Ethnic disparity in HIV prevalence and zidovudine treatment among childbearing women and pediatric AIDS cases in California

Abstract

Objective: 

To compare, by ethnicity, the prevalence of HIV and zidovudine treatment among a cross-sectional survey of childbearing women in California in 1998, and the number of pediatric AIDS cases from 1998 to 2001.

Methods: 

Blood specimens, collected via infant heel-stick for metabolic screening during the third quarter of 1998, were anonymously tested for HIV antibody. Positive specimens were subsequently tested for evidence of zidovudine therapy. Pediatric AIDS cases with diagnosis dates from 1998 to 2001 were obtained from the AIDS case registry.

Results: 

Of the 119 108 specimens tested, 77 (0.65 per 1000) were HIV-antibody positive. Most (37.7%) of the 77 HIV-positive specimens were from newborns of African-American mothers, followed closely by Latina mothers (35.1%). The absence of zidovudine therapy was highest for Latina and African-American women, 29.6 and 24.1%, respectively. Latino and African-American children accounted for the majority of California pediatric AIDS cases diagnosed between 1998 and 2001.

Conclusion: 

Innovative approaches are needed to increase the rate of zidovudine therapy among African-American and Latina HIV-infected childbearing women. These could include a shorter-course zidovudine regimen or rapid HIV testing and counselling of women in late-term pregnancy.

Introduction

Almost all HIV infections in young children (< 14 years) and 90% of pediatric AIDS cases are attributed to mother-to-child (MTC) transmission [¹]. In 1994, the AIDS Clinical Trials Group Protocol 076 (ACTG) demonstrated that multi-course zidovudine treatment for the mother and neonate resulted in a dramatic reduction of MTC HIV transmission, from 25.5 to 8.3% [²]. This treatment regimen consisted of antepartum and intrapartum zidovudine for the mother and 6 weeks of zidovudine for the neonate.

The ACTG Protocol provided the basis for the Public Health Service's guidelines for the use of zidovudine to reduce the risk of MTC HIV transmission [³]. Established in 1995, these guidelines called for routine HIV counselling and voluntary testing for all pregnant women, and outlined a program that healthcare providers should follow to increase counselling and testing [⁴]. These recommendations and guidelines, combined with the widespread use of multi-course zidovudine therapy, have contributed to the decline of perinatal HIV transmission cases in the United States, from over 600 cases in 1995 to 196 cases in 1999 [^5,6]. Large-scale reductions in MTC transmission have also been achieved in other developed countries by introducing zidovudine and other antiretroviral regimens [^7,8].

Recent randomized clinical trials support the safety and efficacy of zidovudine, as found in the ACTG 076 study. The Bangkok Perinatal HIV Transmission Study Group found a 50% reduction in MTC transmission among childbearing women in Thailand who were given a shorter course of zidovudine therapy (from 36 weeks’ gestation until delivery) [⁹]. The 24-month longitudinal study of Leroy et al. [¹⁰] in West Africa also showed reduced MTC HIV transmission with short-course zidovudine therapy among women with CD4 cell counts greater than 500 cells/ml. In addition, the PETRA Study Group reported increased efficacy with the combination of short-course peripartum zidovudine and lamivudine therapy, when compared with zidovudine therapy alone [¹¹].

Although significant progress in reducing MTC HIV transmission has been made in the USA, pediatric HIV infection still elicits public health concern. Women of childbearing age constitute the fastest growing group of newly reported AIDS cases in the USA [¹²]. Collecting population and zidovudine treatment data is crucial in ascertaining the scope of the HIV epidemic among childbearing women and their newborns.

In 1998, the California Department of Health Services, Office of AIDS collaborated with the Genetic Disease Branch and the Viral Rickettsial Disease Laboratory to participate in the national Seroprevalence Survey Among Childbearing Women [¹³]. This study allowed for the collection of dried-blood spot specimens for HIV antibody testing. To assess the prevalence of zidovudine treatment among HIV-infected childbearing women, the Newborn Screening Quality Assurance Program at the Centers for Disease Control and Prevention (CDC) measured zidovudine levels in HIV-positive specimens. The purpose of our study was to compare, by ethnicity, the prevalence of HIV and zidovudine treatment among this 1998 cross-sectional survey of childbearing women, and to examine the number of pediatric AIDS cases from this year until 2001.

