Three critical evidence gaps from AIDS 2018

By Dr Marija Pantelic

This week at the 22nd International AIDS Conference in Amsterdam has no doubt been exciting. It is always inspiring to share a space with thousands of people passionate about ending AIDS – community representatives, scientists, advocates, clinicians, policy makers. But, yet again, we saw critical knowledge gaps, particularly around programmes for young people and other key populations affected by HIV in low and middle-income countries.

CATS in SwazilandGcebile, Phetsile, Ntsiki and Sipho are Community Adolescent Treatment Supporters (CATS) in Swaziland. © International HIV/AIDS Alliance

I will highlight three evidence reviews that the Alliance has conducted with partners and presented at AIDS 2018, because they provide good oversight of the state of the evidence on various topics:

  1. A global evidence gap map of eHealth interventions for key populations and young people (developed together with the Africa Centre for Evidence, University of Johannesburg with support from ViiV Healthcare)
  2. systematic review of programmes to address gender-based violence against young people living with HIV or affected by HIV in low and middle-income countries (developed with the Centre for Evidence-Based Intervention, University of Oxford, for the Alliance-led READY+ programme).
  3. systematic review of programmes to reduce self-stigma among people living with HIV and key populations affected by HIV in low and middle income countries.

Across these three big evidence reviews we are consistently finding three critical gaps in knowledge and programming that need to be addressed urgently.

1) Research continues to ignore young key populations

Firstly, the three evidence reviews listed above identified promising interventions for adults living with HIV, but none found effective strategies for young key populations (such as lesbian, gay, bisexual and transgender people, sex workers, and injecting drug users) in low and middle-income countries (LMIC).

Indeed, across these three reviews, which together represented tens of thousands of research participants from LMIC, we identified only one study focusing on a young key population. The study focused on men who have sex with men in Thailand but unfortunately, it did not result in significant reductions in self-stigma (our outcome of interest).

2) Few studies involve the population of interest in intervention design and delivery

While studies often report piloting questionnaires with the population of interest, they rarely reported whether they engaged the community in the design and delivery of the programme. Among the few studies that reported community engagement in programme design or delivery, none reported sufficient detail to allow replication. Of course, this does not mean that none of these studies meaningfully involved community and indeed I am aware (anecdotally) of many researchers who invest a lot of effort in this.

However, without transparently reporting about the extent to which they engaged with community, how they did it and what were the benefits and challenges, readers are unable to infer or replicate much. From an ethical perspective, it should also be an imperative to adequately acknowledge people’s contributions to the research.   

3) Too many programmes rely on HIV awareness raising

Finally, a striking number of programmes focus solely on raising awareness around HIV, gender norms and stigma. Yet, our preliminary research findings suggest limited effectiveness of such approaches to address gender-based violence or self-stigma.

In our review on gender-based violence, many studies involved programmes that focused exclusively on education and dialogue with survivors of violence. Understandably, of these, few resulted in reductions in gender-based violence. It is unclear what theoretical framework is used to hypothesize ‘if we talk to girls, others will assault them less’.

Our other systematic review on self-stigma programmes is finding similar results. While many programmes focus primarily on educating people, this doesn’t seem to be sufficient for reducing the rife self-stigma they were experiencing. This is not to say that awareness raising isn’t important. However, it is far from sufficient. Beyond ineffectiveness, I am concerned about the messages we are inadvertently sending through this approach. Are we saying that stigma and gender-based violence are a product of victims being uneducated? I hope not.

Similarly, our global evidence map of eHealth for communities most affected by HIV found that the overwhelming majority of studies relied on client education and behaviour change communication. Two studies examined the use of eHealth for healthcare provider education and electronic decision support so we categorised them as structural interventions. But the remaining 73 studies focused entirely on educating key populations and young people. This approach ignores the roles and responsibilities of other actors in the HIV response.

Addressing data gaps

These data gaps might be disheartening but they can easily be addressed if we direct our efforts and funding wisely. We cannot afford to continue making hasty assumptions around what works.
Leaving no one behind in the HIV response also means that it is unacceptable to ignore young people and communities that are disproportionately affected by HIV. All people – no matter where they are from, no matter their age, sexual orientation, gender identity, drug use or line of work – deserve HIV services, policies and programmes that are guided by the best available evidence.

We appreciate that some researchers may think recruiting young key populations is difficult but community-based organisations can help. Talk to us and we will be happy to refer you to country partners. If you involve them meaningfully, take into account their expertise and form partnerships that are fair, they might be willing to take part in your research!

Now let’s get to work and make sure this becomes reality.

Note: this blog listed a number of knowledge gaps without offering examples of effective interventions. This is because we are still in the process of finalising meta-analyses wherever possible and writing up the final manuscripts. Watch our website for updates on findings or get in touch with

Access the systematic evidence map

Read Marija's blog about evidence-based practice