Perinatal exposure to HIV among children born in Australia, 1982–2006

Children born to women with diagnosed HIV infection were notified from 1982 to a national registry of perinatal exposure to HIV through an informal network of clinicians coordinated by the Sydney Children’s Hospital, and, from May 1993, by paediatricians through the Australian Paediatric Surveillance Unit (APSU), using previously described methods.12 Cases were also notified after retrospective assessment of perinatal HIV exposure among children born to women with new diagnoses of HIV infection from July 1995.9,10

Women with HIV infection diagnosed before or at an exposed child’s birth (ie, during the hospital admission for the birth) were classified as having an antenatal diagnosis. Women with HIV infection diagnosed after a child’s birth were classified as having a postnatal diagnosis.

Children born before their mothers’ HIV diagnoses were included as cases of perinatal exposure if no exposure other than mother-to-child transmission was documented for the child and one of the following criteria was satisfied:

• the child was diagnosed with HIV infection;

• the mother was documented as having acquired HIV infection on a specified date before 1985 following receipt of HIV-infected blood or tissue;

• the mother’s HIV infection was diagnosed in an antenatal specimen or dried blood spots collected through newborn screening programs; or

• the mother had a documented history of exposure on a specified date, either before the birth or while breastfeeding.

Newly diagnosed HIV infections, including diagnoses of mother-to-child transmission and AIDS in children, are notifiable in Australia.13 Diagnoses are based on standardised laboratory tests and procedures.14 An exposed child was confirmed to be without infection if HIV antibody testing of a specimen from the child at age 18 months or older gave a negative result. From 1998, the absence of HIV infection was also diagnosed using polymerase chain reaction (PCR) testing for detection of HIV RNA.

Information sought on women whose children were perinatally exposed included the date of HIV diagnosis, the reported source of HIV exposure, mode of delivery, and mode of infant feeding. For women with an antenatal HIV diagnosis, information was also sought on use of interventions to minimise mother-to-child transmission. The available interventions were:

• before 1994, avoidance of breastfeeding alone;4

• from 1994, avoidance of breastfeeding plus antiretroviral therapy during pregnancy and delivery, and treatment of the newborn;1,2 and

• from 1999, delivery by elective caesarean section in addition to other interventions, if appropriate.3

Information sought on perinatally exposed children included the date and country of the child’s birth, and HIV infection status.

Rates of perinatal HIV exposure by state/territory and year of the child’s birth were age-standardised, using the Australian population of women aged 15–49 years who delivered a live baby in 2001 as the standard.15 Rates of perinatal HIV exposure were also age-standardised by the mother’s region of birth using the population of Australian-born women in 2001 as the standard.16 The pattern of perinatal HIV exposure was assessed by year of the child’s birth, timing of the mother’s HIV diagnosis with respect to the child’s birth, and use of interventions by women diagnosed antenatally.

Univariate (χ2 tests) and multivariate logistic regression analyses were performed to identify factors associated with mother-to-child transmission. The project was approved by the Ethics Committee of the University of New South Wales.

The age-standardised rate of perinatal HIV exposure in Australia increased from 2.3 per 100 000 live births in 1982–1986 to 8.3 per 100 000 live births in 2003–2006 (Box 1). Between 1995 and 2006, the exposure rate was stable for children of Australian-born women (0.2 per 100 000 live births). However, it increased for children of women born elsewhere in Oceania and in high-prevalence countries in South-East Asia (to 1.1 and 1.3 per 100 000 live births, respectively, in 2003–2006), and particularly for children of women born in high-prevalence countries in sub-Saharan Africa (to 9.3 per 100 000 in 1999–2002, dropping slightly to 7.4 per 100 000 in 2003–2006).

Sources of maternal HIV exposure are shown in Box 2. For women whose first exposed child was born in Australia in 1982–1986, the predominant source of exposure was receipt of HIV-infected blood or tissue before 1985 (16 of 26 women). In contrast, for those whose first exposed child was born in 2003–2006, heterosexual contact — in Australia or overseas — was the most frequent source (31 of 59 women). This included nine women who reported having a partner from a high-prevalence country, mostly in sub-Saharan Africa), and an increasing number who reported heterosexual contact in Australia with a partner with an HIV risk factor other than coming from a high-prevalence country or injecting drug use. Twenty-two women came from a high-prevalence country (12 from sub-Saharan Africa and seven from South-East Asia). Among 19 Aboriginal and Torres Strait Islander women, HIV exposure was attributed to heterosexual contact (17) or injecting drug use (2).

The mother had an antenatal HIV diagnosis in 75% of reported cases of perinatal exposure in children born in Australia 1982–2006, including 21 of the 25 children (84%) born to Indigenous mothers. The percentage of mother–child pairs with an antenatal diagnosis increased significantly over time (P < 0.001), from 11% in 1982–1986, to 66% in 1991–1998, and 98% in 2003–2006 (Box 3).

Uptake of interventions by women with an antenatal diagnosis is shown in Box 4. Avoidance of breastfeeding was the only intervention used by 52% of women (29/56) whose child was born in 1982–1994. Use of antiretroviral therapy during pregnancy plus avoidance of breastfeeding more than doubled, from 30% (11/37) in 1991–1994 to 69% (25/36) in 1995–1998, and then declined, as the use of three interventions steadily increased — from 5% (2/37) in 1991–1994, to 57% (47/82) in 2003–2006.

Mother-to-child transmission rates for women with an antenatal diagnosis, by uptake of interventions to minimise transmission are shown in Box 5. Transmission occurred in 47% of children (7/15) born to women who reported using no interventions, and 24% (9/37) whose mothers reported avoiding breastfeeding alone. Rates were much lower in children of women who used two interventions (4%; 4/106) or all three interventions (2%; 2/96). In the six cases where transmission occurred despite use of two or three interventions, the duration of antiretroviral therapy during pregnancy was 3 months or less. No cases of mother-to-child transmission occurred among 21 children born to Indigenous women diagnosed antenatally, including 15 whose mother used at least two interventions.

Among children born to women with a postnatal HIV diagnosis, the mother-to-child transmission rate remained high over the period 1982–2006, in comparison with the rate among women diagnosed antenatally (Box 3). Most of these women reported breastfeeding (62/72; 86%). Most were born in Australia (57; 66%), with smaller numbers born in high-prevalence countries (13; 15%) or another country (16; 19%).

Multivariate analysis, adjusting for year of birth, indicated that lack of use of antiretroviral therapy during pregnancy was associated with an increased risk of mother-to-child transmission (odds ratio [OR], 8.7; 95% CI, 2.9–26.2; P < 0.001). Breastfeeding was also associated with increased risk (OR, 2.3; 95% CI, 1.1–4.5; P = 0.03). When use of any antiretroviral therapy during pregnancy was excluded from the model, breastfeeding (OR, 2.29; 95% CI, 1.14–4.60; P = 0.02) and delivery by modes other than elective caesarean section (OR, 4.04; 95% CI, 1.13–14.37; P = 0.01) remained independent predictors of mother-to-child transmission, after controlling for year.

Of the 61 Australian-born children with perinatally acquired HIV infection, 26 had been diagnosed with AIDS, and 16 had died following AIDS by 31 December 2001. Most AIDS diagnoses (20; 77%) and deaths following AIDS (12; 75%) occurred before the introduction of HAART in mid 1996.