The reduction of mother-to-child (perinatal) transmission of the human immunodeficiency virus (HIV) is a great triumph of HIV prevention in industrialized countries.1–3 In New York State, a combination of educational, policy, and regulatory interventions has led to near universal HIV counseling and voluntary testing during pregnancy and access to antiretroviral medication to prevent perinatal transmission (Birkhead G, Kowalski E, Hackel S, Pulver W, Glaros R, Warren B. The contribution of infection acquired during pregnancy to residual mother-to-child HIV transmission in New York State [abstract WEPE0271]. XVI International AIDS Conference; 2006; Toronto, Canada; and Pulver WP, Smith L, Glaros R, Warren BL, Kowlaski EF, Hackel SE, et al. Validation of prenatal HIV testing status as reported to the Newborn Screening Program by birth facilities in New York State. 2005 National HIV Prevention Conference; June 12–15, 2005; Atlanta, GA).4,5
With implementation of the Newborn Screening Law in 1997, the Newborn Screening Program at the New York State Department of Health Wadsworth Center Laboratories began to conduct HIV antibody testing on the filter paper blood specimens submitted for metabolic screening of all newborns with results returned to the hospital of birth and the pediatrician.6.7 At this time, numerous programmatic initiatives were enacted or enhanced to ensure appropriate HIV-related care for pregnant women and their newborns. These included free diagnostic testing for exposed newborns through the New York State Department of Health Diagnostic HIV Laboratory and monitoring of regulatory adherence through review of maternal HIV testing rates reported by birth facilities and review of medical records for each HIV-exposed birth.
In 1999, regulations established an expedited testing program to help identify infected women presenting for delivery who had not received prenatal testing.8,9 Mother-to-child HIV transmission has fallen to the lowest levels in New York since data collection began in 1997.
Despite these efforts, perinatal transmission still occurs (Pulver W, Smith L, Warren B, Birkhead G. Optimum care and prevention of mother to child HIV transmission in New York State: 2003 and 2004. 2007 National HIV Prevention Conference; 2007; Atlanta, GA.).10,11 This analysis uses population-based data to assess factors possibly associated with continued perinatal HIV transmission in New York State with a focus on acquisition of maternal HIV infection during pregnancy and its role as a risk factor for mother-to-child HIV transmission.
MATERIALS AND METHODS
Analysis of data on HIV-exposed births in New York State from January 1, 2002, through December 31, 2006, was conducted by New York State Department of Health staff as part of a continuing evaluation of regulatory compliance and quality of care that is an ongoing assessment of the statewide comprehensive maternal and pediatric HIV prevention and care program. Access to and analysis of the data are authorized by, and is part of, the Department's ongoing regulatory responsibilities and therefore, the New York State Department of Health Institutional Review Board determined it did not require institutional review board review. Racial and ethnic information was collected from medical records and newborn screening specimen collection forms completed by medical personnel in each birth hospital. Race and ethnicity were assessed because of importance of the information in guiding policy and resources given the disproportionate effect of the epidemic among racial and ethnic groups.
The New York State Department of Health Wadsworth Center Laboratories Newborn Screening Program received newborn filter paper blood specimens from live births for metabolic screening and HIV antibody testing. Birth hospitals reported the prenatal and expedited HIV testing history from the pregnancy and delivery on the blood collection form as well as selected demographic variables and birth weight. HIV antibody testing by enzyme immunoassay (EIA) was performed on all filter paper bloodspots with Western blot testing conducted to evaluate an initial reactive EIA.
All specimens from HIV-exposed neonates submitted to the New York State Department of Health Wadsworth Center Diagnostic HIV Laboratory for HIV diagnostic testing first received either an EIA or Multispot test.12,13 Enzyme immunoassay or Multispot reactive specimens received a Western blot on initial specimen submission. DNA polymerase chain reaction (PCR) was performed on confirmed exposed infants for diagnosis.12
After each HIV-exposed birth, the mother's prenatal and delivery records and the newborn's birth and pediatric records up to 6 months of age were reviewed to allow assessment of key variables, including diagnostic testing and HIV-related treatment provided to the mother and infant.4 Data on demographic factors, diagnosis and treatment of HIV infection, prenatal care, and other factors were abstracted using a standard protocol and tool.
