High HIV Prevalence Found Among Male Partners of Thai Sex Workers

An Interview With Neha Shah, M.D., M.P.H.

There's nothing like hearing the results of studies directly from those who actually conducted the research. In this interview, you'll meet one of these impressive HIV researchers and read her explanation of a study she presented at CROI 2009.

My name is Neha Shah. I am a physician with a public health degree. I work at the [U.S.] Centers of Disease Control in Atlanta, Ga., and with the Global AIDS Program. We did a research project looking at male clients of female sex workers in Thailand.1

Thailand had done a great job in the past doing condom prevention programs, trying to look at sex workers and their rates of sexually transmitted diseases. They had a huge program in the 1990s that has been drastically scaled down over the last several years. We wanted to go back and see what was going on with that population -- and specifically their partners, who are likely also at very high risk [for HIV infection].

Neha Shah, M.D., M.P.H.
Neha Shah, M.D., M.P.H.

Currently in Thailand, there is no data on partners of female sex workers. They use a lot of proxy populations, including army recruits or young men [in general], to try to estimate the HIV prevalence of partners of female sex workers. We wanted to specifically recruit those partners and find out what was going on with that population in terms of their sex behaviors, HIV and STD [sexually transmitted disease] prevalences.

The problem is that they're a very hard population to get a hold of. [As a sex worker,] how do you get your partners to come in and agree to being tested? So we had to come up with a unique way to encourage them to come in, get tested and be part of the study. We used a method called respondent-driven sampling. It's been used to recruit difficult, hard-to-reach, hidden populations for several years now. We thought, "Maybe, if we modify it a little bit, we can get a different population in."

Respondent-driven sampling is where you ask your friends who are in the same line of work as you to come in and join a study. You're asking your friends to come in, as opposed to a researcher or a study project officer going and approaching people they don't know. We had the sex workers recruit their sex-worker friends, and then we also asked them to have their partners -- their clients and their regular partners -- come into the study. It was a bit more of a comfortable environment, as opposed to a researcher or a study person.

Did you have problems getting clients to come in?

We didn't. We were actually pretty successful. We were able to stop the study [enrollment] early because we'd reached the sample size of getting clients to come in.

Was there any incentive given?

There was. All the sex workers and everyone who participated got about $11 in U.S. money to come participate in the study, which covered the basic cost of time and travel. Nothing more than that, [but] that seemed to be encouraging enough for people to come.

Were the clients only local men, not people from elsewhere in the world?

We didn't differentiate. We told [sex workers] to recruit one of their clients: whoever they thought they felt most comfortable asking that sort of question to without ruining their own relationship to that person. We asked about 500 of the 700 sex workers, on a first-come, first-serve basis, to recruit their partners. We were able to recruit about 185 clients -- our goal was about 150. We were able to reach our goal, which was very exiting for us.

We found that there was a very high prevalence of HIV in this particular population. The clients themselves were between 17% and 23% in HIV prevalence. The sex workers also recruited just one of their regular sexual partners, and they also had a fairly high HIV prevalence.

Were they aware of their status?

We had asked that question, and there weren't a lot of people who had responded that they knew their HIV status or had been previously tested. We're concerned that might be a sensitive question, so we're not sure how accurate [the response to] that particular question was, but these are mostly new people now being HIV diagnosed.

Aside from the HIV prevalence, we also looked at how many of these clients and sex workers were serodiscordant, where the sex worker was HIV negative and the client was positive, or vice versa. That was also very concerning for us. We found about 31% of the couples were serodiscordant.

The other thing that's concerning is that, in those serodiscordant partners, not only is the sex worker at risk if she's with a positive partner, but that partner probably also has other partners besides the sex workers. The client serves as a bridge, taking HIV from a high-risk, sex worker population to the general population.

Is there no condom use between the sex worker and the client?

We asked what the rates of condom use were between your sex partners, regular partners, casual -- all sorts of different partners. We were very surprised and very excited that there was about 80%, 90% condom use, which is great -- but not 100%, [which is especially important] when you have a positive and negative person. So the condom use is good, but not high enough when you have negative partners.

Were many of these men married?

About 40% of the clients are currently married. There is a high number of people who are married, which is concerning because they could also be spreading HIV to their wives, spouses or [other] permanent partners.

It's curious that this kind of study was never done before, particularly because Bangkok is famous for its sex work industry. People throughout the world go there. Why was a study like this ever done before?