Methods

Blood specimens, routinely collected onto filter paper via infant heel-stick for metabolic screening, were anonymously tested for HIV antibody [¹⁴]. A total of 119 108 dried-blood specimens were collected in California during the third quarter of 1998. For each specimen, the Genetic Disease Branch provided information regarding the mother's race and ethnicity.

Specimens were tested for HIV antibody using enzyme-linked immunosorbent assay. Repeat and confirmatory testing (mini Western blots) were also performed by the Viral Rickettsial Disease Laboratory, in accordance with the 1989 CDC Neonatal HIV-1 Laboratory Procedures [¹⁵]. HIV-positive specimens were sent to the CDC Newborn Screening Quality Assurance Program. To measure zidovudine in HIV antibody-positive specimens, the CDC modified the zidovudine Trac Radioimmunoassay Kit, an assay used to determine zidovudine levels in serum [^16,17]. Calibration and quality control of the modified assay was assured by preparing dried-blood specimens from blood with known zidovudine concentrations, and confirming known zidovudine concentrations by mass spectrometry. The CDC also used the assay to measure zidovudine in the eluate that remained after testing the dried-blood specimens for HIV antibodies. The samples consisted of eluates from previously dilated and tested specimens.

Pediatric AIDS cases with diagnosis dates from 1998 to 2001 were obtained from the AIDS case registry in April 2002. In California, over 90% of AIDS cases are reported to the Office of AIDS 6 months from diagnosis.

Results

Of the 119 108 specimens tested, 77 were HIV antibody positive, for an HIV seroprevalence rate of 0.65 per 1000 women giving birth in California. Most (37.7%) of the 77 HIV-positive specimens were from newborns of African-American mothers, followed closely by Latina mothers (35.1%) (Table 1). African-Americans represent 6.4% and Latinos account for 32.4% of Californian individuals, according to Census 2000 data [¹⁸]. Specimens from infants delivered by white women accounted for 16 cases (20.8%); 46.7% of Californian individuals are white [¹⁸].

Table 1: Number of HIV infections and absence of zidovudine among HIV-positive childbearing women in California, 1998, and reported pediatric AIDS cases, 1998–2001.

Overall, the prevalence of zidovudine among HIV-infected childbearing women was 76.6%; the absence of zidovudine was detected among 23.4% of HIV-positive mothers. The absence of zidovudine therapy was highest for Latinas and African-Americans, 29.6 and 24.1%, respectively.

From 1998 to 2001, 41 pediatric AIDS cases were diagnosed and reported to the Office of AIDS. Latino (n = 18) and African-American (n = 15) children accounted for the majority of California pediatric AIDS cases during this 4-year period.

Discussion

Latina and African-American mothers in California were more likely to have HIV in 1998, but were less likely to have had zidovudine therapy. The ethnic disparity of HIV cases and zidovudine therapy among childbearing women may have played a role in the disproportionate number of Latino and African-American children subsequently diagnosed with AIDS. We found that the majority of pediatric AIDS cases from 1998 to 2001 were among Latino and African-American individuals. Together, African-American and Latina individuals accounted for 72.7% of HIV cases among childbearing women, 83.3% of cases in which HIV-infected mothers had not had zidovudine therapy, and 80.5% of recently reported pediatric AIDS cases.

