National audit of perinatal HIV infections in the UK, 2006-2013: what lessons can be learnt?

OBJECTIVES: The aim of the study was to investigate circumstances surrounding perinatal transmissions of HIV (PHIVs) in the UK. METHODS: The National Study of HIV in Pregnancy and Childhood conducts comprehensive surveillance of all pregnancies in women diagnosed with HIV infection and their infants in the UK; reports of all HIV-diagnosed children are also sought, regardless of country of birth. Children with PHIV born in 2006-2013 and reported by 2014 were included in an audit, with additional data collection via telephone interviews with clinicians involved in each case. Contributing factors for each transmission were identified, and cases described according to main likely contributing factor, by maternal diagnosis timing. RESULTS: A total of 108 PHIVs were identified. Of the 41 (38%) infants whose mothers were diagnosed before delivery, it is probable that most were infected in utero, around 20% intrapartum and 20% through breastfeeding. Timing of transmission was unknown for most children of undiagnosed mothers. For infants born to diagnosed women, the most common contributing factors for transmission were difficulties with engagement and/or antiretroviral therapy (ART) adherence in pregnancy (14 of 41) and late antenatal booking (nine of 41); for the 67 children with undiagnosed mothers, these were decline of HIV testing (28 of 67) and seroconversion (23 of 67). Adverse social circumstances around the time of pregnancy were reported for 53% of women, including uncertain immigration status, housing problems and intimate partner violence. Eight children died, all born to undiagnosed mothers. CONCLUSIONS: Priority areas requiring improvement include reducing incident infections, improving ART adherence and facilitating better engagement in care, with attention to addressing the health inequalities and adverse social situations faced by these women.

Presentation for care and antenatal management of HIV in the UK, 2009-2014.

OBJECTIVES: Despite very low rates of vertical transmission of HIV in the UK overall, rates are higher among women starting antenatal antiretroviral therapy (ART) late. We investigated the timing of key elements of the care of HIV-positive pregnant women [antenatal care booking, HIV laboratory assessment (CD4 count and HIV viral load) and antenatal ART initiation], to assess whether clinical practice is changing in line with recommendations, and to investigate factors associated with delayed care. METHODS: We used the UK's National Study of HIV in Pregnancy and Childhood for 2009-2014. Data were analysed by fitting logistic regression and Cox proportional hazards models. RESULTS: A total of 5693 births were reported; 79.5% were in women diagnosed with HIV prior to that pregnancy. Median gestation at antenatal booking was 12.1 weeks [interquartile range (IQR) 10.0-15.6 weeks] and booking was significantly earlier during 2012-2014 vs. 2009-2011 (P < 0.001), although only in previously diagnosed women. Overall, 42.2% of pregnancies were booked late (≥ 13 gestational weeks). Among women not already on treatment, antenatal ART commenced at a median of 21.4 (IQR18.1-24.5) weeks and started significantly earlier in the most recent time period (P < 0.001). Compared with previously diagnosed women, those newly diagnosed during the current pregnancy booked later for antenatal care and started antenatal ART later (both P < 0.001). Multivariable analyses revealed demographic variations in access to or uptake of care, with groups including migrants and parous women initiating care later. CONCLUSIONS: Although women are accessing antenatal and HIV care earlier in pregnancy, some continue to face barriers to timely initiation of antenatal care and ART.

Syphilis screening in pregnancy in the United Kingdom, 2010-2011: a national surveillance study.

OBJECTIVE: To evaluate the national antenatal syphilis screening programme and provide evidence for improving screening and management strategies. DESIGN: National population-based surveillance. SETTING: United Kingdom (UK). POPULATION: All pregnant women screening positive for syphilis, 2010-2011. METHODS: Demographic, laboratory and treatment details for each pregnancy were collected from UK antenatal units (~210), along with follow-up information on all infants born to women requiring syphilis treatment in pregnancy. MAIN OUTCOME MEASURES: Proportion of women with newly or previously diagnosed syphilis among those with positive screening tests in pregnancy; proportion requiring treatment. RESULTS: Overall, 77% (1425/1840) of reported pregnancies were confirmed syphilis screen-positive. Of these, 71% (1010/1425) were in women with previously diagnosed syphilis (155 requiring treatment), 26% (374/1425) with newly diagnosed syphilis (all requiring treatment) and 3% (41/1425) required treatment but the reason for treatment was unclear. Thus 40% (570/1425) required treatment overall; of these, 96% (516/537) were treated (missing data: 33/570), although for 18% (83/456), this was not until the third trimester (missing data: 60/537). Follow up of infants born to treated women was poor, with at least a third not followed. Six infants were diagnosed with congenital syphilis; two mothers were untreated, three had delayed treatment and one had incomplete treatment (first trimester). CONCLUSION: Over 2 years, among pregnant women with confirmed positive syphilis screening results in the UK, a quarter had newly diagnosed infections and 40% required treatment. Despite high uptake of treatment, antenatal syphilis management could be improved by earlier detection, earlier treatment, and stronger links between healthcare teams. TWEETABLE ABSTRACT: 25% of pregnant women screening positive for syphilis in the UK were newly diagnosed and 40% needed treatment.

