In the multivariable analysis (n = 4892; Table 4) including ART, mode of delivery, gestational age and sex (but not viral load), being untreated (adjusted odd ratio (AOR) = 9.08, P < 0.001) was the strongest risk factor for transmission, and girls were more likely to be infected than boys (AOR = 1.91, P = 0.023). Prematurity (less than 32 weeks) was a significant risk factor for transmission (AOR = 3.55, P = 0.010); however, of the seven women who had infected infants born at less than 32 weeks gestation, all were either untreated (n = 3) or treated for less than 3 weeks (range: 1–19 days), and all but one delivered vaginally. Prematurity was not a significant risk factor for transmission among women on HAART (AOR = 1.32, 95% CI: 0.30–5.82, P = 0.714).
In the multivariable model, vaginal delivery was associated with a nonsignificant 1.8-fold increased risk of transmission compared with elective Caesarean section (AOR = 1.82, P = 0.076). Information on whether vaginal deliveries were planned or unplanned was available for 69.1% (775/1122). After adjusting for ART, gestational age and sex, unplanned vaginal delivery was associated with a significantly increased risk of transmission (AOR = 4.16, 95% CI: 1.66–10.41, P = 0.002) compared with elective Caesarean section, but planned vaginal delivery was not (AOR = 1.56, 95% CI: 0.65–3.72, P = 0.319).
The association between mode of delivery and transmission varied by type of therapy, but owing to small numbers could not be explored adequately using interaction terms; analyses stratified by therapy are presented later.
Viral load was reported for 79.5% (4096/5151) of women, but for significantly fewer untreated than treated women (49.7%, 71/143, versus 82.3%, 4021/4884, P < 0.001), and infected than uninfected infants (67.2%, 41/61, versus 79.7%, 4055/5090, P = 0.024). The closest viral load to delivery was a median 23 days before delivery (IQR: 10–44 days). In the multivariable analysis (n = 4084) controlling for ART, mode of delivery, gestational age and sex, each log10 increase in viral load was associated with a 2.4-fold increased risk of transmission (AOR = 2.41, P < 0.001) (Table 4). In this model, lack of ART (AOR = 3.17, P = 0.023) and vaginal delivery (AOR = 2.40, P = 0.033) were strongly associated with transmission, but gestational age and sex were not.
Although HIV-infected women in the UK and Ireland are recommended to formula feed, breastfeeding was reported in 0.6% (29/4399) of infants with information provided; three were infected, all born to untreated women (two declined ART, one was diagnosed close to delivery).
Transmission following highly active antiretroviral therapy
MTCT rate for women on HAART was 1.0% (40/4120) (Table 2) and was not significantly different when HAART included an NNRTI (0.9%, 18/1959), a protease inhibitor (1.1%, 20/1795, P = 0.625), both an NNRTI and a protease inhibitor (0.8%, 2/258), or neither (0%, 0/108, P = 0.847). Comparing protease inhibitor with NNRTI regimens, there was no difference in MTCT rates after adjusting for mode of delivery, sex and viral load (AOR = 1.31, 95% CI: 0.62–2.76, P = 0.482) (as previously noted, gestational age was not a significant risk factor for transmission in women on HAART and was therefore excluded from these analyses). HAART at conception was associated with a lower risk of transmission than HAART started in pregnancy (0.1%, 1/928, versus 1.3%, 39/2967, P = 0.001), but this was only of borderline significance after adjusting for mode of delivery, sex and viral load (AOR = 0.18, 95% CI: 0.02–1.33, P = 0.093). Among women who started HAART during pregnancy, those who transmitted (n = 34) started later than those who did not (median gestational age at initiation: 30.1 weeks, IQR: 27.4–32.6 weeks, versus 25.9 weeks, IQR: 22.4–28.7 weeks, P < 0.001). Each additional week of treatment corresponded to a 10% (AOR = 0.90, 95% CI: 0.84–0.97, P = 0.007) reduction in the risk of transmission after adjusting for viral load (closest to delivery), mode of delivery and sex.