One explanation for the lack of equity in zidovudine therapy may be because African-American and Latina women do not receive such timely prenatal care. Inadequate prenatal care among African-American and Latina women is well documented [^19,20], and may be partly the result of differing perceptions of when prenatal care should be initiated [^21,22]. Prenatal care later in pregnancy (or no prenatal care) probably reduces the chance that women will receive HIV counselling and be offered an HIV test. In such cases providers may feel compelled to spend time during clinical visits on issues perceived to be of greater importance than counselling patients on the risks and benefits of taking an HIV test. Alternatively, prenatal care providers in California may be spending insufficient time discussing HIV with patients. One recent study [²³] suggested that the difference between providers’ stated HIV counselling and testing practices and the related experiences of patients may be caused by providers not placing sufficient emphasis and relevance on HIV counselling and testing. As such, our finding for Latina and African-American women may not be because they refused or were not informed of zidovudine treatment, but because they were not presented with the option of learning their HIV serostatus. Similarly, prenatal care providers in California may not explain zidovudine therapy in culturally sensitive ways to non-white HIV-positive women. Finally, factors for which we could not control, such as socioeconomic status or a history of drug use, may be responsible for the lack of evidence of zidovudine therapy among HIV-positive Latina and African-American mothers. The anonymous nature of the survey did not facilitate the gathering of other variables that are associated with a lack of zidovudine therapy in HIV-infected childbearing women.

Innovative approaches are needed to increase the rate of zidovudine therapy among African-American and Latina HIV-infected childbearing women. The Thailand clinical trial of a shorter-course zidovudine regimen, which is simpler, shorter, and less expensive than the ACTG 076 protocol, may be more feasible in a healthcare setting where HIV-infected women do not receive adequate prenatal care [⁹]. In addition, in light of recent research that post-exposure zidovudine prophylaxis within 12–24 h of birth can reduce MTC transmission, the initiation of rapid HIV testing and counselling of women in late-term pregnancy would be prudent [²⁴]. Recently, the CDC has funded this type of innovative rapid testing program – MIRIAD (Mother Infant Rapid Intervention at Delivery). MIRIAD focuses on five minority communities with inadequate prenatal care and a relatively high seroprevalence of HIV among women of childbearing age [⁵]. MIRIAD will give women the opportunity to be rapidly screened and tested for HIV before labor. If the woman tests HIV-positive, then she will be counselled about receiving intrapartum and neonatal zidovudine therapy. This rapid intervention before labor may prevent several pediatric AIDS cases each year.

Assessing the prevalence of zidovudine therapy in HIV-infected childbearing women is crucial in allocating HIV prevention, counselling, testing, and treatment resources. The recent identification of seven HIV-infected children (four Latino, three African-American) within a one-month period in Los Angeles County underscores the need for the detection of childbearing women who do not receive HIV testing or zidovudine prophylaxis, and the development of effective community-based interventions and outreach programs [²⁵]. The successful implementation of these programs may dramatically reduce the number of pediatric AIDS cases in California.

Acknowledgements

The authors would like to thank the following individuals: Mr Michael Montgomery, Ms Barbara Bailey, Mr Jim Creeger, and Ms Mi Chen from the California Department of Health Services, Office of AIDS; Ms Vanessa Baird from the Office of Multicultural Health; Dr Richard K. Sun from the California Department of Social Services; Dr Yvonne A. Maldonado and her staff from the Stanford University School of Medicine; Ms Paola Grasso, Ms Jennifer Marino, and Ms Shira Shafir from the University of California, Berkeley; Dr Jeffrey Chang from the University of California, San Francisco; Dr George Cunningham, Dr Fred Lorey, and Ms Corinna Tempelis from the Genetic Disease Branch; Dr John Sherwin and Ms Terry Kennedy from the Genetic Disease Laboratory; Dr Michael S. Ascher, Ms Cynthia K. Cossen, Ms Elizabeth Baylis, Ms Natasha Huntziker, Ms Jennifer Josel, Ms Michelle Cameron, Mr Chris Rogers, Ms Caroly Floyd, Mr Brent Nakagiri, Mr Gordon Shell, Ms Caroline Madrilejo, Mr James Barry, Mr Oliver Oyler, Ms Analisa Weston, and Ms Ruth Fukuchi from the Viral and Rickettsial Disease Laboratory.

The authors will to express their appreciation to Dr Joanne Mei, Ms Nancy Meredith, and Ms Denise Kothe from the Newborn Screening Quality Assurance Program at the Centers for Disease Control and Prevention for screening for zidovudine in California dried-blood spot specimens.