in 1997. Byers and colleagues feel that these decreases are consistent with, and in large part reflect, widespread implementation of the ACTG 076 regimen.

CDC scientists feel that, collectively, these trend data point to the conclusion that declines in pediatric AIDS, particularly among infants and particularly since 1994, are principally related to declines in perinatal transmission rates with increasing use of maternal and newborn zidovudine (ZDV). While the declines actually precede some of the PHS recommendations, they likely reflect the impact of pregnant women using ZDV for their own health. In addition, since ACTG 076 results were published in February 1994, four months before the PHS recommendations for use of ZDV to reduce perinatal transmission (published in August 1994), some women may have received ZDV in early 1994 based on the clinical trial findings. Also, women were treated for their own health in the 1990s, including as many as 20% of pregnant HIV-infected women. Other factors such as increasing use of therapy among HIV-infected children may also be playing a role by delaying the onset of AIDS; however, it should be noted that the use of combination therapy with potent protease inhibitors was not the standard of care for children during the period of rapid decline.

As a framework for understanding the impact of efforts to prevent perinatal HIV transmission, CDC representatives presented its data in terms of a chain of events or steps that must be taken to ensure prevention success. The chain is based on ensuring timely and complete implementation of the ACTG 076 regimen and includes the following steps: (1) receipt of early prenatal care (depends upon access to and utilization of care); (2) provider offering of counseling and testing (depends upon health care provider knowledge, attitudes, beliefs, and practices); (3) client acceptance of testing; (4) HIV-positive client acceptance of ZDV (depends upon provider offering therapy); (5) ZDV adherence (requires taking ZDV during the antepartum, intrapartum, and postpartum periods); and (6) follow-up care for both mother and baby.

Compared to the general population, HIV-infected women are much more likely to receive late or no prenatal care. Provisional STEP data indicate that only 63% of HIV-infected women giving birth received prenatal care prior to the third trimester. This compares to 95% to 97% of women in the general population (based on National Center for Health Statistics 1994 natality data and PRAMS data from 11 reporting states). As in the general population, prenatal care use among HIV-infected women varies by race and ethnicity, with African-American and Hispanic women likely to have fewer prenatal visits. The strongest predictor of inadequate prenatal care among HIV-infected women, however, is illicit drug