The cohort of New York State newborns with HIV exposure at birth was established primarily through confirmed positive Newborn Screening Program HIV antibody tests. However, several other reports routinely received by the New York State Department of Health were reviewed to ensure as complete an ascertainment as possible: 1) forms completed by birth facilities documenting reactive EIA or rapid test results from expedited testing of mothers during labor and delivery, expedited testing of newborns, or both; 2) pediatric HIV-related testing results generated from diagnostic testing of clinical specimens by the New York State Department of Health Diagnostic HIV Laboratory; and 3) routine clinician reports of new HIV infections submitted to the surveillance unit of the New York State Department of Health as required by law.14 Multiple variables inclusive of name, sex, date of birth, and birth facility were used to establish a deduplicated list of exposed newborns for the cohort; if questions remained, medical record review was used to clarify identity. The prenatal and expedited testing history, Newborn Screening Program test results, pediatric HIV diagnostic testing, and medical record abstraction data on each case were aggregated to allow epidemiologic analysis, assessment of regulatory adherence, and follow-up of exposed newborns. To ensure that each pregnancy of an HIV-infected woman that resulted in the live birth of a newborn was represented only once in the cohort, only one newborn from the pregnancy was assessed. Two pregnancies involving multiple births had a mother-to-child HIV transmission with twins discordant for HIV infection. In both of these cases, the infected twin's data were kept for analysis; in neither case did the mother acquire infection during the pregnancy. Cases were excluded if the newborn's HIV infection status could not be determined or if chart review provided insufficient information for analysis.
Mother-to-child HIV transmission cases were identified by 1) newborn DNA PCR test results obtained through the New York State Department of Health Diagnostic HIV Laboratory; 2) newborn DNA PCR, RNA PCR, or viral culture test results found on review of medical records; or 3) physician report of an infected newborn. In all cases of potential mother-to-child transmission, medical record reviews were conducted to verify transmission. Confirmed mother-to-child transmission was defined as a newborn with a positive DNA or RNA PCR test at any age followed by a second positive DNA or RNA PCR test collected on a different day. Presumed mother-to-child transmission was defined as a newborn with a single documented positive DNA or RNA PCR test with no subsequent negative DNA or RNA PCR. Absence of HIV infection was confirmed if newborns had two negative HIV DNA or RNA PCR tests after 1 month of age with at least one test after 4 months of age. Newborns were presumed to be uninfected if they had at least one negative DNA or RNA PCR test documented after 1 month of age and no positive tests documented by medical record review, the New York State Department of Health Diagnostic HIV Laboratory, or the surveillance unit of the New York State Department of Health. For this analysis, presumed and confirmed infant infection statuses were combined.
Maternal acquisition of HIV during pregnancy was defined by a documented negative maternal HIV test during the pregnancy with subsequent documentation of HIV infection by 1) a second prenatal HIV test, 2) confirmation of a reactive expedited HIV test of either the mother or newborn at delivery, 3) a positive Newborn Screening Program HIV antibody test, or 4) a reactive maternal antibody test 10 days postdelivery with testing prompted by paternal HIV diagnosis (one case).
Univariable analysis of characteristics associated with HIV-exposed births with and without mother-to-child transmission was conducted to identify factors associated with transmission. A univariable analysis of demographic, risk, and medical care information was conducted on HIV-exposed births with and without maternal acquisition of HIV during pregnancy to identify factors related to HIV acquisition during pregnancy in comparison to all HIV-infected mothers. The odds ratios (ORs) and the corresponding 95% confidence intervals (CIs) were calculated using the SAS 9.1 statistical software version (SAS Institute Inc, Cary, NC). A stepwise logistic regression model was fit for the outcome of mother-to-child transmission with inclusion of independent variables with P<.05 in univariable analysis; adjusted ORs and the corresponding 95% CIs were reported.
