A RANDOMIZED TRIAL OF MODE OF DELIVERY IN WOMEN INFECTED WITH THE HUMAN-IMMUNODEFICIENCY-VIRUS

Objective During the pilot phase of a trial to evaluate the effectiven ess of caesarean section delivery compared with vaginal delivery in re ducing mother-to-child transmission of human immunodeficiency virus (H IV) infection, the feasibility of randomisation to mode of delivery wa s assessed. Design At 36 weeks of pregnancy, women infected with HIV w ere randomly allocated to either caesarean section delivery at 38 week s or vaginal delivery. Information was also collected on the reasons w hy women were not enrolled, either because they refused or had a contr aindication. Setting Fifty-one centres in six European countries. Popu lation Pregnant women with confirmed HIV-1 infection Main outcome meas ures Randomisation. Results Three-hundred and thirty-nine women had be en randomised by the end of 1996, the large majority from Italy (n = 2 50) and France (n = 54), with 22 from South Africa, three from Sweden, nine from Barcelona and one from London. A further 150 women were eli gible but had not been randomised. Forty-eight women (14%) were not de livered according to the arm to which they were randomised; the majori ty (n = 44) were changed from vaginal to caesarean section delivery. T here is wide variation between European countries in the acceptability and adherence to the mode of delivery trial. Conclusion The pilot pha se of this trial has shown that in some settings randomisation to mode of delivery is feasible and acceptable, but that in other setttings c linicians and pregnant women are more reluctant to be randomised. Pend ing further information on transmission rates and accrual, enrollment into the trial continues.