Review of uptake of interventions to reduce mother to child transmission of HIV by women aware of their HIV statusBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7127.268 (Published 24 January 1998) Cite this as: BMJ 1998;316:268
- E G Hermione Lyall, lecturer in paediatric infectious diseasesa (, )
- Christopher Stainsby, research assistantb,
- Graham P Taylor, lecturer in genitourinary medicine and communicable diseasesb,
- Mounir Ait-Khaled, research fellowb,
- Samantha Bingham, research assistantb,
- Jennifer A Evans, lecturer in paediatric infectious diseasesa,
- Anne Wright, research fellowb,
- Jonathan N Weber, professor of genitourinary medicine and communicable diseasesb,
- Myra O McClure, senior lecturer in virologyb,
- Sam Walters, senior lecturer in paediatric infectious diseasesa,
- Gareth Tudor-Williams, senior lecturer in paediatric infectious diseasesa
- a Department of Paediatrics, Imperial College School of Medicine at St Mary's, London W2 1PG
- b Department of Genitourinary Medicine and Communicable Diseases, Jefferiss Research Trust Laboratories, Imperial College School of Medicine at St Mary's, London W2 1PG
- Correspondence to: Dr Lyall
- Accepted 22 December 1997
Objectives: To examine the change in uptake of interventions to reduce transmission of HIV from mothers to infants from January 1994 to July 1997.
Design: Review of mother-infant pairs who presented for infant diagnosis of HIV infection.
Setting: Central London hospital with facilities for diagnosis of infant HIV infection.
Subjects: 57 consecutive mother-infant pairs, mainly from central London but also referred from surrounding hospitals.
Interventions: Data were collected on mother's country of origin; CD4 count at delivery; plasma HIV RNA copies/ml; mode of delivery; antiretroviral therapy; infant feeding; and HIV infection in infants.
Main outcome measures: HIV infection of infants.
Results: The vertical transmission rate was 12% (7 pairs; 95% confidence interval 3% to 22%). All mothers chose not to breast feed. The caesarean section rate was 53% (30/57). Antiretroviral therapy was taken by 68.5% (39/57) of mother-infant pairs. With antiretroviral therapy or caesarean section, or both, transmission occurred in 6% (0% to 13%) of pairs (3/50). During the 24 months of 1994 and 1995, 21% (4/19) of infants were infected with HIV; 7.9% (3/38) were infected over the 19 months January 1996 to July 1997. The caesarean section rate did not change over these periods. Use of antiretroviral therapy increased from 31.5% (6/19) to 86.8% (33/38) (P<0.0001).
Conclusion: Women with a diagnosis of HIV infection acted to reduce the risk of transmission to their infants. Uptake of antiretroviral therapy increased significantly over time, and the caesarean section rate was persistently high.
In this London cohort, pregnant women who knew they had HIV all acted to reduce vertical transmission
No mothers breast fed, and about half chose to have a caesarean section
Mothers' use of antiretroviral therapy increased from 32% in 1994–5 to 87% in 1996-7
With antiretroviral therapy or caesarean section, or both, vertical transmission occurred in 6% of mother-infant pairs
Improving uptake of antenatal testing for HIV continues to be important
The Paediatric AIDS Clinical Trials Group published the results of their ACTG 076 study of zidovudine (AZT) versus placebo for prevention of vertical transmission of HIV in 1994, and use of zidovudine in pregnancy was recognised as an appropriate standard of care for prevention of vertical transmission in the United States that same year.1 2 This study's result has now been widely reproduced in general clinical practice and in women with more advanced HIV disease.3 4 5 In the United Kingdom, use of antiretroviral therapy during pregnancy has gradually increased over the past three and a half years. However, this has been slower than in the United States or other European countries because of a low rate of antenatal diagnosis of HlV infection.6 7
We retrospectively reviewed all the mother-infant pairs presenting to our service from January 1994 to July 1997 for diagnosis of HIV infection in the infant by polymerase chain reaction and culture. The aim of this review was to examine the uptake of interventions to prevent transmission over this period. The mother-infant pairs fell into two groups: those in which mothers were known to have HIV infection in pregnancy or before and could consider options to reduce transmission; and those in which a family member, most often the infant (17 of 24), presented with HIV related symptoms. Those 24 mothers were unaware of their HIV infection antenatally and are only briefly presented here: all but one breast fed; 18 (75%) of their infants were infected, and 11 developed AIDS during infancy.
Women known to have HIV in pregnancy were counselled about the risks of transmission and the protective effects to the infant of antiretroviral therapy, avoidance of breast feeding, and the possible protective effect of caesarean section. The mother decided whether to opt for caesarean section.
Data were collected on mother's country of origin; CD4 count at delivery; plasma HIV RNA burden before antiretroviral therapy and at delivery; mode of delivery; gestation at delivery; antiretroviral therapy regimen; infant feeding; and infection outcome for the infant. For mothers from some hospitals other than St Mary's not all data were available. Plasma concentrations of HIV RNA were measured by at least one of three commercial assays (Chiron, Organon Technica, and Roche) and the highest reading obtained was used for analysis. HIV RNA measurements were available for mothers delivering at St Mary's Hospital and for some mothers delivering elsewhere. Diagnosis of HIV infection in infants was by polymerase chain reaction of HIV DNA and HIV culture of blood mononuclear cells taken at 1 day, 4 weeks, 3 months, and 6 months of age.8
Medians were compared by Mann-Whitney U test, and the numbers of patients in distinct groups were compared with Fisher's exact test; 95% confidence intervals were generated treating the observed number as a Poisson variable.