The incidence of congenital syphilis in the United Kingdom: February 2010 to January 2015.

OBJECTIVE: To estimate the incidence of congenital syphilis in the UK. DESIGN: Prospective study. SETTING AND POPULATION: United Kingdom. METHODS: Children born between February 2010 and January 2015 with a suspected diagnosis of congenital syphilis were reported through an active surveillance system. MAIN OUTCOME MEASURES: Number of congenital syphilis cases and incidence. RESULTS: For all years, reported incidence was below the WHO threshold for elimination (<0.5/1000 live births). Seventeen cases (12 male, five female) were identified. About 50% of infants (8/17) were born preterm (<37 weeks' gestation): median birthweight 2000 g (865-3170 g). Clinical presentation varied from asymptomatic to acute disease, including severe anaemia, hepatosplenomegaly, rhinitis, thrombocytopaenia, skeletal damage, and neurosyphilis. One infant was deaf and blind. Median maternal age was 20 years (17-31) at delivery. Where maternal stage of infection was recorded, 6/10 had primary, 3/10 secondary and 1/10 early latent syphilis. Most mothers were white (13/16). Country of birth was recorded for 12 mothers: UK (n = 6), Eastern Europe (n = 3), Middle East (n = 1), and South East Asia (n = 2). The social circumstances of mothers varied and included drug use and sex work. Some experienced difficulty accessing health care. CONCLUSION: The incidence of congenital syphilis is controlled and monitored by healthcare services and related surveillance systems, and is now below the WHO elimination threshold. However, reducing the public health impact of this preventable disease in the UK is highly dependent on the successful implementation of WHO elimination standards across Europe. TWEETABLE ABSTRACT: Congenital syphilis incidence in the UK is at a very low level and well below the WHO elimination threshold.

Duration of ruptured membranes and mother-to-child HIV transmission: a prospective population-based surveillance study.

OBJECTIVE: To investigate the association between duration of rupture of membranes (ROM) and mother-to-child HIV transmission (MTCT) rates in the era of combination antiretroviral therapy (cART). DESIGN: The National Study of HIV in Pregnancy and Childhood (NSHPC) undertakes comprehensive population-based surveillance of HIV in pregnant women and children. SETTING: UK and Ireland. POPULATION: A cohort of 2398 singleton pregnancies delivered vaginally, or by emergency caesarean section, in women on cART in pregnancy during the period 2007-2012 with information on duration of ROM; HIV infection status was available for 1898 infants. METHODS: Descriptive analysis of NSHPC data. MAIN OUTCOME MEASURES: Rates of MTCT. RESULTS: In 2116 pregnancies delivered at term, the median duration of ROM was 3 hours 30 minutes (interquartile range, IQR 1-8 hours). The overall MTCT rate for women delivering at term with duration of ROM ≥4 hours was 0.64% compared with 0.34% for ROM <4 hours, with no significant difference between the groups (OR 1.90, 95% CI 0.45-7.97). In women delivering at term with a viral load of <50 copies/ml, there was no evidence of a difference in MTCT rates with duration of ROM ≥4 hours, compared with <4 hours (0.14% for ≥4 hours versus 0.12% for <4 hour; OR 1.14, 95% CI 0.07-18.27). Among infants born preterm with infection status available, there were no transmissions in 163 deliveries where the maternal viral load was <50 copies/ml. CONCLUSIONS: No association was found between duration of ROM and MTCT in women taking cART. TWEETABLE ABSTRACT: Rupture of membranes of more than 4 hours is not associated with MTCT of HIV in women on effective ART delivering at term.

British HIV Association and Children's HIV Association position statement on infant feeding in the UK 2011.

