Using HIV/AIDS Surveillance to Monitor Public Health Efforts to Reduce Perinatal Transmission of HIV

Abstract

To the Editor: In 1994, AIDS Clinical Trials Group protocol 076 demonstrated that the risk of perinatal transmission of HIV can be reduced from 25% to 8% by administering zidovudine (ZDV) to selected HIV-infected pregnant women and their newborns (1,2). Timely counseling and voluntary HIV testing of women, particularly pregnant women, will play a critical role in reducing perinatal transmission of HIV. Public Health Service guidelines published in 1995 recommend voluntary HIV counseling and testing of all pregnant women (3). We describe here a means of monitoring progress toward full implementation of the prevention guidelines by using as a baseline the 1993 HIV/AIDS surveillance data, which were collected before the use of ZDV was recommended to prevent perinatal transmission of HIV. In 1993, 21 of 25 states that required reporting of HIV infection (4) also conducted the Survey of Childbearing Women (SCBW), a population-based blinded HIV serosurvey of women giving birth (5). Pregnancy at the time of HIV/AIDS diagnosis is indicated on the HIV/AIDS case report form. Comparing the number of pregnant HIV-infected women reported each year in these states with the number of HIV-infected women giving birth as determined by the SCBW yields a minimum estimate of the proportion of HIV-infected women known to be infected during pregnancy. In 1993, an estimated 1,588 HIV-infected women gave birth and 340 pregnant women were reported with newly diagnosed HIV in the SCBW states. Thus, a minimum of 21% of HIV-infected women giving birth were known to be infected. This proportion varied by state (range 6-47%, median 19%; Table 1). Similar proportions of pregnant women reported with HIV and women determined to be HIV-infected by the SCBW were black (70% and 76%, respectively, in 13 states with adequate SCBW race data) and aged 20-29 years (56% and 61%, respectively, in nine states with adequate SCBW age data). Twenty-seven states require reporting of pediatric HIV infection (4). In many of these states, all HIV-exposed infants (those born to women known to be HIV-infected) are followed to determine infection status. An estimate of the number of perinatally exposed/infected infants identified by the health-care system can be obtained by comparing the proportion of perinatally exposed/infected children born in a given year and subsequently reported to HIV/AIDS surveillance with the estimated number of infants born to HIV-infected women that same year as determined by the SCBW. In 1993, in the 18 states that required reporting of HIV infection in children, followed perinatally exposed infants, and conducted the SCBW, 1,655 infants were born to HIV-infected mothers; of these, 796 (48%) were reported to HIV/AIDS surveillance by March 1995 (range 9-85%, median 57%; Table 2). Similar proportions of perinatally HIV-exposed/infected infants followed through surveillance and infants born to HIV-infected mothers were black (70% and 69% black, respectively, in 14 states with adequate SCBW race data). Overall, in these states, which accounted for approximately 25% of the 7,000 births to HIV-infected women, at least 21% of HIV-infected women who gave birth in 1993 were known to be infected and 48% of HIV-exposed infants born in 1993 had been identified by March 1995. The variation by state in the proportion of pregnant HIV-infected women known to be infected and in the proportion of HIV-exposed infants identified is the result of differences among states in counseling and testing practices and in surveillance methods, which range from receiving only limited information from laboratory reports to conducting active surveillance and follow-up. Increases in prenatal counseling and voluntary testing will result in increases in the number of pregnant HIV-infected women referred to HIV care and reported to surveillance. Increases in the number of reported perinatally exposed/infected infants are also expected as increasing numbers of women known to be HIV-infected give birth. Furthermore, in response to recent public health recommendations, more states are likely to implement pediatric HIV infection reporting. Data presented here combined with other strategies, such as matching of HIV/AIDS registries to birth registries and reviewing medical records of pregnant women reported with HIV and of mothers of HIV-exposed infants, will provide estimates of the number of known HIV-infected women giving birth and the number receiving ZDV. This information is critical for evaluating progress toward implementation of the guidelines to reduce perinatal transmission of HIV in a timely manner. Pascale M. Wortley; Patricia L. Fleming; Mary-Lou Lindegren; Patricia A. Sweeney; Susan Davis Division of HIV/AIDS Prevention; National Center for HIV/STD/TB Prevention; Centers for Disease Control and Prevention; Atlanta, Georgia