What does “first ever” research look like?

Starting from nothing

When we talk about where there is no data, it is not strictly true that there is NO data.  Even if information has not been collected or analysed or published, it still exists.  But collecting, analysing and publishing information is often an important step to getting health problems to be better understood and to dealing with them effectively.

I was once asked to do some research on key populations for HIV – particularly sex workers and men who have sex with men – in a small country where no structured research had ever been carried out on these groups.  Of course, if you asked someone in the ministry of health, or even someone on the street, whether sex workers and men who have sex with men existed in the country, they would have an opinion – or indeed several opinions.  But if you want to start building programmes to work with these populations on issues such as human rights and HIV you need more than that.  Our initial brief was to carry out a population size estimate – to try to figure out roughly how many sex workers and MSM there were – but given the complete lack of any documented information, it was clear that this was not going to be possible.  So instead, we decided to try and find a few people to talk to and to write up a few case studies – just to document for the first time that there were indeed sex workers and MSM, to get a sense from them of the human rights and HIV situation, and possibly to lay the groundwork for some more formal research.

By word of mouth, and by identifying people through internet chat sites, we managed to find a handful of MSM who said they would be happy to talk.  Given the sensitivity of the topic and the fact that sex between men was illegal and highly stigmatised in the country, we arranged to meet in private locations like out-of-hours health facilities or hotel rooms, or even in one case in the back of a taxi.  Similar arrangements were made to speak with sex workers, although finding the way in to the group was less straightforward, as it involved working through local intermediaries.

We had an idea of how many people we wanted to speak to in the time available, and had developed both a questionnaire and a focus group discussion guide in order to be able to adapt to different situations.  We also prepared an informed consent statement and form, designed to make sure participants understood the purpose of the exercise and were happy to participate.

In reality, it is very hard to control what is going to happen in this sort of exercise, and nothing worked out quite the way we had planned.  Here are a few of the stumbling blocks we came up against:

  • Around half of the appointments were not kept.
  • Although participants were asked to come individually, a number came in pairs or groups of three, making it difficult to use either the interview or the focus group discussion guides.
  • A number of participants avoided responding to questions about sexual behaviour and sex partners – in some cases they declined to talk about themselves but talked about people they knew instead.
  • One interview had to be cut short because we were being devoured by mosquitoes.

Limited results; managing expectations

There are many good reasons why the process was so fraught. It is quite likely that participants were not reassured by the explanation of the research or the consent agreement, or that they did not feel able to talk openly despite the initial agreement.  Although the work did produce a few case studies, the amount of data and insight produced was fairly minimal.  And although it was not the only work we did during the trip, it ended up being relatively expensive for what it produced.

It may seem, given these challenges, that there is very little point in conducting this sort of research.  However, in situations where virtually nothing is known, and where the government or other funders use this fact to sidestep sensitive issues, it is important to start somewhere.  We were able to take the scant information to decision makers, to demonstrate that these groups existed and that even though it was challenging, it was possible to work with them to find out more about their lives.  We also made a handful of contacts who helped out, much later, in developing a more structured piece of research.  Conversely, as the difficulties we faced suggest, it would have been impossible to conduct a more structured piece straight off: we would not have been able to identify enough people and implement the research in a rigorous or acceptable way.

One of the biggest challenges in this process was, of course, that the information we were able to obtain fell far short of the client’s expectations.  In this case we realised before starting that this would be the case and was able to discuss this with the client and come to an agreement on what was realistic.  However, it is often not until research projects get going that challenges like these are identified.

What is the lesson?

The lesson here is that when trying to gather data on sensitive topics, it is often necessary to begin with very small and informal studies.  Being clear about the methods, and ensuring participants understand the process and consent to participate in it remains essential.  But don’t set your expectations in terms of results too high.  And in an environment where those who fund research are often removed from realities by some distance, it is essential to be clear from the start about what is and isn’t possible.  If a client wants a large behavioural study or a population size estimate, but the starting point is similar to the one described above, it is important to develop some type of “roadmap” that plots out the different processes and sub-studies needed to get there.  We’ll post more about this “roadmap” idea on Where there is no data over time.