Correlates of prenatal HIV testing in women with undocumented status at delivery.

OBJECTIVE: To determine factors associated with prenatal HIV testing in women who accepted rapid testing at delivery. METHODS: The mother-infant rapid intervention at delivery (MIRIAD) protocol offered counseling and voluntary HIV testing in six US cities including New York City (NYC). These hospitals are required to document the HIV status of pregnant women or their infants. From January 2002 to January 2005, 653 HIV-negative women were interviewed post-partum. RESULTS: 63% of women reported prior HIV testing during the index pregnancy, although their results were not available at delivery. Multivariate logistic modeling identified receipt of prenatal care and delivery in NYC as being associated with having been offered prenatal HIV testing. In a model restricted to women receiving medical care, emergency department (ED) use and delivery outside of NYC were associated with not having been offered testing. In a model restricted to women who were offered prenatal testing, acceptance was associated with delivery outside of NYC. CONCLUSIONS: Improved documentation of prenatal test results, expanded prenatal testing in non-traditional settings like EDs, and routine voluntary "opt-out" testing during pregnancy may further reduce perinatal HIV transmission.