Summary: 434 male and 145 female prisoners were available to participate in crosssectional, voluntary anonymous HIV surveillance (using saliva samples) with linked self-completion questionnaire at HMP (Her Majesty's Prison) Perth on 17 May and at HMP Cornton Vale on 18 May 1995. Three hundred and four men (70%) and 136 women (94%) completed a risk-factor questionnaire and 304 and 135 samples were received for HIV antibody testing. Two hundred and eighty-two and 132 questionnaires passed logical checks. Six saliva samples from Perth (all injectors) out of 304 and none from Cornton Vale out of 134 tested were HIV antibody positive. Four were presumptively from known HIV-infected male inmates; the other 2 were local men, under 26 years, who began injecting in 1989-91, and both reported having had a recent HIV test. Overall HIV prevalence was estimated at 2% compared to a known prevalence of 1.4% (6/434), giving a 1.5 ratio of overall: disclosed HIV prevalence at HMP Perth. HIV prevalence was estimated at 7% (6/82) for injector-participants and 14% (5/35) for local injector participants. At Cornton Vale, where both known HIV-infected inmates abstained, overall and disclosed HIV prevalence were equal at 1.4%. At Perth Prison, 29% of prisoners had injected drugs (82/278); 85% of injector-inmates reported having injected inside (some) prison and 31% (25/80) had started to inject while inside, 7 during their present sentence. Of all 21 injector-inmates who first injected after 1991, 10 hadstarted to inject inside, including one of 69 male inmateswho had never been inside before. The corresponding figures for Cornton Vale, where 46% of inmates were injectors (58/132), were that 57% of injector-inmates had injected inside (32/56) but only one woman, for whom this was not her first sentence, had started to inject inside. Twenty-eight per cent of male prisoners (78/277) and 57% of male injector-inmates (47/82) had had a personal HIV test since January 1993, as had 35% of female prisoners (43/124) and 57% of female injector-inmates (30/53). A much higher proportion of Glasgow's female prisoners (64%: 38/60) were injectors than of women prisoners from the Edinburgh, Dundee and Fife area (21%: 5/26) or from elsewhere (34%: 15/45). Rape was reported by 23% of women (30/130). Women who had been raped had a more polarized distribution of male sexual partners (none to 2 plus) in the year before sentencing than other women and were more likely to report anal sex (11/30 vs 11/100, P< 0.001). Prostitution had been engaged in by 19% of female injector-inmates (11/57) and was acknowledged by one other woman. However, only 5% of women (6/130) reported ever having been treated for an STD.
The vulnerability of Brazilian female prisoners to HIV infection