Newborn screening HIV antibody test results were available for 1,209,441 of 1,224,058 (98.8%) live singleton or the first of multiple births in New York State between January 2002 and December 2006. From data provided on the newborn screening collection form, 95.0% of the women giving birth had documentation of a prenatal HIV test or were known to be HIV-infected before the pregnancy. The prenatal HIV testing variable was not completed for 1.0%. For 4.0%, the status of prenatal testing was unknown at delivery or no prenatal testing was reported; 96.4% of these received expedited testing in the labor, delivery, or newborn setting. The Newborn Screening Program identified 3,385 neonates born to HIV-positive women for an HIV exposure rate at birth of 0.28%.
Figure 1 outlines the criteria that led to the final analysis cohort of 3,102 HIV-exposed birth events. In addition to the 3,385 newborns identified through the Newborn Screening Program, 11 HIV-exposed newborns were identified through positive expedited HIV test reports received from New York State birth facilities, HIV-related testing reports from the State Diagnostic HIV Laboratory, clinician reports, or both of HIV infection in a newborn for a total of 3,396 HIV-exposed newborns identified. Two hundred ninety-four HIV-exposed births (8.7%) were not included in the analysis. Only one newborn from a multiple birth was kept for analysis purposes, leading to the exclusion of 66 newborns. Also excluded were 56 cases with incomplete chart review data and 167 newborns with insufficient data to determine infection status. An additional five HIV-infected newborns with negative maternal prenatal tests and negative Newborn Screening Program tests who appeared to have mother-to-child HIV transmission were excluded from analysis as a result of the inability to determine if the mother became infected just before delivery or shortly after giving birth. Characteristics of the analysis cohort of HIV-exposed delivery events are outlined in Table 1.
Mother-to-child HIV transmission occurred in 65 neonates (2.1%) born to 63 HIV-infected mothers; 3,037 neonates (97.9%) were not infected. Two mothers became pregnant and transmitted HIV twice each during the study period. Transmission was confirmed in 64 newborns and was presumed to have occurred in one additional newborn. In the newborn with presumed infection, the mother acquired HIV during pregnancy. DNA PCR testing through the New York State Department of Health HIV Diagnostic Laboratory served as the diagnostic test(s) for 57 (87.7%). Documentation of infection for the remaining eight infected newborns was obtained through a combination of DNA PCR testing done at the New York State Department of Health Laboratory and testing documented through medical record review or the surveillance unit of the New York State Department of Health. Absence of HIV infection was confirmed in 2,419 and presumed in 618 newborns.
Among the 2,031 mother–infant pairs receiving a level of care inclusive of at least three prenatal visits and all three arms (prenatal, intrapartum, and neonatal) of antiretroviral medication, 18 (0. 9%) had mother-to-child transmission, whereas among those 1,071 mother–infant pairs without this level of care, 47 had mother-to-child transmission for a transmission rate of 4.4%. Further mother–newborn characteristics and univariable analyses are shown in Table 1.
The results of stepwise logistic regression analysis assessing key risk factors significantly associated with mother-to-child transmission on univariable analysis (exclusive of year of birth and length of membrane rupture) is shown in Table 2. Transmission was significantly more likely to occur when mothers acquired infection during pregnancy (OR 15.19, 95% CI 3.98–56.30), maternal diagnosis was at delivery or after the child's birth (OR 3.24, 95% CI 1.15–8.15), when mothers had documented illicit substance use during the pregnancy (OR 2.66, 95% CI 1.33–5.27), when mothers did not have three or more prenatal care visits (OR 2.37, 95% CI 1.11–4.91), and when newborns weighed less than 2,500 g at birth (OR 2.46, 95% CI 1.26–4.74). As a result of one or more missing variables, 660 (21.3%) cases were excluded from this analysis.
