been passed, according to the committee's site visits. Women also fear losing custody of their children if their drug use is discovered.

Correctional settings—prisons and jails—offer a unique opportunity for prevention activities targeted at hard-to-reach women at risk for, or already infected with, HIV. The total number of incarcerated women was 74,730 in 1996, a threefold increase from 1985 (Bureau of Justice Statistics, 1997). The prevalence of HIV infection among incarcerated women is far higher than in the general community: 4% of female state prison inmates nationwide are known to be HIV-positive, with the proportion exceeding 10% in nine states. Women are more likely than men to be incarcerated for drug offenses. In Rhode Island, for instance, nearly half of incarcerated women are imprisoned for drug-related charges (Flanigan, 1998). Consequently, they generally serve shorter sentences and return to the community, where many will re-offend.

The proportion of pregnant women in correctional settings who are HIV-infected is not known, but can be inferred to be higher than that in the general community. The median age of incarcerated women (31 years) places them squarely in the reproductive period. Furthermore, 6.1% of women in state prisons in 1994 were pregnant upon admission. Women in correctional settings thus represent an important population for targeted prevention efforts. Despite the relatively high-prevalence of HIV and pregnancy, only 85% of pregnant women received a gynecological exam related to pregnancy upon admission, and only 69% received any prenatal care while incarcerated.

Many interventions could be introduced in correctional settings either for general primary prevention of HIV transmission or for prevention of perinatal transmission among HIV-infected pregnant women in particular. Interventions could focus on HIV testing and treatment, drug testing and treatment, prenatal care, and efforts to ensure continuity of care for HIV-positive patients who leave the correctional setting. Given the realities of the correctional system, however, utmost care must be taken so that interventions are seen to be in the best interest of those incarcerated. Interventions that take advantage of prisoners to protect others, especially if the interventions lack confidentiality and may put prisoners at risk for harm, can be counter productive.

The Rhode Island prison system provides a model comprehensive HIV testing and care program that is integrated with the community. The proportion of HIV-positive women in Rhode Island prisons at any given time is between 8% and 12% (Flanigan, 1998). As outlined to the committee at one of its workshops, Rhode Island mandates HIV testing for all individuals upon prison intake. For infected individuals, complete HIV care is available, and HIV patients are successfully linked to follow-up care in the community (Appendix D; see also Flanigan, 1998). The Rhode Island program has had a tremendous impact on HIV