Public health officials, donors, epidemiologists, researchers and clinicians need to factor in the relevance of the social and the cultural if they are to see biomedical HIV prevention strategies – including treatment as prevention (TasP) and pre-exposure prophylaxis (PrEP) – succeed in real-world settings, speakers told the 2nd International Conference for the Social Sciences and Humanities in HIV, held recently in Paris.
“TasP is a simple acronym that masks considerable complexity,” commented Professor Gary Dowsett of La Trobe University in Melbourne, Australia.
Although modelling studies have ‘demonstrated’ that treatment as prevention (a programme of HIV testing and treatment) could eliminate HIV infections in societies such as South Africa and Vietnam, they have relied on utopian assumptions, such as 100% of people getting tested and 100% of diagnosed people taking antiretroviral treatment.
But the gaps in the ‘treatment cascade’ suggest that the world is much more complicated than this.
Unpacking the treatment cascade
For treatment as prevention to work, large numbers of people who are at risk of HIV infection need to understand that they are personally at risk; find the idea of testing for HIV acceptable; and be able to go to trusted, convenient and affordable HIV testing services. They also need to be willing to test repeatedly, perhaps once every year (and continue to have access to the necessary health facilities).
Those who are diagnosed with HIV need to come to terms with the result and believe that there are benefits to engaging with health services now, even though they may not feel ill. Once again, trusted, convenient and affordable health services need to be available.
Moreover, those diagnosed need to stay in touch with their doctors and nurses, and attend appointments regularly. Again and again, they need to overcome the same barriers which may have made it hard for them to attend a clinic in the first place – the stigma of HIV, other urgent personal priorities, difficulties getting to a health service, and so on.
When a doctor suggests that they start taking HIV treatment – which may be very soon after diagnosis – individuals need to believe that it will benefit them personally. Alternatively, they need to understand the possible advantage for a sexual partner, or for other people, and consider this a good enough reason to take the drugs. Any perceived problems with HIV treatment, such as side-effects or other people finding out about the person’s HIV status, need to feel manageable in comparison with the advantages.
An efficient healthcare system is needed to provide the drugs, in a way that is affordable for the patient, without interruptions in drug supply. In case a particular drug doesn’t work for somebody, alternative anti-HIV drugs need to be available.
And all these things need to happen, not just for a short period of time when the conditions are favourable, but for decades.
All interventions are behavioural
Professor Susan Kippax, conference co-chair, therefore argued that we cannot expect to see similar results to HPTN 052 (the study which found that HIV treatment reduced the risk of passing on HIV to a regular partner by 96%) when early HIV treatment is used a population strategy.
For Kippax, all effective interventions require a social transformation that a randomised controlled trial (RCT) cannot tell us how to achieve. “Some of the biomedics seem to fail to understand that people live in cultural and social worlds,” she said. “Populations are likely to differ from one another and may respond very differently from one another.”
In Gary Dowsett’s words: “It is neither novel nor simplistic to point out that no technology exists without or outside human behaviour, and that all human behaviour is socially determined and culturally comprehended.”
The great diversity of results seen in trials of PrEP (from 73% reduction in risk in one trial to having no effect in two others) suggest the importance of attending to the social and cultural. But delegates said that this concern should not be reduced to a question of ‘how to get people to take pills’ – more is needed to achieve effectiveness. Dr Marsha Rosengarten, another conference co-chair, said that the conference’s key message was that “biomedical technologies do not – in and of themselves – prevent transmission of the virus”.
Numerous speakers argued that the difference between efficacy and effectiveness is key.
‘Efficacy’ refers to the impact of a drug or an intervention during a research study, when extra financial and human resources are usually available, among individuals selected to take part in an RCT.
‘Effectiveness’ describes impact in real-world settings, in entire populations, and under resource constraints. It tells us about the results achieved when an intervention is implemented in a 'normal' healthcare system.
“TasP is a simple acronym that masks considerable complexity” Gary Dowsett
So, even though RCTs of early HIV treatment and PrEP have shown that those interventions have efficacy in certain conditions, this does not mean that they will be effective in each setting in which they are implemented.
Effectiveness is not just about the drug, but also the system that delivers the drug, and the response of the people who are supposed to take it.
Although ‘treatment as prevention’ requires a complex set of conditions for it to be effective, it is improbable that such a mix of behaviours, cultural values, health services and political conditions will be alike in different settings.
How these factors come together in any one setting is subject to chance and will inevitably change over time. Moreover, there are obvious differences between populations that have greater or fewer financial resources, civil rights or access to healthcare. Circumstances will be particular and local.
Real-world effectiveness data
Susan Kippax noted that, although observational studies in British Columbia (Canada), San Francisco (United States) and KwaZulu Natal (South Africa) have shown a correlation between expanded HIV treatment and reduced HIV diagnoses, these findings have not been replicated everywhere.
"Effective interventions require social transformation" Susan Kippax
In particular, there is a lack of similar evidence from gay communities in Australia, the United Kingdom or the Netherlands. In places where coverage of antiretroviral therapy is already very good, it may be impossible to bring new infections down without substantial changes in testing behaviour, sexual behaviour, or both.