Maternal acquisition of HIV during pregnancy was confirmed in 41 cases for an estimated prevalence of 3.4 cases per 100,000 live births. The Newborn Screening Program testing was the primary source for detection of maternal acquisition of HIV during pregnancy, identifying 29 (70.7%) of the cases. A second prenatal test identified 6 (14.6%) instances of maternal acquisition of HIV during pregnancy. An additional 5 (12.2%) cases were identified through expedited testing at labor or delivery. One case (2.4%) was identified shortly after delivery with testing prompted by paternal diagnosis.
Case details of women identified as having acquisition of HIV during pregnancy differed from those who appeared to have acquired their HIV infection before pregnancy (Table 1). Mothers who acquired HIV during pregnancy were more likely to have a vaginal delivery (63.4%) compared with mothers who did not acquire HIV during pregnancy (43.8%) (OR 2.17, 95% CI 1.14–4.09). Mothers who acquired their HIV infection during pregnancy were more likely to have membrane rupture of 2–4 hours (19.5%) compared with mothers who did not acquire HIV during pregnancy (8.2%) (OR 3.02, 95% CI 1.31–6.92). No difference was observed in race, maternal age, prenatal care, sexually transmitted infections, illicit drug use, or infant birth weight. However, these mothers were much more likely not to have received prenatal antiretroviral medication (82.9%) compared with mothers who did not acquire their infection during the pregnancy (8.1%) (OR 72.78, 95% CI 28.21–187.75), and they were much less likely to receive antiretroviral medication during labor or delivery (80.5%) compared with women not acquiring infection during pregnancy (10.5%) (OR 34.53, 95% CI 15.81–75.40). The newborns of mothers who acquired HIV during pregnancy were more likely not to have received antiretroviral medication (29.3%) compared with the newborns whose mothers did not acquire HIV during pregnancy (0.4%) (OR 187.81, 95% CI 73.44–480.29).
Twenty-two percent of neonates born to mothers who acquired HIV during pregnancy became infected with HIV compared with 1.8% of newborns whose mothers did not acquire HIV during pregnancy (OR 15.09, 95% CI 6.88–33.10). Although maternal acquisition of HIV during pregnancy was documented in only 1.3% of perinatal HIV exposures, it was associated with 9 of the 65 (13.8%) mother-to-child HIV transmission cases.
New York State data show that a multifaceted approach emphasizing voluntary routine prenatal testing and prenatal, intrapartum, and neonatal antiretroviral medication can lead to sustained low levels of mother-to-child HIV transmission. For women who receive standard health care for prevention of perinatal transmission (prenatal care and three-part antiretroviral medication), the risk of transmission is very low. However, previously identified factors, including lack of maternal and newborn antiretroviral medication, maternal illicit drug use, and breastfeeding, continue to be associated with perinatal transmission.11
Although acquisition of HIV during or shortly after pregnancy has been well documented, based on our search of MEDLINE (search terms: “HIV,” “Seroconversion,” “Pregnancy,” “Acute,” “Infection,” “Incidence,” “Risk Factors,” and “Primary Infection,” 1966–2009), this is the first population-based report to assess maternal acquisition of HIV during pregnancy and the increased risk of perinatal HIV transmission.15–21 The results demonstrate that maternal acquisition of HIV during pregnancy is a significant factor in residual perinatal HIV transmission in New York State with one of every seven transmissions associated with maternal acquisition of HIV during pregnancy. Our analysis likely understates the frequency of maternal acquisition of HIV during pregnancy because maternal acquisition of HIV occurring late in pregnancy may not have been detected as a result of absent or low levels of maternal antibody at delivery. Moreover, we excluded from our formal analysis the five mother–newborn pairs with negative newborn screening tests followed by newborn HIV infection that was detected as a result of newborn illness with unknown timing of maternal infection. Based on available data, we judge that these mothers acquired HIV in the very late prenatal or very early postnatal period. Given these possible cases of maternal acquisition of HIV during pregnancy, as many as 20% of perinatal transmission cases may have been associated with maternal acquisition of HIV during pregnancy and transmission through in utero or intrapartum exposure or breastfeeding.
