Randomised, controlled, community-level HIV-prevention intervention for sexual-risk behaviour among homosexual men in US cities

Summary

Background

Community-level interventions may be helpful in population-focused HIV prevention. If members of populations at risk of HIV infection who are popular with other members can be engaged to advocate the benefits of behaviour change to peers, decreases in risk behaviour may be possible. We assessed a community-level intervention to lower the risk of HIV infection, focusing on men patronising gay bars in eight small US cities.

Methods

We used a randomised community-level field design. Four cities received the intervention and four control cities did not. Participants were men from each city who went to gay bars. Men completed surveys about their sexual behaviour on entering the bars during 3-night periods at baseline and at 1-year follow-up. In the control cities, HIV educational materials were placed in the bars. In the intervention cities, we recruited popular homosexual men in the community and trained them to spread behaviour-change endorsements and recommendations to their peers through conversation.

Findings

Population-level rates of risk behaviour decreased significantly in the intervention cities compared with the control cities at 1-year follow-up, after exclusion of surveys completed by transients and men with exclusive sexual partners. In a city-level analysis, in the intervention cities we found a reduction in the mean frequency of unprotected anal intercourse during the previous 2 months (baseline 1·68 occasions; follow-up 0·59; p=0·04) and an increase in the mean percentage of occasions of anal intercourse protected by condoms (baseline 44·7%; follow-up 66·8%, p=0·02). Increased numbers of condoms taken from dispensers in intervention-city bars corroborated risk-behaviour self-reports.

Interpretation

Popular and well-liked members of a community who systematically endorse and recommend risk-reduction behaviour can influence the sexual-risk practices of others in their social networks. Natural styles of communication, such as conversations, brought about population-level changes in risk behaviour.

Introduction

Efforts to curtail the incidence of HIV infection require that individuals at risk of infection change their sexual behaviour. Studies have shown that intensive small-group interventions combining risk education with activities to improve safer-sex attitudes, behaviour-change intentions, risk-reduction skills, and problem-solving can decrease the rate of high-risk sexual behaviours among homosexual or bisexual men,1, 2 women,3, 4 and adolescents.5, 6 However, intensive counselling of individuals or small-group interventions, although necessary, are not sufficient to curtail a major public-health epidemic. Community-level prevention efforts that focus on HIV-vulnerable populations are also essential and urgently needed.

Few randomised, controlled trials have studied community-level interventions aimed at decreasing the rate of risk behaviour in HIV-vulnerable populations. Among populations in developing countries where even basic knowledge about HIV is commonly lacking, outreach education has produced positive effects in some single-city campaigns.7, 8 However, approaches are needed that extend beyond risk education. Several small-scale trials of community-level interventions to prevent HIV infection have shown promising results. Previous studies9, 10 tested a community intervention in which popular members of gay communities in three small cities were asked to endorse, through conversations with their friends, the social desirability of lessening high-risk behaviour. Cross-sectional surveys revealed downward trends in the sexual-risk behaviour of homosexual men in each of the three cities after intervention. Kegeles and colleagues¹¹ used a similar single-city intervention, supplemented by HIV-prevention social events in gay communities, and found a change in the amount of risk behaviour. HIV-prevention community projects have shown that continuous outreach programmes focused on changing the norms of sexual behaviour can influence the intentions, readiness to change, and behaviour of some populations at risk of HIV.¹²

We investigated the effectiveness of a community-level intervention to prevent HIV infection among homosexual men in 16 small-sized to moderate-sized cities in the USA. Eight cities—four intervention and four control cities—constituted the first wave of the trial and we report our findings for these cities here. The other eight cities—four intervention and four control cities—were entered in the trial 6 months after the start of the first wave. The baseline and 1-year follow-up surveys in the first-wave cities were carried out during winter months, and in the second-wave cities were carried out during the summer months. Baseline risk-behaviour results in the second-wave cities were much different than in the first-wave cities and the proportions of transient respondents were higher. In addition, a bar in one of the second-wave intervention cities closed during the trial, and in a second-wave control city, the intervention was implemented by a local organisation. Therefore, the expansion of the trial in the second wave was unsuccessful and led us to report only the outcomes of the first-wave cities.

By contrast with previous studies that have examined single-city interventions, we used a randomised field-trial design in which four city populations received the intervention and four control cities did not. In each city, we targeted men without exclusive partners, who visited gay bars. Although incidence of HIV infection among older, white, homosexual men has levelled in many original, large, urban HIV epicentres, HIV risk-behaviour rates13, 14 and new HIV infections,15, 16 remain high among homosexual men in many areas traditionally deemed “second tier” areas for HIV. Men who visit gay bars tend to be young and have higher rates of HIV infection than their older counterparts.17, 18 These men have also been shown to have higher rates of high-risk sexual behaviour than men who do not visit gay bars,¹⁹ and, therefore, are at particularly high risk of HIV infection.

The rate of HIV risk behaviour may be high in certain communities because behaviour change through risk reduction is not an accepted social norm. Several studies have shown that perceptions of whether safer-sex practices are viewed as norms among one's friends, sexual partners, and other peer-group members, are important factors in an individual's own adoption of behaviour change.13, 20, 21 These findings suggest that interventions that redefine social norms of sexual behaviour towards avoidance of high-risk practices may have substantial impact on the rate of high-risk behaviour. Social norms are commonly initially formed by observing the behaviour or standards of visible, popular, and well-liked members of one's own peer network.²² Because the standards communicated by such opinion leaders serve as modelling influences that can be spread to others,23, 24 interventions that engage popular opinion leaders to endorse behaviour change in communities may be especially powerful for altering risk-behaviour practices. We investigated whether an intervention in which popular opinion leaders endorsed risk avoidance could reduce the rate of sexual-risk behaviour among homosexual men visiting gay bars.