Behavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in high-income countries.

backgroundInterventions to change behaviour among sex workers and their clients have been identified as a strategy to reduce HIV transmission. However, there has been no systematic review that has examined and summarized their effects.

objectivesTo identify and evaluate the effects of the studies performed on behavioural interventions to reduce the transmission of HIV infection among sex workers and their clients in high-income countries. SEARCH

methodsElectronic searches were undertaken using MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL) and other databases between January 1980 and July 2010. Experts in the field were contacted to locate any other studies. SELECTION CRITERIA: Randomised controlled trials or specified quasi-experimental designs with comparison groups that examined the effects of behavioural interventions aimed at reducing the risk of HIV or sexually transmitted infections (STIs) transmission among sex workers in high-income countries. We reviewed studies for outcome relevance and methodological rigor. DATA COLLECTION AND ANALYSIS: Two reviewers independently applied the inclusion criteria to potential studies, and any disagreements were resolved by discussion. Studies were assessed for completeness of reporting and extracted data. MAIN

resultsA total of four studies were included, comprising two randomised controlled trials and two quasi-experimental pretest-posttest trials with control groups involving 1795 participants. No trials reported HIV prevalence/incidence as outcomes.Overall, the effects of behavioural interventions for sex workers in high-income countries on STI incidence did not differ significantly among two studies using a random effects model (risk ratio (RR) 0.46, 95% confidence interval (CI) 0.11 to 1.98). Only one study found that the self-reported STI prevalence in clients of female sex workers was statistically significant (RR 0.09, 95%CI 0.01 to 0.72, P=0.02). There was no significant difference after behavioural intervention for condom use. Two studies demonstrated the effectiveness of intervention for knowledge of HIV transmission among sex workers (RR 1.82, 95%CI 1.55 to 2.14) and clients of sex workers (RR 1.93, 95%CI 1.46 to 2.55). AUTHORS'

conclusionsThere is limited evidence from randomised controlled trials for the effectiveness of behavioural interventions to reduce the transmission of HIV infection among sex workers and their clients in high-income countries. Further randomised controlled trials are very likely to have important impacts on our confidence in the estimates of the effects, and are likely to change the estimates for effective interventions with outcomes of HIV incidence or prevalence and a variety of different settings among sex workers and their clients in high-income countries. Randomised controlled trials that test for the identification of effective interventions for HIV prevention with outcomes of biological endpoints, such as HIV incidence or prevalence, are needed for these neglected populations. More research is also needed for male or transgender sex workers and their clients in high-income countries.