Routine Newborn Screening Program testing played a key role in identifying 29 cases of HIV exposure after negative prenatal HIV tests. This information allowed for cessation of breastfeeding and appropriate diagnostic testing and opportunistic infection prophylaxis for the newborn as well as medical evaluation of the mother. Consistent with Centers for Disease Control and Prevention recommendations for recognition of maternal acquisition of HIV during pregnancy, New York State now recommends repeat HIV testing in the third trimester22,23; this may be of value for other areas of high prevalence as well.17,20,23,24 The 29 HIV-exposed newborns identified through routine Newborn Screening may have been identifiable through third-trimester testing, and earlier recognition would have allowed prophylaxis to prevent mother-to-child transmission at delivery and in the newborn period. However, ensuring third-trimester testing is problematic in premature births or in the absence of routine prenatal care.
Several limitations of our investigation must be acknowledged. Inherent in any retrospective analysis is the potential for misclassification of maternal acquisition of HIV during pregnancy. When HIV infection occurs in early pregnancy such that antibodies are present at the first prenatal test, the association is missed. As previously mentioned, missed diagnoses of maternal acquisition of HIV in late pregnancy are also likely. Cases of transmission may have been missed in the 223 mother–newborn pairs that were excluded from the analysis cohort as a result of indeterminate newborn infection status or incomplete chart review. Newborn infection status for 619 (20%) was classified using a presumed status. However, this limitation is likely to be inconsequential because a prior review of our data showed the presumptive definitions to be highly predictive of final status for statewide surveillance purposes (Pulver W, et al. Timing of diagnosis and HIV infection in infants in New York State [NYS] 1997–2004. Presented at the XVI International AIDS conference, Toronto, Canada, August 13–18, 2006). Assessing the possibility that maternal acquisition of HIV during pregnancy leads to an increased risk of transmission as compared with untreated chronic HIV in pregnancy proved difficult. The numbers of mother–newborn pairs in both groups who received no antiretroviral medication were too small for meaningful direct comparison. In the logistic regression model, an interaction between the variable designating acquisition of HIV during pregnancy and the newborn antiretroviral medication variable could not be adequately assessed as a result of the small numbers of untreated newborns who became infected in both groups. Thus, the model results may not adequately address the relationship between these closely related factors associated with mother-to-child transmission.
Comprehensive perinatal HIV prevention programs that stress prenatal HIV testing, accompanied by expedited testing in the labor setting in the absence of documented prenatal testing, can be successful in dramatically lowering perinatal HIV transmission. However, achieving the goal of elimination of mother-to-child HIV transmission depends on successful approaches to very difficult multifaceted problems,25 including prevention, diagnosis, and management of HIV infection acquired during pregnancy.26
1. Fowler MG, Lampe MA, Jamieson DJ, Kourtis AP, Rogers MF. Reducing the risk of mother-to-child human immunodeficiency virus transmission: past successes, current progress and challenges, and future directions. Am J Obstet Gynecol 2007;197(suppl 1):S3–9.
2. European Collaborative Study. Mother-to-child transmission of HIV infection in the era of highly active antiretroviral therapy. Clin Infect Dis 2005;40:458–65.
3. Lindegren ML, Byers RH Jr, Thomas P, Davis SF, Caldwell B, Rogers M, et al. Trends in perinatal transmission of HIV/AIDS in the United States. JAMA 1999;282:531–8.
4. Public Health Service Task Force recommendations for use of antiretroviral drugs in pregnant HIV-infected women for maternal health and interventions to reduce perinatal HIV transmission in the United States. November 2, 2007. Available at: http://AIDSinfo.nih.gov/ContentFiles/perinatalGL.pdf
. Retrieved February 18, 2008.
5. New York State Department of Health Memorandum, AI 99–01: Maternal–Pediatric HIV Prevention and Care Program: HIV counseling and voluntary testing of pregnant women; routine HIV testing of newborns. August 1, 1999.
6. Newborn screening in New York State. A guide for health professionals: Newborn Screening Program. Wadsworth Center, New York State Department of Health; 2003.
8. 10 NYCRR §69–1.3(l)(2).
