Between 1996 and 2000, more than half of HIV-infected infants in a large pediatric HIV surveillance project were born to women who missed opportunities to obtain perinatal HIV prevention care--18% to women who received no prenatal care, 29% to women who had prenatal care but were not tested for HIV before delivery, and 9% to women who had HIV diagnosed during pregnancy but did not receive antiretroviral treatment.1 The risk of mother-to-child transmission of the virus was sharply reduced among women who received any preventive intervention during pregnancy; it was virtually eliminated among those who were treated with zidovudine in combination with other antiretroviral drugs.

The multisite project, funded by the Centers for Disease Control and Prevention, collects medical record data on children born to HIV-infected women. It is partly hospital-based and partly population-based, and the children on whom it gathers data are representative of all children born to U.S. women with HIV. Analysts used data on the 4,755 singleton births documented by the project during 1996-2000 to examine trends in use of perinatal HIV prevention methods, mother-to-child transmission rates and missed opportunities for perinatal preventive measures.

Six in 10 mothers included in the analyses were black, three in 10 were Hispanic and most of the rest were white. Seventy-nine percent had Medicaid or other public insurance coverage for their children, 7% had private insurance and 4% had no coverage; for 10%, insurance coverage was unknown. About half of mothers supplied information on drug use; 23% overall reported using illicit drugs.

The proportion of women known to have received prenatal care rose from 79% of those who gave birth in 1996 to 88% of those who delivered in 2000; data were not available for 10% of women. The prevalence of cesarean delivery, which is recommended for women with HIV infection, also grew over the study period-- from 20% to 48%.

Among women who received prenatal care, 90% in 1996 and 94% in 2000 had HIV diagnosed during pregnancy; 78% and 87%, respectively, had antiretroviral drugs prescribed to prevent perinatal transmission of the virus. Whereas 71% of preventive drug regimens prescribed in 1996 relied on zidovudine alone, only 9% took that approach in 2000; the proportion combining zidovudine with other antiretroviral drugs increased from 6% to 70%. Seventy-nine percent of all infants born in 1996 received zidovudine treatment, compared with 92% in 2000.

Nine percent of infants born during the study period were HIV-infected, and 77% were not; the infection status of 14% could not be determined. Among the infected infants whose records included information on the services their mothers received during pregnancy, 44% were born to women who had prenatal care, an HIV diagnosis during pregnancy and treatment to prevent transmitting the virus to their newborn. For the other 56%, the opportunity to take preventive measures was missed because the mother either received no prenatal care (18%), received prenatal care but not an HIV test (29%), or received care that included an HIV test but not antiretroviral therapy (9%). By comparison, among uninfected infants, only 16% had any missed opportunity for preventive care.

A greater proportion of women who reported illicit drug use than of those reporting no such use received no prenatal care (17% vs. 3%). Results of logistic regression analysis confirmed this association: When project site, year, and mother's race or ethnicity and insurance status were controlled for, users of illicit drugs were significantly more likely than nonusers to have received no prenatal care (odds ratio, 8.0).

In a second logistic regression analysis, adding prenatal interventions and type of delivery to the set of controls, the researchers examined the factors associated with mother-to-child HIV transmission. The strongest associations were with prenatal interventions: Compared with the risk of transmission in the absence of HIV testing and antiretroviral therapy, the risk was significantly reduced if the woman had been tested and had used zidovudine alone during pregnancy (odds ratio, 0.1) and, especially, if she had been tested and taken a prenatal drug regimen combining zidovudine and other antiretroviral agents (0.08). The transmission risk was lower in each of the last three years of the study period than in the first two (0.5-0.6) but was not associated with any of the other factors examined.

While the analysts stress that perinatal HIV transmission is largely preventable through the use of recommended prenatal interventions, they also emphasize that failures of these interventions are not uncommon and that investigation into the causes of those failures is imperative. Given their findings on missed opportunities for and failures of preventive measures, they conclude that the "most important" challenge in preventing mother-to-child HIV transmission is to ensure that all women are offered HIV screening prenatally or, if their HIV status is unknown, during labor and delivery, and that "comprehensive interventions are then aggressively offered."--D. Hollander


1. Peters V et al., Missed opportunities for perinatal HIV prevention among HIV-exposed infants born 1996-2000, Pediatric Spectrum of HIV Disease cohort, Pediatrics, 2003, 111(5):1186-1191.