Antenatal HIV testing

Has been done badly in Britain and needs to improve

The advantages of ascertaining a pregnant woman's HIV positive status before delivery are clear: transmission to the baby can be roughly halved by avoiding breast feeding1 and reduced by a further two thirds by the administration of zidovudine.2 Yet, as several papers in this week's issue show, in Britain we are failing to test pregnant women for HIV and, as a result, to reduce the rate of vertical transmission.

Undoubtedly there are psychological and social disadvantages to a woman in discovering that she is HIV positive, but these will inevitably occur at some time. The advantages of knowing are particularly great in pregnancy. As well as through avoiding breast feeding and using zidovudine, further reductions in the risk of transmission may be possible by offering caesarean section,3 using other antiretrovirals, and avoiding invasive procedures during vaginal delivery. The paper by Lyall et al shows that women will take these measures to prevent transmission (p 268).4 Some women may choose to terminate their pregnancy,5 and all can make informed decisions about further pregnancies: on p 271 Richardson and Sharland show that in many mothers infection is not diagnosed until their child has reached 1 year of age, when a subsequent pregnancy may already have started.6

It has been estimated that the number of infected babies in London could have been reduced from 40 to 13 a year if all positive women had been identified and their uptake of interventions to reduce vertical transmission had been at the rate observed in 1995.5 Despite an increase in prevalence among pregnant women in London to 0.18% in 1995 and Department of Health guidelines encouraging the offer of testing to all women in higher prevalence areas,7 Nicoll et al report antenatal detection rates of only 7%, with no signs of improvement over time (p 253).8 How can this situation in Britain be changed?

Compulsory testing is undesirable and illegal. There is no evidence that it would work, and it risks deterring women from seeking any antenatal care. Voluntary testing could be offered to all women antenatally or only to those at high risk. These approaches have not been subject to a randomised trial, but each has potential advantages and disadvantages. In particular, offering the test only to women with reported risk factors may miss cases,9 result in poor uptake,10 and be perceived as discriminatory.

Despite debate about the best approach, some things are already clear. In London at least, where the prevalence is highest, the test must be “normalised,” as advocated by De Cock and Johnson (p 290)11 by being offered and recommended to all women at their booking visit, alongside other blood tests. Simpson et al show that this does not cause anxiety or impair satisfaction with the booking visit (p 262).12 Improved and continuing training for midwives and other health care workers providing antenatal care is crucial to the success of this approach. Variability in the uptake of the HIV test according to the midwife seen (p 272),12 13 the fact that some pretest discussion actually dissuades women from testing (p 270),14 and a recent article in a nursing journal advocating breast feeding for women known to be seropositive15 suggest that some midwives remain to be persuaded of the benefits of testing. (Midwives rarely witness the consequences for the baby of vertical transmission of HIV.) The argument that HIV is a difficult subject to broach with women whose first language is not English may deny testing to those at highest risk and is untenable, as many other sensitive and complicated issues necessitate the provision of adequate interpreting services.

Accurate information for all pregnant women explaining the benefits, to themselves and their babies, of making a positive diagnosis before delivery is needed. Gibb et al show that any discussion of HIV transmission with a pregnant woman increased the likelihood of testing (p 259),16 and Simpson et al show that offering the test with or without a detailed discussion increases uptake.12 This discussion will add time to the booking visit, although this need not be excessive.12 13 16 Women's concerns about confidentiality must be addressed seriously and may require amendments to local policies. In other countries there is no evidence of widespread refusal of the HIV test, and uptake rates reach over 90% (p 293).17 18 Importantly, Gibb et al show no evidence of increased refusal rates among those at highest risk.16 There is anecdotal evidence that many women assume they are tested.

Gibb et al show that the booking unit is the strongest predictor of testing.16 Units claiming to have policies of universally offering an HIV test have rates of testing of 3.4-51.2%, suggesting widely divergent implementation. For those units participating in the unlinked anonymous surveys, comparisons are already available between actual and detected rates of HIV infection, but this information needs to reach the healthcare workers offering the HIV test. Furthermore, information systems in departments of pathology and antenatal clinics must be improved to facilitate routine audit of the rate of testing and to allow public health doctors to exert leverage by setting expected testing rates as part of the commissioning process.

An intercollegiate working party on antenatal HIV testing, including representatives of the royal colleges of midwives, obstetricians and gynaecologists, physicians, and general practitioners, among others, has prepared recommendations to reduce vertical transmission of HIV in the United Kingdom by increasing voluntary confidential HIV testing. It is addressing the issues considered above, but to be more effective than the current Department of Health guidelines it will need to be followed by a detailed implementation plan.

The indifference of some obstetricians and an unwillingness by many midwives to broach the issue of testing has meant that Britain has fallen behind other countries in providing pregnant women with access to HIV testing. It is shameful and negligent that we have counted the number of babies at risk of infection since 1990 without acting to reduce their risk.