Commercial sex and the spread of HIV in four cities in sub-Saharan Africa

Objective: To examine whether commercial sex transactions were more common and/or transmission between sex workers and clients more efficient in two A frican cities with high HIV prevalence (Kisumu, Kenya and Ndola, Zambia) co mpared with two with relatively low HIV prevalence (Cotonou, Benin and Yaou nde, Cameroon). Methods: Data on sexual behaviour, HIV and sexually transmitted infections were collected from representative samples of around 300 female sex workers in each city. Sexual behaviour data from a population-based study of aroun d 1000 men aged 15-49 in each city were used to estimate the extent of cont act with sex workers. Results: The number of sex workers per 1000 males was highest in Kisumu and Ndola, but other estimates of the extent or characteristics of sex work co ntact showed no consistent differences between high or low prevalence citie s. HIV prevalence among sex workers was 75% in Kisumu, 69% in Ndola, 55% in Cotonou and 34% in Yaounde. The prevalence of genital ulceration and trich omoniasis was higher among sex workers in Kisumu and Ndola but no clear pat tern was seen for the other sexually transmitted infections. Around 70% of sex workers in Cotonou reported use of a condom with the last client, marke dly higher than in the other cities. Conclusions: Although sex work is likely to have played an important role i n the spread of HIV in all four cities, differences in present patterns of sex work do not appear to explain the differential spread of HIV. However, high levels of condom use among sex workers may have slowed the spread from sex workers to the general population in Cotonou, highlighting the importa nce of interventions among sex workers and their clients. (C) 2001 Lippinco tt Williams & Wilkins.