Moreover, questions were raised about the generalisability of the British Columbia data. The reduced diagnoses are only in people who inject drugs, not in diagnoses associated with heterosexual or homosexual transmission. The Canadian consultant San Patten suggested that the province had had success by targeting the “low hanging fruit” of injecting drug users in Vancouver’s Downtown Eastside.
Interventions have achieved a considerable improvement in the uptake of HIV testing and treatment in this geographically limited area marked by drug use, high HIV prevalence and other social problems. But at the same time, harm-reduction services have expanded dramatically in the city (syringe borrowing fell from 39.6% in 1996 to 1.7% in 2011), raising questions about what has caused drug-related infections to fall.
British Columbia data were used to produce one of the mathematical modelling studies which predicted excellent outcomes, should access to testing and treatment be expanded exponentially. Dr Annick Borquez of Imperial College London suggested that social scientists could help improve models with better inputs on medication adherence, the acceptability of programmes and the reachability of different populations.
But models will never be able to predict social responses such as stigma, or guide us on ethics, she said.
Consequences and repercussions
"Mathematical models could be improved with better inputs on adherence, programme acceptability and population reachability."
As well as understanding the circumstances in which biomedical interventions might be effective, speakers at the conference were also interested in what the long-term consequences of their implementation could be. These social, behavioural and economic repercussions will be different in different places.
As treatment’s prevention benefit becomes more widely understood, how will people’s understanding of the impact of having HIV, and of the drugs used to treat it, change? And how will that further evolve as HIV-negative people are increasingly using the same pills to prevent infection? These developments are likely to be subtle and may take time to occur.
The idea that some people’s sexual behaviour may change in response to PrEP or TasP is already widely discussed, but usually reduced to a single thing called ‘risk compensation’. Dean Murphy of the University of New South Wales was more interested in how people will respond to the new possibilities that are opened up by PrEP. An earlier example would be how, in response to the availability of HIV testing, gay men created ‘negotiated safety’ and ‘strategic positioning’.
In relation to stigma, numerous possible futures were suggested. Will people who do not take medication become a new focus for HIV stigma? Would wider public understanding of reduced infectiousness limit stigma, especially of HIV-positive people’s sexuality? Could universal testing lead to HIV-status disclosure becoming the norm, making it harder for people with HIV to keep that information private? Will early treatment push people to disclose their status before they are ready?
Some delegates had considerable anxieties about what the consequences of HIV prevention being medicalised would be, but a criticism that could be levelled at the conference itself is that it included relatively few studies in which social scientists actually explored these issues with empirical data. One useful exception was a presentation from a randomised controlled trial in Ivory Coast.
"The idea that sexual behaviour may change is widely discussed, but usually reduced to ‘risk compensation’."
The Temprano trial compares results for individuals starting antiretroviral treatment immediately and those starting when their CD4 cell count has dropped below 350 cells/mm3 (other study details have been previously described on aidsmap.com). The researchers wished to examine the social impact of beginning therapy early. Most of the participants are female.
The results after one year’s follow up are reassuring. No statistically significant differences in disclosure of HIV status to sexual partners or to other people, relationship break-up, unprotected sex, sex with multiple partners, fears of forced disclosure or experiences of stigma were recorded at month 12.
Data will be collected for a further year; the researchers called for studies of the longer-term impact.
Dowsett G HIV Prevention, Hyperbole, Culture and Practice. 2nd International Conference for the Social Science and Humanities in HIV, Paris, session CS01, 2013. (View the abstract on the conference website; an earlier version of the talk was published in HIV Australia.)
Kippax S Does TasP Spell the End of the Epidemic? 2nd International Conference for the Social Science and Humanities in HIV, Paris, session CS30, 2013. (Download presentation slides on the conference website; the author has previously published a fuller version of these arguments in the Journal of the International AIDS Society.)
Patten S Canadian Caveats to Treatment as Prevention. 2nd International Conference for the Social Science and Humanities in HIV, Paris, session CS30, 2013. (Download presentation slides on the conference website.)
Borquez A Situating Social Science Enquiry in the HIV Treatment as Prevention Era. 2nd International Conference for the Social Science and Humanities in HIV, Paris, session CS30, 2013. (Download presentation slides on the conference website.)
Desgrees du Lou A et al. Social Impact Of Very Early ART Treatment Within The TEMPRANO Trial, Abidjan. 2nd International Conference for the Social Science and Humanities in HIV, Paris, session CS34, 2013. (View the abstract and download presentation slides on the conference website.)
Jean K et al. Early antiretroviral therapy, sexual behaviours and HIV-1 transmission risk: estimates from the Temprano-ANRS 12136 Randomized Controlled Trial, Abidjan, Côte d´Ivoire. 7th IAS Conference on HIV Pathogenesis, Treatment and Prevention, Kuala Lumpur, abstract MOAC0201, 2013. (View the abstract, Powerpoint slides and webcast on the IAS conference website.)