Among women on HAART, there was no statistically significant difference in MTCT rates between elective Caesarean section (0.7%, 17/2286) and planned vaginal delivery (0.7%, 4/559; AOR = 1.24, 95% CI: 0.34–4.52, P = 0.746, adjusted for sex and viral load); viral load was undetectable in 58.7% (1135/1934) of those having elective Caesarean sections, and 79.0% (417/528) having planned vaginal deliveries (P < 0.001). Only three transmissions were reported among 2117 infants born to women on HAART with undetectable viral load (0.1%, 95% CI: 0.0–0.4%); two were born by elective Caesarean section (0.2%, 2/1135) and one by planned vaginal delivery (0.2%, 1/417). Two of the three infants (one born vaginally) had positive PCR results within 72 h of birth, suggesting possible in-utero transmission. MTCT rates were higher in women on HAART who had emergency Caesarean sections (1.7%, 15/877) or unplanned vaginal deliveries (3.3%, 4/122) compared with elective Caesarean sections (P = 0.027 and P = 0.019, respectively).
Eighteen other women had infected infants despite HAART and either planned vaginal delivery or elective Caesarean section. For 10 of them, one or more of the following applied: short duration of treatment, high viral load near delivery (range: 8500–285 000), adherence problems. There were six transmissions from women with low but detectable viral load (≥50 and <1000 copies/ml); two had planned vaginal deliveries (2.5%, 2/81) and four elective Caesarean sections (0.8%, 4/471, P = 0.215). Two of these six infected infants had positive PCR results within 72 h of birth (both born by elective Caesarean section). The remaining two women received HAART for at least 1 month, but no viral loads were reported.
Information on neonatal prophylaxis was provided for 20 of the 21 infants infected despite maternal HAART and planned vaginal or elective Caesarean section delivery, and all were treated; 18 had breastfeeding status reported, and none was breastfed.
Transmission following zidovudine monotherapy
The mothers of 638 infants received prophylactic zidovudine monotherapy in pregnancy, and three infants were infected (0.5%, 95% CI: 0.1–1.4%); all three women were treated for less than 1 month, had detectable viral load (range: 474–3000 copies/ml) and delivered vaginally. The transmission rate following monotherapy and elective Caesarean section was 0% (0/464, 95% CI: 0–0.8%), with median viral load of 400 copies/ml (IQR: 61–1992 copies/ml); this transmission rate was not significantly different from that following HAART and planned vaginal delivery (0.7%) or elective Caesarean section (0.7%, P = 0.150). The transmission rate following monotherapy and emergency Caesarean section was 0% (0/116, 95% CI: 0–3.4%), with median viral load of 597 copies/ml (IQR: 84–3195 copies/ml).
Adjustment for unreported infection status
Potential bias introduced by excluding infants with unreported infection status was investigated by computing likely infection status based on maternal treatment and viral load categories. A crude transmission rate of 9.1% was assumed for all untreated women (Table 2) and those with missing treatment information (viral load was unavailable for over 70% of these women). Among treated women, actual transmission rates by viral load levels were 0.1% (3/2299) for viral load of less than 50 copies/ml, 1.2% (12/1003) for 50–999 copies/ml, 1.4% (6/414) for 1000–9999 copies/ml, 4.6% (14/305) for at least 10 000 copies/ml and 1.0% (9/863) for those where viral load was not recorded. Using these transmission rates to compute likely infection status, an estimated 1.8% (14/779) of children with unreported infection status would be infected, which would increase the overall transmission rate marginally to 1.3% (75/5930, 95% CI: 1.0–1.6%).
These results are based on observational data collected in the UK and Ireland. Routine antenatal HIV testing and professional guidelines  were available throughout this period, outlining a range of strategies for the management of HIV in pregnancy. Although these guidelines recommend that most women should receive HAART in pregnancy, zidovudine monotherapy with planned Caesarean section remains an alternative for asymptomatic women with high CD4 cell count and low viral load. A more tolerant approach to mode of delivery for women who achieve full viral suppression on HAART has become evident, with rates of planned vaginal delivery rising .