In 57 consecutive mother-infant pairs, 81% of mothers (46) were of African origin. Vertical transmission of HIV occurred in seven pairs (12%; 95% confidence interval 3% to 22%). Thirteen mothers were diagnosed as HIV infected during the current pregnancy, 34 were known to have HIV before this pregnancy, and the exact timing of diagnosis was unknown for 10. Eleven mothers already had a child with HIV infection. Three children developed AIDS within the first year of life. Four children were delivered prematurely, of whom one was infected.
All mothers chose not to breast feed. The caesarean rate was 53% (30/57), including six emergencies. The rate of vertical transmission was 3% (1/30) by caesarean section and 22% (6/21) by vaginal delivery (1).
Antiretroviral therapy was taken by 69% (39/57) of mother-infant pairs. Twenty six (46%) took a full 076-type course of zidovudine and seven took an incomplete course with only one or two parts of the regimen (antepartum, intrapartum, or postpartum). Six mothers took combination antiretroviral therapy antenatally (five took zidovudine plus lamivudine; one took zidovudine plus lamivudine and indinavir) and then completed the 076 course. Eighteen of the mothers (32%) did not take or did not tolerate antiretroviral therapy. Transmission rate was 8% (3/39) for mothers who took any kind of antiretroviral therapy and 22% (4/18) for those who did not. In the seven pairs who took an incomplete antiretroviral course, two infants were infected. Transmission occurred in three of 50 pairs (6%; 0% to 13%) with antiretroviral therapy or caesarean section, as opposed to four of seven pairs (57%) with vaginal delivery without antiretroviral therapy (P=0.003) (1).
Although this was a mainly African cohort of mothers, ethnic origin did not seem to affect mothers' choice of interventions. CD4 counts were available for only 34 mothers and HIV RNA load for 36; although this is only 60% of the group, we have no reason to suspect it is not representative of the cohort. The median CD4 count at delivery was 305x106/l, and vertical transmission occurred at all levels. Median plasma HIV RNA was log 3.6 copies/ml at delivery, and transmission occurred across the range of levels. The viral load was slightly less in 26 women who took antiretroviral therapy (log 3.43 v log 3.71; P=0.06). Five mothers had a viral load below the level of detection of the kits (<log 3.0 copies/ml); all took antiretroviral therapy and none transmitted HIV infection. Four of these mothers were taking combination antiretroviral therapy for either more advanced HIV disease or higher viral load at presentation.
Using the infant's date of birth as the reference point, we assessed 19 mother-infant pairs over the 24 months 1994–5 and 38 over the 19 months January 1996 to July 1997. During the first time period 21% (4/19) of infants were infected with HIV, and 8% (3/38) were infected in the second period. The rate of caesarean section did not change (58% v 50%). Use of antiretroviral therapy increased from 32% (6/19) to 87% (33/38) (P<0.0001).
This small observational study shows that women diagnosed as HIV infected before or during pregnancy all acted to reduce the risk of transmission to their infants. No mother who knew she had HIV infection breast fed. Uptake of antiretroviral therapy increased significantly over time, and the caesarean section rate was persistently high. Mothers who received antiretroviral therapy or caesarean section, or both, had a transmission rate of 6%; the overall transmission rate for the group was 12%. This cohort is too small to study the separate effects of variables such as viral load, CD4 count, antiretroviral therapy, and caesarean section on transmission. HIV transmission occurred at all levels of viral load and CD4 count; this concurs with the findings of the 076 study, where zidovudine did not alter this pattern but reduced transmission at all levels.9
Our findings complement those of an earlier study of uptake of interventions in the United Kingdom that showed a continued increase in uptake of interventions by mothers.7 In our study ethnic group did not affect mothers' choice of interventions.
With improved antenatal detection of HIV infection and use of antiretroviral therapy, vertical transmission should greatly decline. This will result in increased exposure of uninfected infants to potent drugs in utero and around the time of birth. Although there is no evidence of short term deleterious effects of zidovudine, the children from the 076 study will be followed to adulthood to monitor late effects. There is much less data on effects of other drugs now being used in pregnancy. Clinicians prescribing antiretroviral therapy should report to the International Registry of Antiretroviral Use in Pregnancy (c/o Glaxo Wellcome, Greenford, Middlesex UB6 OHE) so that information can be gathered on the increasing number of exposed but uninfected children.
We are grateful to St Mary's Hospital Trust midwives, especially Ms S Dick, Ms M Hornby, and Ms F Clayton, for their assistance with the study and to specialist nurse Paula Seery, Chelsea and Westminster Hospital, for collection of data. We are grateful to clinicians in the following hospitals for allowing us to study their patients: Chelsea and Westminster, Hillingdon, Central Middlesex, North Middlesex, West Middlesex, Milton Keynes, Northampton General, Northwick Park, Royal Free, and Watford General. Thanks to R Booy for fruitful discussions.
Funding: No additional funding.
Conflict of interest: None.
Contributors: JNW, MOMcC, JAE, and SW set up and have monitored the infant HIV diagnostic service. JNW and GT-W were the instigators of this paper. EGHL wrote the paper and CS, GPT, JNW, MOMcC, SW, and GT-W took part in its editing. CS, MA-K, and SB carried out laboratory testing of the infants and mothers. EGHL, GPT, and CS collated the results; Paula Seery collated results from the Chelsea and Westminster Hospital. Contributors from the clinical paediatric service were SW, GT-W, JAE, Stephen Marriage, and EGHL; contributors from the clinical adult service were GPT, Siobhan Crowley, and AW; contributors from obstetrics and midwifery were Ms S Dick, Ms M Hornby, Ms F Clayton, and Mr J Smith. Other clinicians from surrounding hospitals who referred their patients also contributed clinically (see acknowledgements). The guarantor is Hermione Lyall.