To prevent the transmission of HIV infection during the postpartum period, the British HIV Association and Children's HIV Association (BHIVA/CHIVA) continue to recommend the complete avoidance of breast feeding for infants born to HIV-infected mothers, regardless of maternal disease status, viral load or treatment.

Antiretroviral therapy and preterm delivery-a pooled analysis of data from the United States and Europe.

OBJECTIVE: To investigate reported differences in the association between highly active antiretroviral therapy (HAART) in pregnancy and the risk of preterm delivery among HIV-infected women. DESIGN: Combined analysis of data from three observational studies. SETTING: USA and Europe. POPULATION: A total of 19, 585 singleton infants born to HIV-infected women, 1990-2006. METHODS: Data from the Pediatric Spectrum of HIV Disease project (PSD), a US monitoring study, the European Collaborative Study (ECS), a consented cohort study, and the National Study of HIV in Pregnancy and Childhood (NSHPC), the United Kingdom and Ireland surveillance study. MAIN OUTCOME MEASURE: Preterm delivery rate (<37 weeks of gestation). RESULTS: Compared with monotherapy, HAART was associated with increased preterm delivery risk in the ECS (adjusted odds ratio [AOR] 2.40, 95% CI 1.49-3.86) and NSHPC (AOR 1.43, 95% CI 1.10-1.86), but not in the PSD (AOR 0.92, 95% CI 0.67-1.26), after adjusting for relevant covariates. Because of heterogeneity, data were not pooled for this comparison, but heterogeneity disappeared when HAART was compared with dual therapy (P = 0.26). In a pooled analysis, HAART was associated with 1.5-fold increased odds of preterm delivery compared with dual therapy (95% CI 1.19-1.87, P=0.001), after adjusting for covariates. CONCLUSIONS: Heterogeneity in the association between HAART and preterm delivery was not explained by study design, adjustment for confounders or a standard analytical approach, but may have been the result of substantial differences in populations and data collected. The pooled analysis comparing HAART with dual therapy showed an increased risk of preterm delivery associated with HAART.

Vertically acquired HIV diagnosed in adolescence and early adulthood in the United Kingdom and Ireland: findings from national surveillance.

OBJECTIVE: The aim of the study was to describe the characteristics of young people with vertically acquired HIV diagnosed aged > or =13 years. METHODS: A retrospective review of HIV diagnoses reported to well-established national paediatric and adult HIV surveillance systems in the United Kingdom/Ireland was conducted. RESULTS: Forty-two young people with vertically acquired HIV diagnosed aged > or =13 years were identified; 23 (55%) were female, 40 (95%) were black African and 36 (86%) were born in sub-Saharan Africa. The median age at HIV diagnosis was 14 years (range, 13-20 years). Half of the patients presented with symptoms; the remainder were screened for HIV following diagnosis of a relative. The median CD4 count at diagnosis was 210 cells/microL (range, 0-689 cells/microL), 12 patients (29%) were diagnosed with AIDS at HIV diagnosis or subsequently, and 34 (81%) started combination antiretroviral therapy (ART), most (31 of 34) within a year of diagnosis. CONCLUSION: A small number of young people with vertically acquired HIV survive childhood without ART and are diagnosed at age > or =13 years in the United Kingdom/Ireland. Half of the patients were asymptomatic, highlighting the importance of considering HIV testing for all offspring of HIV-infected women, regardless of age or symptoms. Increased awareness among clinicians and parents is required to reduce delayed presentation with advanced disease and to avoid onward transmission as these young people become sexually active.

Rubella seroprevalence in pregnant women in North Thames: estimates based on newborn screening samples.

OBJECTIVES: Routine screening for rubella susceptibility is recommended in the UK so that women found to be susceptible can be offered immunization in the post partum period. We demonstrate the use of newborn dried blood spot samples linked to routine vital statistics datasets to monitor rubella susceptibility in pregnant women and to investigate maternal characteristics as determinants of rubella seronegativity. SETTING: North Thames region of England (including large parts of inner London). METHODS: Maternally acquired rubella IgG antibody levels were measured in 18882 newborn screening blood spot samples. Latent class regression finite mixture models were used to classify samples as seronegative to rubella. Data on maternal country of birth were available through linkage to birth registration data. RESULTS: An estimated 2.7% (95% CI 2.4%-3.0%) of newly delivered women in North Thames were found to be seronegative. Mothers born abroad, particularly in Sub-Saharan Africa and South Asia, were more likely to be seronegative than UK-born mothers, with adjusted odds ratios of 4.2 (95% CI 3.1-5.6) and 5.0 (3.8-6.5), respectively. Mothers under 20 years were more likely to be seronegative than those aged 30 to 34. CONCLUSION: Our findings highlight the need for vaccination to be targeted specifically at migrant women and their families to ensure that they are protected from rubella in pregnancy and its serious consequences.