MTCT rates remained very low: 1.2% overall (slightly lower in recent years) and 0.8% for women on treatment for at least the last 2 weeks of pregnancy. These low rates are consistent with reports from elsewhere in Europe and the United States [3,5]. We observed similar transmission rates among women on HAART who had elective Caesarean sections or planned vaginal deliveries (0.7% in both groups), and the transmission rate was particularly low (0.1%) in the 2100 women on HAART who achieved viral suppression. In cases where transmission occurred despite HAART and planned vaginal or elective Caesarean section delivery, most infections could be explained by failure to reduce viral load, mainly because of short duration of treatment or adherence problems, or to in-utero transmission. An audit undertaken to investigate circumstances surrounding cases of perinatal HIV transmission in England reported similar findings .
In this population, zidovudine monotherapy is likely to have been used selectively according to the BHIVA guidelines , a strategy associated with low rates of drug resistance [18,19]. Our findings suggest that selective zidovudine monotherapy combined with elective Caesarean section is a reasonable approach for preventing MTCT (no infections reported in over 450 infants).
Lack of ART and vaginal delivery, particularly unplanned, remained significantly associated with transmission after controlling for viral load near delivery. Before the widespread use of HAART, prematurity was identified as a risk factor for transmission . In this analysis, the increased risk of transmission associated with very premature delivery was not observed in women on HAART. An increased risk of MTCT for girls has been previously reported, both for HIV and for hepatitis C virus [21–23], but in this study, the increased rate for girls was no longer significant when the analysis was restricted to cases where maternal viral load was recorded, possibly owing to reduced numbers.
Among women on HAART, we found no difference in transmission rates by type of regimen, even though differences in response to protease inhibitor and NNRTI-based regimens in pregnant women have been reported . Any effect, however, might have been attenuated in recent years by the restriction of nevirapine to women with CD4 cell counts below 250 cells/μl [6,25], who may also have higher baseline viral loads.
There was evidence that being on HAART at conception and starting HAART earlier in pregnancy were associated with a lower risk of transmission after adjusting for viral load; this could be due to an increased risk of in-utero transmission before initiation of treatment. In these analyses we used viral loads closest to delivery, which might not reflect the infant's exposure to high maternal viral load at earlier stages of pregnancy. Of the nine infected children born to women with low or undetectable viral load, four had detectable virus at birth and were probably infected during fetal life. With postpartum transmission in developed countries now rare and intra-partum transmission highly preventable, in-utero transmission is likely to account for an increasing proportion of perinatal infections. The benefits of early treatment, however, must be weighed against concerns about drug toxicities and adverse pregnancy outcomes [26–28]; we have previously reported an increased risk of premature delivery associated with HAART in this population .
This national surveillance study is designed to include all pregnant women diagnosed with HIV any time up to delivery, regardless of maternal characteristics or uptake of interventions. Alignment of NSHPC reports with unlinked anonymous neonatal seroprevalence data suggests that about 95% of HIV-infected women are diagnosed before delivery . Given the active nature of the complementary obstetric and paediatric reporting systems, and the high estimated overall detection rates, very few cases of diagnosed HIV infection in pregnancy are likely to remain unreported.
Children whose infection status had not yet been reported were more likely than those with known infection status to have recognized risk factors for transmission. We have, however, shown that although this bias could potentially lead to an underestimate of the overall transmission rate any effect is likely to be small. Owing to the observational nature of these data they should be interpreted in the context of current guidelines, in which interventions are recommended on the basis of timing of maternal diagnosis, clinical presentation, response to treatment, personal circumstances and choice. Since a randomized trial would be inappropriate, these population data provide important evidence to support the targeted use of different combinations of interventions for preventing MTCT.
