HIV Screening and Treatment of Pregnant Women and Their Newborns: A Simulation-Based Analysis

Abstract

Clinical trials have shown that HIV-positive preg nant women treated with zidovudine during preg nancy and delivery have a much lower rate of HIV transmission to their newborns. Recent studies have also shown the effectiveness of antiretroviral therapy combined with elective cesarean delivery and a short- term antiretroviral therapy. In this study a dynamic computer model was built. The model was used to estimate economic consequences of testing pregnant women for HIV virus and treating the HIV-positive women and their newborns with three different reg imens. Outcomes of data analysis included costs and the number of avoided cases of perinatal transmission of the HIV virus. The study found that while the over all costs of the three treatment programs were simi lar, the short-term antiretroviral program was the least cost-effective and prevented fewer cases of peri natal transmission of the virus. Sensitivity analysis found that the cost-effectiveness of the treatment pro grams were differentially sensitive to screening and treatment rates. Cost-effectiveness of all regimens is highly sensitive to the HIV infection rate among preg nant women. When the rate reaches four to five per 1,000, treatment cost savings from the reduction in pediatric HIVcases exceed the costs of screening and treatment. The model presented provides a useful tool to be used to evaluate treatment programs designed to pre vent mother-to-infant transmission of the HIV virus.