Objective: To test the efficacy of lower doses of mifepristone and gemeprost for medical induction of early abortion.
Design: Randomised controlled trial. Participants were blinded as to the therapy and physicians to the dose of mifepristone.
Setting: Thirteen hospital gynaecological units in different continents.
Participants: 1,224 healthy pregnant women requesting medical abortion at <57 days from last menses.
Intervention: Random allocation to one of four regimens: mifepristone 50 mg by mouth followed by either 0.5 mg or 1.0 mg gemeprost vaginally on day 3; mifepristone 200 mg by mouth followed by either 0.5 mg or 1.0 mg gemeprost vaginally. We concealed the allocation sequence from clinicians enrolling participants, and maintained double blinding throughout.
Main outcome measures: Incidence of complete abortion; subordinate outcome measures included side effects such as vomiting and fall in haemoglobin, as well as the need for emergency curettage and blood transfusion.
Results: The success rate was significantly related to the dose of mifepristone. The relative risk of failure to have a complete abortion with the lower dose of mifepristone was 1.6 (95% CI: 1.1-2.3) times that with the higher dose. The relative risk of failure with the lower dose of gemeprost (1.3; 95% CI: 0.9-1.8) did not reach statistical significance.
Conclusions: A single dose of mifepristone 50 mg followed by gemeprost is inadequate for early medical abortion. There was no significant difference in side effects between the four treatment groups.