Published online by Cambridge University Press: 12 January 2001
The diffusion potential of a targeted HIV/AIDS intervention that enlisted peer educators to disseminate ‘safer sex’ messages and condoms among female commercial sex workers and their clients was evaluated in the Dominican Republic. Levels of interurban interaction potential were ascertained that linked the targeted city of La Romana with the proximate cities of San Pedro de Macoris and Guaymate. Weekly service statistics generated over an 8-month period were analysed to establish activity areas for the peer educators. Data were entered and analysed using a geographic information system and interurban linkages were established. Project outcomes were examined via a series of three cross-sectional Knowledge, Attitudes and Practices (KAP) surveys conducted among convenience samples of commercial sex workers at the start of the intervention and at 4 and 8 months. The results attest to a high degree of interconnectivity between the targeted and proximate cities, and a pattern of interurban mobility that links commercial sex workers, clients and establishments in all three cities. The examination of project outcomes revealed statistically significant changes in condom use in the targeted city of La Romana among commercial sex workers, as well as among their counterparts interviewed in the proximate cities of San Pedro de Macoris and Guaymate. These data suggest a diffusion effect. It is concluded that a targeted intervention may influence proximate cities within a relatively compressed period of time. The findings suggest the importance of considering geographic diffusion principles, such as urban hierarchies, regional nodes and transportation linkages, when designing HIV/AIDS prevention efforts. It also has important implications in the selection of control sites when conducting experimental studies of HIV/AIDS interventions.