Trends in management and outcome of pregnancies in HIV-infected women in the UK and Ireland, 1990-2006.

OBJECTIVE: To describe the changing demographic profile of diagnosed HIV-infected pregnant women over time and trends in pregnancy outcome, uptake of interventions and mother-to-child transmission. DESIGN: National surveillance study. SETTING: UK and Ireland. POPULATION: Diagnosed HIV-infected pregnant women, 1990-2006. METHODS: Active surveillance of obstetric and paediatric HIV conducted through the National Study of HIV in Pregnancy and Childhood. MAIN OUTCOME MEASURES: Maternal characteristics, pregnancy outcome, use of antiretroviral therapy, mode of delivery and mother-to-child transmission. RESULTS: A total of 8327 pregnancies were reported, increasing from 82 in 1990 to 1394 in 2006, with an increasing proportion from areas outside London. Injecting drug use as the reported risk factor for maternal HIV acquisition declined from 49.2% (185/376) in 1990-1993 to 3.1% (125/4009) in 2004-2006 (P < 0.001), while the proportion of women born in sub-Saharan Africa increased from 43.5% (93/214) in 1990-1993 to 78.6% (3076/3912) in 2004-2006 (P < 0.004). Reported pregnancy terminations decreased from 29.6% (111/376) in 1990-1993 to 3.4% (135/4009) in 2004-2006 (P < 0.001). Most (56.4%, 3717/6593) deliveries were by elective caesarean section, with rates highest in 1999 (66.4%, 144/217). Vaginal deliveries increased from 16.6% (36/217) in 1999 to 28.3% (321/1136) in 2006 (P < 0.001). Use of antiretroviral therapy in pregnancy increased over time, reaching 98.4% (1092/1110) in 2006, and the overall mother-to-child transmission rate declined from 18.5% (35/189) in 1990-1993 to 1.0% (29/2832) in 2004-2006. CONCLUSIONS: The annual number of reported pregnancies increased dramatically between 1990 and 2006, with changing demographic and geographic profiles and substantial changes in pregnancy management and outcome.

HIV-1 subtypes in pregnant women in the UK.

We analyse the distribution of HIV-1 subtypes in HIV-1-seropositive samples from 333,270 residual neonatal dried blood spot samples tested for routine newborn screening tests in the UK between July 1999 and December 2002. Of the 813 antibody-positive samples shown to contain passively acquired, maternal HIV-1 for which subtyping was attempted, 333 (41%) could not be subtyped due to cross-reactivity or low values of the assay results, and 480 (59%) were classified as B (35, 7.3%) or non-B (445, 92.7%). The proportions of subtyped B samples differed significantly (P=0.004) between those from neonates whose mothers were born in the UK (21.4%) and those from neonates whose mothers were known to be born abroad (7%). Using a serological approach to establish viral serotype, we document the distribution of HIV-1 subtypes in infected pregnant women in the UK.

Pneumocystis carinii pneumonia and cytomegalovirus infection in children with vertically acquired HIV infection.

OBJECTIVES: The outcome of Pneumocystis carinii pneumonia (PCP) in HIV-infected infants is poor, and the role of cytomegalovirus (CMV) co-infection in the course and outcome of PCP is unclear. This study describes the prevalence, clinical characteristics, management and changes in survival over time of vertically HIV-infected infants developing PCP and/or CMV infection. METHODS: Data on children with HIV, born in the UK and Ireland and reported to the National Study of HIV in Pregnancy and Childhood, with PCP and/or CMV were combined with clinical information collected from reporting paediatricians. RESULTS: By April 1998, 340 vertically HIV-infected children had been reported, of whom 93 had PCP and/or CMV, as their first AIDS indicator disease; 85 (91%) were infants. Among infants with PCP, 79% were born to mothers not diagnosed as HIV infected, and there was an independent and statistically significant association with breast-feeding, being black African, and developing CMV disease. Median survival after PCP and/or CMV was significantly better in those born between 1993 and 1998 compared with those born before 1993 (P = 0.009), and worse than after other AIDS diagnoses (P = 0.01). Infants with dual infection were more likely to be ventilated (P = 0.003) and receive corticosteroids (P = 0.002) than those with PCP alone. CONCLUSION: Although survival from PCP and CMV has improved over time, these remain serious and potentially fatal infections among infants in whom maternal HIV status is not recognized in pregnancy. Breast-feeding increases the risk of combined PCP and CMV infection, which is associated with severe disease.