9. Letter regarding regulatory changes for expedited HIV testing in Labor & Delivery. Available at: http://www.health.state.ny.us/diseases/aids/testing/perinatal/expedited/2004_prenatal_hiv_testing_letter.htm
. Retrieved April 28, 2009.
10. Wade NA, Zielinski MA, Butsashvili M, McNutt LA, Warren BL, Glaros R, et al. Decline in perinatal HIV transmission in New York State (1997–2000). J Acquir Immune Defic Syndr 2004;36:1075–82.
11. Peters VB, Liu K-L, Robinson L-G, Dominguez KL, Abrams EJ, Gill BS, et al. Trends in perinatal HIV prevention in New York City, 1994–2003. Am J Public Health 2008;98:1857–64.
12. Charbonneau TT, Wade NA, Weiner L, Omene J, Frenkel L, Wethers JA, et al. Vertical transmission of HIV in New York State: a basis for statewide testing of newborns. AIDS Patient Care STDS 1997;11:227–36.
13. New York State Department of Health Wadsworth Center, Diagnostic HIV Laboratory, Pediatric HIV Testing Service. Available at: http://www.wadsworth.org/divisions/infdis/hiv/pedhivtest.html
. Retrieved April 28, 2009.
14. New York State Public Health Law § 2130.
15. Berkeley JS, Fogiel PC, Kindley AD, Moffat MA. Peripartum HIV seroconversion: a cautionary tale. Lancet 1992;340:58–9.
16. Goon PK, Watkins RP, Lyall EG, Parry J, Taylor GP. Third trimester screening or safer sex to prevent mother to child transmission of HIV. Sex Transm Infect 2001;77:226–7.
17. Moses SE, Tosswill J, Sudhanva M, Poulton M, Zuckerman M. HIV-1 seroconversion during pregnancy resulting in vertical transmission. J Clin Virol 2008;41:152–3.
18. Nicholson O, Michalik DE, Patel S, Larussa P, Neu N. Acute human immunodeficiency virus infection in a breast-fed infant in New York City. Pediatr Infect Dis J 2007;26:653–5.
19. Patterson KB, Leone PA, Fiscus SA, Kuruc J, McCoy SI, Wolf L, et al. Frequent detection of acute HIV infection in pregnant women. AIDS 2007;21:2303–8.
20. Nesheim S, Jamieson DJ, Danner SP, Maupin R, O'Sullivan MJ, Cohen MH, et al. Primary human immunodeficiency virus infection during pregnancy detected by repeat testing. Am J Obstet Gynecol 2007;197:149.e1–5.
21. Gray RH, Li X, Kigozi G, Serwadda D, Brahmbhatt H, Wabwire-Mangen F, et al. Increased risk of incident HIV during pregnancy in Rakai, Uganda: a prospective study. Lancet 2005;366:1182–8.
22. New York State Department of Health. Health alert: steps to further reduce mother-to-child HIV transmission in New York State. Issued February 2007. Available at: www.health.state.ny.us/diseases/aids/testing/healthalert.htm
. Retrieved February 18, 2008.
23. Centers for Disease Control and Prevention (CDC). Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep 2006;55:1–17; quiz CE1–4.
24. Sansom SL, Jamieson DJ, Farnham PG, Bulterys M, Fowler MG. Human immunodeficiency virus retesting during pregnancy: costs and effectiveness in preventing perinatal transmission. Obstet Gynecol 2003;102:782–90.
25. Guar A, Dominguez K, Kalish M, Rivera-Hernandez D, Donohoe M, Mitchell C. Practice of offering a child pre-masticated (pre-chewed) food: an unrecognized possible risk factor for HIV transmission. Programs and Abstracts Book of the 15th Conference on Retroviruses and Opportunistic Infections; February 3–6, 2008; Boston, Massachusetts.
26. Struik SS, Tudor-Williams G, Taylor GP, Portsmouth SD, Foster CJ, Walsh C, et al. Infant HIV infection despite “universal” antenatal testing. Arch Dis Child 2008;93:59–61.