In The X Files, FBI agent Fox Mulder had a poster on his office wall with a slogan that became a blockbuster 90s catchphrase, and a mantra for a generation of desperate people swayed by pseudoscientific bollocks like the law of attraction:
I Want To Believe.
Chris Carter’s hit show operated within the realm of science fiction, though, and you would hope that in the world of HIV prevention, we’d be relying more on science fact. Unfortunately, the global zeitgeist around testing-as-prevention that’s now infected Australian AIDS organisations has been taken on board while listening to a looped tape of Dusty Springfield’s “Wishin’ and Hopin’”.
After two blogs on ACON’s latest campaign Ending HIV (If condoms are punishment and You can sell anything), I’ve been taken to task on my scepticism around this, not least of which by ACON’s Director of Sexual Health and HIV Prevention Geoff Honnor, so I feel this is important enough to tackle comprehensively: in part 2 (next week) I’ll look at the question of whether investing in this approach is realistic compared to comprehensive condom-based social marketing as a primary choice, and in part 3 I’ll look at the implications of HIV prevention being taken out of the hands of communities (us) and into the hands of the medical industry.
This week, I’ll be asking – what evidence do we have to show that testing-as-prevention even works in the first place?
Research has shown testing as prevention to be ineffective
On February 1, The Lancet published a nationwide population study that looked at whether testing-as-prevention had been effective among gay men in England and Wales. It used data spanning most of the previous decade, from 2001 to 2010.
Its findings should make any government seeking to fund such an initiative nervous: despite a four-fold increase in HIV testing during this period, there was no effect on HIV incidence or the number of undiagnosed HIV infections. This is also despite an increase in men going onto treatment – from 69% in 2001 to 80% in 2010.
Also published in February was a paper funded by the UK government’s National Institute for Health Research. It too observed that despite an increase in treatment uptake, numbers of new infections diagnosed in the UK each year among gay men have not decreased.
The reasons for this, the researchers say, is unclear although they note a “modest” increase in sex without condoms (26%). Through computer modelling, taking into account risk behaviour, testing rates, and the percentage of men on treatment, the researchers explored a number of potential scenarios.
The most alarming of these found that, had all condom use ceased in 2000, HIV incidence would be a whopping 424% higher in the UK and Wales than it is now, a staggering disaster compared to the estimated reduction in incidence when compared to ideal testing-as-prevention scenarios.
“Increased condom use should be promoted to avoid the erosion of the benefits of ART [drug treatment] and to prevent other serious sexually transmitted infections,” the researchers concluded.
Their conclusion about condom use being promoted to avoid “erosion” is a crucial point: prioritising one prevention strategy over another does have an effect, and if that strategy is less effective, the results will be bad.
Promotion of condom use will not be effective if it’s buried within the messages of a wider campaign, or tacked on summarily in a ritual genuflection. Indeed, in a 2008 paper published in the Lancet and funded by the Australian Research Council, a group of Sydney researchers cautioned precisely this: that if people came to widely believe that treatment stopped infectiousness, and condom use declined as a result, “then there is the potential for substantial increases in HIV incidence.”
The key proponent of testing as prevention advises caution
The modern prophet of testing-as-prevention is Dr Reuben Granich of the WHO. When he first introduced the idea that the HIV epidemic could be controlled by testing-as-prevention in a 2009 Lancet paper, many of his peers felt it was more utopian than realistic, and in his reply to their scepticism he conceded that his paper was “a hypothetical exercise and further research is required to assess whether
the studied approach has merit.”
Needless to say, the 2013 Lancet study of the UK/Wales situation mentioned above has given him even further pause for thought. In an editorial published in the same issue, he concludes that “increased testing and earlier treatment is no quick and easy solution.”
With their Ending HIV campaign, ACON expect it to reduce incidence by 80% in less than seven years, despite even the most optimistic studies (such as this one from San Francisco cited by NAM) noting there’s no proof that increased treatment causes HIV rates to drop.
In 2009, Granich presented a plenary talk on testing-as-prevention at an international AIDS conference in South Africa. His theoretical model, based on data from sub-Saharan countries in Africa, estimated that HIV could be reduced 95% over ten years if everyone got tested at least once a year and went immediately onto treatment if diagnosed positive.
But, he added, “it’s a model that is specifically designed to look at questions among a heterosexual epidemic, and hasn’t been applied elsewhere. I would emphasize again that it’s a theoretical model, and doesn’t represent a change in WHO policy.”
Why should it make a difference that the model is based on a heterosexual population?
Studies of heterosexual HIV risk cannot be applied to gay men
Because our risk profile as gay men is completely different. The rise of the internet and hookup apps give us a much greater ability to get sex on tap. Sexual behaviour research in all Western countries shows that, on average, we have sex with more people, more often than heterosexuals. This is not a moral judgment, it is merely a fact.
Add on top of this that anal sex, particularly if you’re the bottom, is around twenty times riskier than vaginal sex in terms of HIV transmission (why do you think emergency medicines in hospital are delivered up the arse instead of orally?).
Finally, the third layer in our rapidly-collapsing cake is the 1-3 month window period during which HIV is undetectable through testing.
Put all these three things together and you’re left with the uncomfortable truth that most gay men who become infected will infect at least one other person (probably unknowingly) before they reach the golden target of diagnosed + on treatment + viral load low enough to reduce transmission risk.
But according to Geoff Honnor, ACON “accept the research findings that underpin the treatment as prevention approach”.
They must really want to believe.