According to UNAIDS, each day 1000 children are infected with HIV; 90% of those younger than 15 years receive the virus through mothertochild transmission. If HIV transmission during pregnancy and delivery is not reduced, AIDS may increase infant mortality by as much as 75%. A range of interventions is therefore being tried out to reduce such HIV transmission.

Because of some reports that low vitamin-A levels increase the chance of perinatal transmission, some advocate supplemental vitamin-A administration to pregnant women (e.g., in Zambia and Zimbabwe). A study among 95 American women by the Wadsworth Center in Albany, NY, however, concluded that there is no relationship between low vitamin-A levels and HIV transmission during pregnancy. In any event, care must be taken in this regard as too much vitamin A can lead to abnormalities in the child. The Wadsworth team recommends that women living in countries where vitamin-A deficiency is not a problem should avoid taking such supplements.

In light of some evidence that transmission risks are greater if a baby is born before 37 weeks of pregnancy, attention should be given to promoting good antenatal care for all women. Measures might include decreasing her workload, improving her diet, increasing access to antenatal care provision and monitoring, identifying and  treating STDs.

Some research has explored whether Caesarean delivery may reduce HIV transmission. The data at present are mixed, with some studies showing an advantage and some a disadvantage. A study is currently under way, but until definitive data are at hand and in the absence of definitive findings, there may be problems carrying out this procedure.

Data show that women living with HIV who have Caesarean sections are more likely to develop complications related to operative delivery. Therefore, Caesarean sections are generally recommended only for other appropriate indications (e.g., prolonged labour with ruptured membranes, with a higher time interval for women who are giving birth to their first child).

Currently, the procedure receiving the greatest attention in the literature is administration of zidovudine (ZDV) to pregnant women living with HIV for a few weeks before birth, during delivery and 6 weeks afterwards to both the woman and her baby. This requires that a woman's serostatus is known; for this reason, antenatal clinic (ANC) HIV testing is being promoted. A number of factors, however, argue against ANC testing:

• it is traumatic; women are approached at a time when they are very vulnerable
• men (fathers) are rarely tested
• the principle of informed consent is often violated
• adequate counselling is often overlooked (a review of the literature revealed that only two published studies on ANC testing mentioned that counselling took place and in one case the counselling took place during labour!)
• when counselling is done, the content may be inappropriate: women are only told the test will be done "to see that everything is ok" (pretest counselling) or risk reduction and behaviour change are not addressed (posttest counselling)
• the costs are considerable.

It is still uncertain whether and to what extent the major study underlying the promotion of ANC testing and ZDV administration is generalizable. Moreover, it should be remembered that transmission in the American study was reduced to 8%  not 0%  and the longterm effects of the drug on uninfected children are unknown. Also, what are the implications for mothers, especially since ZDV monotherapy is no longer recommended? The mother may build up resistance and be unable to benefit from the new combination therapies in the future. Thus, the very act of reducing the chances of an infected baby may increase the likelihood of early orphanhood for that child.

It is not my opinion that ANC testing and ZDV administration should be avoided at all costs. All women should have a right to request an antenatal HIV test. ANC testing may be useful in order to predict the spread of HIV, understand perinatal transmission and monitor pregnancies; moreover, many women may still choose the option of ZDV administration if they are properly informed about the possible advantages and disadvantages. In any event, ANC testing should not be done unless free care for the mother and child can be offered for at least two years: during the nine months of pregnancy and 18 months thereafter until the baby can be tested for HIV with some certainty.

Also of concern is that, from 1985-1995, only about 7% of all studies in the field of HIV/AIDS concerned women, strongly pointing to the need for more attention to women in connection with HIV/AIDS. Only when this is done will the various measures being taken to reduce HIV perinatal transmission show a true chance of success.

Lorraine Sherr, Royal Free Hospital, School of Medicine, Rowland Hill Street, London NW3 2PF, United Kingdom; tel: 441718302129; fax: 441813462961; email: lsherr@rfhsm.ac.uk.