Segmentation of heterosexual prostitution into various forms: a barrier to the potential transmission of HIV.

The results of an earlier Dutch study on prostitutes' clients, interviewed face-to-face, were cross-validated by means of telephone interviews, which both guaranteed anonymity and produced a larger sample of respondents (n = 559). Depending on their preference and financial resources, they had visited different types of prostitution; but about half of them had restricted themselves to one type. Inconsistent use of condoms was reported by 14% of the clients who had had vaginal or anal intercourse. Condoms were most frequently used in clubs, brothels and window prostitution, and least often in street, home and escort prostitution; in the last two sorts especially not with regular prostitutes. Whether or not prostitutes came from non-Western countries had no influence on protective behaviour of these clients. The formation of networks between different sorts of prostitution, through unsafe sexual contacts in two or more sorts, involved only 3% of respondents. Thus, prostitution in The Netherlands should be seen as a number of sparsely-connected networks rather than as a single network. The formation of networks between prostitution and the population at large is made possible by the 10% of the respondents who had had unsafe contacts with both commercial and private partners.