Uptake of interventions to reduce mother-to-child transmission of HIV in the United Kingdom and Ireland.

OBJECTIVES: To describe the uptake of interventions to reduce mother-to-child transmission of HIV infection. DESIGN: Voluntary confidential reporting of HIV infection in pregnancy and childhood; telephone interview with key professionals in all London maternity units. SUBJECTS AND SETTING: HIV-infected pregnant women and children in the United Kingdom and Ireland. MAIN OUTCOME MEASURES: Trends in breastfeeding, use of zidovudine, mode of delivery and terminations of pregnancy. RESULTS: Between 1990 and 1995, 14 (4%) out of 314 women diagnosed with HIV infection before delivery breastfed compared with 109 (77%) out of 142 diagnosed after delivery. Since 1994, zidovudine use has increased in each 6-month period (14, 39, 67, and 75%; chi 2 = 17.5, P < 0.001), although in 1995 it was the policy of only 48% of London maternity units to offer zidovudine to HIV-infected women. During 1995, 44% of HIV-infected women were delivered by elective Cesarean section. Since 1990, 20% of women first diagnosed in pregnancy were reported to have their pregnancy terminated. CONCLUSIONS: Although detection of previously undiagnosed HIV infection in pregnancy remains low in the United Kingdom, and particularly in London, HIV-infected pregnant women who are aware of their status are increasingly active in taking up interventions to reduce transmission to their infants. If all HIV-infected women attending for antenatal care in London consented to testing and took up interventions and termination of pregnancy at the rates observed in this study, the number of vertically infected babies born in London each year could be reduced from an estimated 41 to 13.

Performance of antenatal HIV screening strategies in the United Kingdom.

BACKGROUND: In 1996 only 13.5% of previously undiagnosed HIV infected women were detected in pregnancy. In this study, all 265 maternity units in the United Kingdom were surveyed to determine the relation between screening strategy, uptake of testing, and detection rate. METHODS: Data on HIV screening strategy and uptake of testing were collected in 1997 by postal questionnaire. The proportion of women with previously undiagnosed HIV infection identified during pregnancy (detection rate) was calculated using data from national obstetric HIV surveillance and unlinked anonymous seroprevalence studies. RESULTS: 239 (90%) units responded; 25 of these (10%) had a universal offer strategy, 89 (37%) a selective offer, and 125 (52%) tested only women who requested it. All selective units offered testing to injecting drug users, but only 26% to women who had lived abroad in high prevalence areas. Uptake was over 10% in only eight units, all with a universal strategy, and in 76% of selective units it was below 0.1%. The detection rate was 14.7% in universal units, 7.8% in selective units, and 7.7% in on request units. In universal units, detection increased by 6.3% (95% confidence interval 3.7% to 8.8%) for every 10% increase in uptake of testing. There was evidence of both selective presentation for testing and avoidance of testing among infected women. CONCLUSIONS: All current antenatal HIV testing strategies fail to identify most infected women. Universal offer strategies achieve a very low uptake and a poor detection rate. Units with selective strategies tend to test only a minority of women at high risk and do not target all the main high risk groups.

Perinatal mortality and morbidity among babies delivered in water: surveillance study and postal survey.

