%
- Using PrEP: 21%
- Tested for HIV and STIs in the last three months: 20%
- Reduced HIV concern and increased pleasure because of PrEP: 15%
Comparing the two surveys' figures for 2017 is interesting. In the PrEPARE Project, 54% were eligible versus 36% in GCPS (maybe reflecting the slightly wider criteria in PrEPARE and the participants' different risk profiles); nearly everyone who was eligible was aware; and more were using PrEP.
Cascades can identify 'break points' where the proportion of people staying in the cascade declines most abruptly, and the PrEPARE Project isolates that at a point not included in the GCPS – willingness to use PrEP. Although 97% of those eligible had heard of PrEP, only 68% of those were willing to use it – a relative decline of one-third from those eligible, and almost as great a decline as from those who were aware. There was also a further 14% decline from willingness to having a discussion with a doctor, meaning that less than half of eligible respondents had discussed PrEP with their doctor. Seventy-eight per cent of those who had discussed it with their doctor actually started PrEP – or 35% of those eligible.
Virtually everyone who started PrEP (97%) tested regularly for HIV and STIs. Of those, 76% reported reduced HIV concern and more pleasure. That’s 29% of those eligible – or 15% of all respondents, which coincidentally is the same figure as the ‘end point’ (using PrEP) of the 2018 GCPS survey.
The surveys also showed that participant characteristics had some influence on the cascades. In the GCPS, PrEP users were less likely to come from Canberra or Perth and more likely to be of Anglo-Australian ethnicity, have a degree and be employed. They were unsurprisingly more likely to have behavioural characteristics that made them eligible, apart from using crystal meth – which, interestingly and possibly worryingly, was higher among non-PrEP users. In the PrEPARE project, men over 40 were less willing to use PrEP, whereas men born in Australia and those who knew someone who already used it were more willing.
What the GCPS figures show that awareness of PrEP does not translate straightforwardly into taking it, and while their cascade is easier to compile, it misses out two crucial steps – one cognitive (being willing to take PrEP) and one action-based (going to a doctor to discuss it). The authors comment that adding willingness to take PrEP to the Gay Community Periodic Surveys would be useful.
They comment that, in future, it would be useful to standardise PrEP cascades so that studies use the same sequence of steps. The GCPS three-step cascade is easier to calculate as its steps – eligibility, awareness, and use – are already enumerated in a fair number of studies. However, as the PrEPARE Project shows, cascades with more steps may be needed to identify the 'break points' which are the reason why high eligibility may end up with low usage.
In this case, the significant drop-off was in the proportion of men willing to take PrEP. This indicates a need for research into reasons for this – whether due to perception of low risk, fear of side effects, social stigma or some other reason – so that interventions to address these reasons can be devised.
The authors recognise that developing these cascades was possible because of both the large numbers of PrEP users in Australia and because of the regular behavioural surveillance of gay men there. However, they do comment that a shorter cascade like the GCPS one could be constructed in other locations.
Holt M et al. HIV Preexposure Prophylaxis Cascades to Assess Implementation in Australia: Results From Repeated, National Behavioral Surveillance of Gay and Bisexual Men, 2014–2018. Journal of Acquired Immune Deficiency Syndromes (JAIDS), January 2020.