As I mentioned earlier, clients are a very hard population to sample. It's hard to convince them to come into surveys. It's obviously not an open culture; they're not congregated, so it's hard to identify who clients are. There have been studies in the past that tried to recruit them, but they've been mostly researcher-initiated: I, as a researcher, would go out and find somebody who I think might have been a client of a sex worker. It's hard to say whether that person really was, and it's a certain type of person you would recruit, versus having the sex workers themselves recruit their clients.

Have these kinds of studies been done in the United States and Canada?

They are looking at partner studies. Right now there are several studies going on to try to find the best way to recruit partners for many different subpopulations: sex workers, injection drug users. They're going on around the world.

It's been thought that it's very hard for women to transmit HIV, and that as a result we don't see huge numbers of men who contract HIV after using sex workers. The sex industry is just as big in the United States as in Thailand, and we're not seeing huge HIV infection numbers among heterosexual men who use sex workers. By comparison, the numbers look very large in this study.

Yeah, that's a good point. I think, for this population, we really looked at clients of street-based sex workers, and they're at a very high risk. With this study, because we were able to directly recruit their clients using the sex workers, we found that there was a high prevalence. I think perhaps in some of the other studies, we're not sure whether the clients themselves were recruited, or some other partner. It's hard to know, and every country has a different epidemic.

These partner studies that are now happening are going to be very important to understand what the epidemic is like in these high-risk populations.

What were the behaviors associated with HIV infection?

I think because our population was fairly small, we weren't able to associate too much. We were able to find out that having had your first sexual intercourse at a young age, meaning less than 19, was associated with being HIV positive, as well as using condoms with your regular partners in addition to your clients, or just casual partners. But because the sample size was small, we can't make any definitive conclusions we'd be able to generalize to this entire population at this point. Definitely a population we need to do some more research on, get some bigger numbers, get a wider sample -- a bigger representation of the population -- to understand exactly what's going on here.

You're going to stay focused in Thailand and not the United States?

My group works internationally, but I'm sure that there is work going on in the domestic side as well.

How did you compensate for the undersampling of clients?

Because there's no way to actually analyze this group just yet, we were trying to come up with the best method of analysis. We had to weight it based on the female sex worker: what she considers her network of other female sex workers. She would recruit a sex worker, and that sex worker would recruit a client. Depending on how many people you know, you're more likely to be connected to more clients.

We took that, and then we also asked each sex worker how many clients they have in a given period -- for our study, it was a month. We asked how many clients they had, and since they were only allowed to recruit one, we took the inverse of [their total client number] and then weighted that with the female sex worker's network size.

What did you find was most unexpected about the study's end results?

I think we were surprised to find out how high the HIV prevalence was. They were using proxy populations previously, which had a very low prevalence. This was really surprising, to see how high it was, and really concerning that we don't have enough information on what kinds of programs to do for that population. Given that it's the first study, we really need to do more partner studies to understand the epidemic in this group and how it can affect the general population.

I know that in Thailand, and particularly in Bangkok, they have a very strong program specifically for sex workers. I think I met the head of the sex worker union there at an International AIDS Conference. She's very active, she's doing conference programs, she has displays of many condoms, and she explains to these women the importance of condoms. So there is a lot of outreach there. It's not like other countries in the area, where there's nothing being done.

There is a lot that Thailand has been doing. They are one of the leaders of the condom promotion program, especially with female sex workers. But again, we didn't know what was going on with their partners -- maybe it was the partners spreading it to the sex workers and then going to the public. It's hard to understand. This was why we first started this project, to see what was going on with the partners and how that might explain anything else.

Do the sex workers and the clients who participated in this study know the results? Are they aware of what the percentage is, so they could change their practices?

They were given the opportunity to come back and get their results -- many opportunities to come back, which we really tried to stress. Unfortunately, there wasn't a very high rate of return for people coming for their results. For clients, about 23% came back and got their HIV results. That was one of the other things that we're really trying to emphasize for the research: how to get people to come back for the results, so they're aware -- and then can change their behavior.

You could probably bring the poster to the sex work association and just give them the numbers. Once they see the numbers, it might compel other women to be a little more vigilant about protection.

I certainly hope so. The organizations that we worked with for this project are aware of the results, and they work with these populations, too. So it's getting out there.

Thank you very much.

This transcript has been lightly edited for clarity.


  1. Shah N, Subhachaturas W, Anand A, et al. Recruiting an emerging high-risk hidden population: male clients of female sex workers in Bangkok, Thailand, August to October 2007. In: Program and abstracts of the 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montréal, Canada. Abstract 1031.
    View poster: Download PDF

Download the audio file