The low rate (1.2%) of MTCT of HIV among diagnosed pregnant women in the UK and Ireland is a remarkable achievement. Continuing to improve the offer and uptake of antenatal HIV testing could have a significant impact on further reducing MTCT, since most perinatally acquired infection is now in infants whose mothers are among the approximately 5% of infected women who remain undiagnosed at delivery [12,17]. Ensuring that women are diagnosed in time to take up appropriate interventions remains a priority, and early testing for all pregnant women should continue to be promoted. Our findings suggest that offering HIV-infected women choices about HIV treatment and mode of delivery, according to current guidelines, has led to very low rates of MTCT.
National surveillance of obstetric and paediatric HIV is undertaken through the National Study of HIV in Pregnancy and Childhood (NSHPC) in collaboration with the Health Protection Agency Centre for Infections, and Health Protection Scotland. We gratefully acknowledge the contribution of the midwives, obstetricians, genito-urinary physicians, paediatricians, clinical nurse specialists and all other colleagues who report to the NSHPC through the British Paediatric Surveillance Unit of the Royal College of Paediatrics and Child Health, and the obstetric reporting scheme run under the auspices of the Royal College of Obstetricians and Gynaecologists. We wish to thank Janet Masters who co-ordinates the study and manages the data, Icina Shakes for administrative support, and Marie-Louise Newell, Claire Thorne, Claire Hankin and Barbara Willey for their helpful comments on drafts of this paper.
Sponsorship: The National Study of HIV in Pregnancy and Childhood was previously funded by AVERT and the Department of Health, and is currently funded by the Health Protection Agency (grant number GHP/003/013/003). C.L. Townsend is funded by the UK Medical Research Council (MRC) (grant number G0501895). This work was undertaken at the Centre for Paediatric Epidemiology and Biostatistics which benefits from funding support from the MRC in its capacity as the MRC Centre of Epidemiology for Child Health. The UCL Institute of Child Health receives a proportion of funding from the Department of Health's National Institute for Health Research Biomedical Research Centres funding scheme.
The views expressed in this paper are those of the authors alone.
Ethics approval for the NSHPC was renewed following review by the London Multi-Centre Research Ethics Committee in 2004 (ref. MREC/04/2/009).
C.L.T. and P.A.T. participated in the data collection and drafted the paper. C.L.T. carried out the statistical analyses with support from M.C.-B. All authors participated in developing the concept of the paper and interpreting the results. All authors commented on drafts of the paper and approved the final version. P.A.T. is responsible for the NSHPC and is the guarantor.
There are no conflicts of interest.
1. Duong T, Ades AE, Gibb DM, Tookey PA, Masters J. Vertical transmission rates for HIV
in the British Isles: estimates based on surveillance data. BMJ 1999; 319:1227–1229.
2. European Collaborative Study. HIV-infected pregnant women and vertical transmission in Europe since 1986
3. European Collaborative Study. Mother-to-child transmission of HIV infection in the era of highly active antiretroviral therapy
. Clin Infect Dis
4. Cooper ER, Charurat M, Mofenson L, Hanson IC, Pitt J, Diaz C, et al
. Combination antiretroviral strategies for the treatment of pregnant HIV
-1-infected women and prevention of perinatal HIV
-1 transmission. J Acquir Immune Defic Syndr 2002; 29:484–494.
5. Centers for Disease Control and Prevention. Achievements in public health. Reduction in perinatal transmission of HIV infection – United States, 1985–2005
. MMWR Morb Mortal Wkly Rep
6. British HIV
Association. Guidelines for the management of HIV infection in pregnant women and the prevention of mother-to-child transmission of HIV.
March 2005. Available at: www.bhiva.org/cms1191558.asp
. Accessed 9 October 2007.
7. The European Mode of Delivery Collaboration. Elective caesarean-section versus vaginal delivery in prevention of vertical HIV-1 transmission: a randomised clinical trial
8. Newell ML, Thorne C. Antiretroviral therapy and mother-to-child transmission of HIV
-1. Expert Rev Anti Infect Ther 2004; 2:717–732.
9. Jamieson DJ, Read JS, Kourtis AP, Durant TM, Lampe MA, Dominguez KL. Cesarean delivery for HIV
-infected women: recommendations and controversies. Am J Obstet Gynecol 2007; 197:S96–S100.
