Intended for healthcare professionals


Descriptive survey of antenatal HIV testing in London: policy, uptake, and detection

BMJ 1996; 313 doi: (Published 31 August 1996) Cite this as: BMJ 1996;313:532
  1. Sandra E MacDonagh, research midwifea,
  2. Janet Masters, research assistanta,
  3. Barbara A Helps, research fellowa,
  4. Pat A Tookey, research fellowa,
  5. A E Ades, reader in epidemiology and biostatisticsa,
  6. Diana M Gibba, senior lecturer in paediatric epidemiology
  1. a Epidemiology and Biostatistics Unit, Institute of Child Health, London WC1N 1EH
  1. Correspondence to: Ms MacDonagh.
  • Accepted 12 April 1996

Avoidance of breastfeeding1 and use of antiretroviral therapy2 can substantially reduce the risk of a mother transmitting HIV to her child. The Department of Health has recommended that antenatal HIV testing should be offered to all women in areas of high prevalence such as London.3 We examined HIV testing policy and practice in London maternity units and related policy to uptake of testing and detection of previously undiagnosed HIV infection.

Methods and results

In spring 1995 one interviewer conducted a telephone interview with a professional involved with HIV issues at each of London's 33 maternity units. Information on women booking for antenatal care over the previous 12 months and the number who elected to have an HIV test were collected by postal questionnaire from 32 units. Copies of written information available for staff and women were also collected.

The total number of HIV infected women delivering in 1993 and 1994 in each unit was estimated from unlinked anonymous antenatal and neonatal seroprevalence studies (unlinked anonymous HIV seroprevalence monitoring programmes in England and Wales, data to the end of 1994, Public Health Laboratory service, Institute of Child Health). Confidential reports of pregnant HIV infected women4 over the same period were examined to determine how many women had been diagnosed before or during the reported pregnancy. The ratio of the number of women first diagnosed during the reported pregnancy to the total number of infected women minus those previously diagnosed was calculated to assess the efficiency of the screening policy.5

In 13 units all women were offered an HIV test (universal policy), in 14 only those perceived to be at high risk of infection were offered testing (selective policy), and in 6 only those who asked were tested (request policy). In 15 units at least one staff member had specific responsibility for HIV issues. Training on HIV in pregnancy was provided for staff, mainly midwives, in 12 units, but only 9 units had written protocols that included guidelines on the procedure for offering HIV testing. Written information about HIV infection in pregnancy was available for women attending 23 units and is summarised in table 1.

Table 1

Antenatal HIV testing and seroprevalence by policy type in London maternity units

View this table:

Overall only 22 of an estimated 322 previously undiagnosed HIV infected women booked in these units during 1993 and 1994 were identified, 2/126 in units with selective or request policies and 20/196 in units with a universal policy. These women were identified in 9 of the 13 units with a universal policy compared with only 1 of the 20 units with other policies (P = 0.0002, two tailed Fisher's exact test).

Information on uptake of HIV testing over the previous year could be supplied by only 24 units. In most the uptake was low, being under 10% in 7/11 units with universal policies and never exceeding 1.5% in the units with other policies (table 1). Detection of previously undiagnosed infected women in units with a universal policy was significantly increased with higher uptake of testing. This finding was largely attributable to those two units where uptake of testing was over 40% and detection of previously undiagnosed women exceeded 20%.


In view of the high prevalence of HIV infection among pregnant women in London (1 in 580) and the real potential to reduce mother to child transmission the failure of maternity services to identify infected women is of concern. In many units no written information on HIV and pregnancy was available for women and, where it was, information on the potential to reduce transmission was often omitted. In addition in most units midwives had received no training to enable them to discuss these issues with women. The inability of nine units to supply data on the number of women tested is also worrying. Only after both the uptake and process of offering antenatal HIV testing is routinely audited can more detailed study of the cost effectiveness of different approaches be undertaken.

We thank Patsy Tuck, colleagues at the Public Health Laboratory Service, all the directors of midwifery services and interviewees in the maternity units, and Professor Catherine Peckham for her comments and support.


  • Funding Department of Health and the AIDS Education and Research Trust (AVERT).

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
View Abstract