AIM: To compare perinatal morbidity and mortality for babies delivered in water with rates for babies delivered conventionally (not in water). DESIGN: Surveillance study (of all consultant paediatricians) and postal survey (of all NHS maternity units). SETTING: British Isles (surveillance study); England and Wales (postal survey). SUBJECTS: Babies born in the British Isles between April 1994 and March 1996 who died perinatally or were admitted for special care within 48 hours of birth after delivery in water or after labour in water followed by conventional delivery (surveillance study); babies delivered in water in England and Wales in the same period (postal survey). MAIN OUTCOME MEASURESE Number of deliveries in water in the British Isles that resulted in perinatal death or in admission to special care within 48 hours of birth; and proportions (of such deliveries) of all water births in England and Wales. RESULTS: 4032 deliveries (0.6% of all deliveries) in England and Wales occurred in water. Perinatal mortality was 1.2/1000 (95% confidence interval 0.4 to 2.9) live births; 8.4/1000 (5.8 to 11.8) live births were admitted for special care. No deaths were directly attributable to delivery in water, but 2 admissions were for water aspiration. UK reports of mortality and special care admission rates for babies of women considered to be at low risk of complications during delivery who delivered conventionally ranged from 0.8/1000 (0. 2 to 4.2) to 4.6/1000 (0.1 to 25) live births and from 9.2 (1.1 to 33) to 64/1000 (58 to 70) live births respectively. Compared with regional data for low risk, spontaneous, normal vaginal deliveries at term, the relative risk for perinatal mortality associated with delivery in water was 0.9 (99% confidence interval 0.2 to 3.6). CONCLUSIONS: Perinatal mortality is not substantially higher among babies delivered in water than among those born to low risk women who delivered conventionally. The data are compatible with a small increase or decrease in perinatal mortality for babies delivered in water.

Vertical transmission rates for HIV in the British Isles: estimates based on surveillance data.

OBJECTIVE: To estimate and interpret time trends in vertical transmission rates for HIV using data from national obstetric and paediatric surveillance registers. DESIGN: Prospective study of HIV infected women reported through obstetric surveillance. HIV infection status of the child and onset of AIDS were reported through paediatric surveillance. Rates of vertical transmission and progression to AIDS rate were estimated by methods that take account of incomplete follow up of children with indeterminate infection status and delay in AIDS reporting. SETTING: British Isles. SUBJECTS: Pregnant women infected with HIV whose infection was diagnosed before delivery, and their babies. MAIN OUTCOME MEASURES: Mother to child transmission of infection and progression to AIDS in children. RESULTS: By January 1999, 800 children born to diagnosed HIV infected women who had not breast fed had been reported. Vertical transmission rates rose to 19.6% (95% confidence interval 8. 0% to 32.5%) in 1993 before falling to 2.2% (0% to 7.8%) in 1998. Between 1995 and 1998 use of antiretroviral treatment increased significantly each year, reaching 97% of live births in 1998. The rate of elective caesarean section remained constant, at around 40%, up to 1997 but increased to 62% in 1998. Caesarean section and antiretroviral treatment together were estimated to reduce risk of transmission from 31.6% (13.6% to 52.2%) to 4.2% (0.8% to 8.5%). The proportion of infected children developing AIDS in the first 6 months fell from 17.7% (6.8% to 30.8%) before 1994 to 7.2% (0% to 15. 7%) after, coinciding with increased use of prophylaxis against Pneumocystis carinii pneumonia. CONCLUSIONS: In the British Isles both HIV related morbidity and vertical transmission are being reduced through increased use of interventions.

Factors affecting uptake of antenatal HIV testing in London: results of a multicentre study.

OBJECTIVES: To measure the uptake of antenatal HIV testing and determine its relation to risk of HIV and to screening practices. DESIGN: Multicentre prospective questionnaire study. SUBJECTS: Pregnant women attending six maternity units. SETTING: Inner London, 1995-6. MAIN OUTCOME MEASURES: Uptake of testing by risk factors for HIV, ethnicity, and factors about the antenatal interview. RESULTS: All units had a "universal offer" policy for HIV testing. In five units forms were completed for 18,791 (88%) of 21,247 pregnant women. The sixth unit, where the response rate was too low to assess uptake, was excluded from the analysis. Uptake ranged from 3.4% to 51.2% (overall 22.9%), in parallel with detection of previously undiagnosed infection in pregnant women (4.9-60%). Controlling for unit, uptake was higher among the 7% who disclosed risk factors. Among those at low risk, uptake varied by ethnic group (South Asian women 9%; Latin American and Mediterranean women 33%). The relation between uptake and HIV risk category varied greatly across units. Despite increased HIV seroprevalence in black African women, uptake was similar in this group to that among women at low risk (24%). Uptake increased 2.1-fold if HIV transmission was discussed. Midwives reported spending 7 (2-15) minutes discussing HIV issues. CONCLUSIONS: Uptake of HIV testing was unacceptably low in all units, with maternity unit the strongest predictor. New approaches to antenatal HIV testing are urgently required and uptake should be audited routinely.