We have recently seen a spurt of advances in biomedical HIV prevention research. These interventions include circumcision, vaccines, microbicides, pre-exposure prophylaxis and suppressive therapy for herpes simplex virus-2 (HSV-2) to reduce risks of HIV transmission through sexual intercourse. While these new technologies provide a variety of prevention options, the clinical trial data has been less than promising (with the exception of medical male circumcision – as demonstrated in the three African randomised controlled trials.) This pattern of flat trials was poignantly illustrated in the recent findings of the FACT 001 microbicide trial in South Africa.

A central and obvious issue at the heart of biomedical prevention is behaviour change, more specifically adherence to medical products of investigation on trial. Recent studies suggest that product adherence represents an ’Achilles heel’ for these prevention trials. Given the challenges with adherence within the prevention realm, we can glean some insights from more established research on adherence to anti-retroviral treatment (ART).

It needs to be stated at the outset that the term ‘adherence’ is a complex dynamic behaviour influenced by characteristics of the individual, the clinical setting, the relationship between the medical provider and individual, the treatment regime, the medical condition, and the surrounding context in which the individual lives, and the health care system. These relationships can be reciprocal and reinforcing and, as a result, an individual’s level of adherence can change over time. Adherence is not solely a social-psychological issue (within the individual and person-to-person relationships), but is also conceptualised as a structural issue, which includes level of access and availability of products.

Where Do We Go From Here?

Some insights from ART adherence research that can be used in HIV prevention research relate to how we measure adherence, understanding social/ structural factors impeding adherence and leanings from recent adherence interventions that show promise.

  • Firstly, measurement has always been a thorny issue. Many clinical trials to date have measured adherence through self-reports of adherence behaviour and/or counts of pills/products, but these measures may overestimate adherence. Our experience suggests that reliance on a single approach is not optimal. We need approaches that work to reduce social desirability through a method of communication that is more sensitive to adherence difficulties, understands social and normative processes on adherence to medication and processes that ensure confidentiality in self-reports. Standard and simple methods of self-reporting (including the use of techniques such as visual analogue scales for low literacy populations) and computer assisted methods enhance validity and address issues of privacy and confidentiality. In our clinical research, we also need process-oriented research to document ecological validity of these approaches.
  • Secondly and equally important, we need formative work that attempts to understand the social context in which we operate as programmers/interventionists/researchers. A first step is to develop appropriate conceptual frameworks. To date, many theories applied to ARV adherence are targeted at individuals to improve knowledge and skills related to HIV and AIDS and ARV treatment. This approach has been shown to be limited. Revised frameworks need to recognise both social support and the social context surrounding individuals can facilitate and maintain a specific health behaviour change.
  • We need more complex theories that address multiple domains associated with behaviour change and include, self-regulation capabilities, responses to internal affective states such as depression or positive affect that have an impact on these capabilities and the environmental context. Regarding the latter, structural factors related to the availability of and ease of access to medications, including pharmacy stock-outs and costs associated with transportation to clinics, and waiting times are important determinants of ARV treatment adherence in resource-limited settings.
  • Finally, some recent and promising findings based on ART adherence research include motivation-based techniques (cognitive-behavioural counseling), conditional incentivisation in exchange for adherence behaviours, SMS reminders, telephone-based unannounced pill counts, electronic drug monitoring, and outreach methods like home visits that provide adherence support and counseling through in-home rather than requiring participants to travel to study sites.
    HEARD’s three country study on mobile populations in southern Africa is currently using SMS messaging to instill sexually protective behaviours in truck drivers, informal traders and sex workers at border posts. This study will provide important insights on SMS methods for HIV prevention in difficult to reach populations.

Each of the methods noted above have implementation challenges. Pilot work on the feasibility of these approaches to particular contexts is recommended, including triangulated approaches which enable researchers to understand the complexities of adherence through access to different types of data. HIV prevention research requires a deeper engagement with the issue of adherence and social context. HEARD seeks to address adherence and broader issues of HIV prevention in the upcoming book initiative, leading up to the International AIDS Conference next year, on Innovations in HIV Prevention.

By Dr Kaymarlin Govender