11. Townsend CL, Cliffe S, Tookey PA. Uptake of antenatal HIV
testing in the United Kingdom: 2000–2003. J Public Health (Oxf) 2006; 28:248–252.
12. The UK Collaborative Group for HIV
and STI Surveillance. A Complex Picture. HIV and other sexually transmitted infections in the United Kingdom: 2006
. London, Health Protection Agency Centre for Infections. 2006. Available at: www.hpa.org.uk/publications/2006/hiv_sti_2006/
. Accessed 9 October 2007.
13. Nicoll A, Lynn R, Rahi J, Verity C, Haines L. Public health outputs from the British Paediatric Surveillance Unit and similar clinician-based systems. J R Soc Med 2000; 93:580–585.
14. Centers for Disease Control and Prevention. 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults
. MMWR Recomm Rep
15. R Development Core Team. R: a language and environment for statistical computing.
Vienna, Austria, R Foundation for Statistical Computing. 2006. Available at: www.R-project.org
16. Townsend CL, Cortina-Borja M, Peckham CS, Tookey PA. Trends in management and outcome of pregnancies in HIV infected women in the United Kingdom and Ireland, 1990–2006
17. AIAU, NSHPC, and CHIVA. Perinatal Transmission of HIV in England 2002–2005. Executive Summary and Recommendations. October 2007
. Available at: www.nshpc.ucl.ac.uk
. Accessed 10 January 2008.
18. Larbalestier N, Mullen J, O'Shea S, Cottam F, Sabin CA, Chrystie IL, et al
. Drug resistance is uncommon in pregnant women with low viral loads taking zidovudine monotherapy to prevent perinatal HIV
transmission. AIDS 2003; 17:2665–2667.
19. Read P, Costelloe S, Mullen J. Zidovudine monotherapy in pregnancy
does not lead to detectable drug resistance when used in accordance with British HIV
Association Guidelines. British HIV Association Spring Meeting
. April 2006.
20. European Collaborative Study. Maternal viral load and vertical transmission of HIV-1: an important factor but not the only one
21. European Collaborative Study. Are girls more at risk of intrauterine-acquired HIV infection than boys? AIDS
22. Piwoz EG, Humphrey JH, Marinda ET, Mutasa K, Moulton LH, Iliff PJ. Effects of infant sex on mother-to-child transmission of HIV
-1 according to timing of infection in Zimbabwe. AIDS 2006; 20:1981–1984.
23. European Paediatric Hepatitis C Virus Network. A significant sex – but not elective cesarean section – effect on mother-to-child transmission of hepatitis C virus infection
. J Infect Dis
24. European Collaborative Study. Time to undetectable viral load after highly active antiretroviral therapy initiation among HIV-infected pregnant women
. Clin Infect Dis
25. Lyons L. Nevirapine tolerability in HIV infected women in pregnancy – a word of caution
. 2nd International AIDS Society Conference on HIV Pathogenesis and Treatment
. Paris, 13–17 July 2003. [Abstract LB27] Available at: www.aegis.org/conferences/IASHIVPT/2003/LB27.html
. Accessed 9 October 2007.
26. Thorne C, Newell ML. The safety of antiretroviral drugs in pregnancy
. Expert Opin Drug Saf 2005; 4:323–335.
27. European Collaborative Study. Increased risk of adverse pregnancy outcomes in HAART-treated HIV infected women in Europe
28. Suy A, Martinez E, Coll O, Lonca M, Palacio M, de Lazzari E, et al
. Increased risk of preeclampsia and fetal death in HIV
-infected pregnant women receiving highly active antiretroviral therapy
. AIDS 2006; 20:59–66.
29. Townsend CL, Cortina-Borja M, Peckham CS, Tookey PA. Antiretroviral therapy and premature delivery in diagnosed HIV
-infected women in the United Kingdom and Ireland. AIDS 2007; 21:1019–1026.
Keywords:© 2008 Lippincott Williams & Wilkins, Inc.
antiretroviral agents; highly active antiretroviral therapy; HIV; pregnancy; vertical disease transmission; viral load