Cost-effectiveness of HIV screening for incarcerated pregnant women.

Antiretroviral therapy (ART) initiated on a prenatal basis in HIV-infected pregnant women is a highly effective method for preventing mother-to-child HIV transmission. We developed a decision analytic model to project the clinical and economic outcomes of alternative HIV screening strategies (voluntary prenatal screening [VPS], routine prenatal screening [RPS], and mandatory newborn screening [MNS]) for a high-risk population of incarcerated pregnant women. Data for the decision model came from the HIV voluntary counseling and testing program at Connecticut's sole correctional facility for women and a comprehensive anonymously linked serosurvey of all inmates who entered the facility during the 2-year period beginning in October 1994. Based on serosurvey results, in the absence of any HIV screening program, 2.5 cases of pediatric HIV infection would be expected per 1000 pregnancies. Multiplied by the discounted lifetime cost per case of $247,000, this translates to a cost of $624 per testing-eligible prison entrant. Entrants were considered eligible if they were pregnant and their HIV status was unknown. MNS would save money, cost $364 per eligible entrant, and simultaneously reduce the rate of infections to 1.1 per 1000 pregnancies. Doing both MNS and RPS is most effective in reducing the rate of new infections (down to 0.2 per 1000 pregnancies). It would, however, increase costs to $430 per eligible entrant. This would result in an incremental cost of $73,603 per additional pediatric HIV case averted when compared with MNS alone. If mandatory newborn testing was not considered a feasible option, RPS would dominate VPS and would be cost-saving compared with no screening. RPS compares favorably with alternative uses of HIV prevention and treatment resources. In correctional facilities where voluntary newborn screening is already in place, our findings show that there remains a small marginal benefit to be realized from switching to RPS. In settings where HIV screening policies are not in place, however, the implementation of RPS can be expected to significantly reduce pediatric HIV cases and net health care expenditures.