Methods	RCT with randomisation of individual women, in blocks of 10 and stratified by site (2 sites involved)
Participants	Inclusion criteria • Women with incomplete spontaneous or induced miscarriage, less than 12 weeks’ gestation, diagnosed using ultrasound. • Uterine size equivalent to a gestation of less than 12 weeks LMP, open cervical os, past or present history of vaginal bleeding during pregnancy and ultrasound evidence of substantial uterine debris with evidence of fetal demise. • Women living or working within the hospital’s geographical area of coverage, no known contraindications to misoprostol, no signs of severe infection, temperature < 38 °C and general good health. • N = 460 women. Exclusion criteria • Women with very high fever; signs of severe infection.
Interventions	Intervention: oral misoprostol. • 600 ug, single dose. • N = 233. Comparison: surgery. • MVA. • N = 227.
Outcomes	Complete miscarriage following initial treatment; adverse effects, bleeding; pain (7-point Likert scale), acceptability (5-point Likert scale) • Assessed at 1 week using clinical assessments and US. Women could wait a further week before surgery (MVA) if they wished.
Notes	1. Setting: 2 large university teaching hospitals in Burkina Faso, Sub-Saharan Africa. 2. Additional outcomes assessed but not pre-specified in the review: pain/cramps; fever; chills; bleeding; overall experience; overall satisfaction; would choose again; would recommend to a friend; hospitalisation; managed pain with paracetamol; would have liked stronger pain killers; sought contact with providers; made phone calls to providers; best and worst features.
Risk of bias
Bias	Authors’ judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	“…computer-generated random sequence provided by Genunity Health Projects…”
Allocation concealment (selection bias)	Low risk	“The assignment was concealed from providers and participants until after informed consent was given when the next sequential opaque sealed study envelope was opened to reveal allocation…”
Blinding (performance bias and detection bias) All outcomes	High risk	“Neither women nor providers were blinded to treatment assignment…”
Incomplete outcome data (attrition bias) All outcomes	Low risk	• Lost after randomisation and before Rx: 10 in misoprostol group and 3 in the MVA group. • Exclusion after randomisation: 5 in the misoprostol group and 1 in the MVA group. • Overall, there were uneven loses to follow and some exclusions, but as numbers are small and we think this is unlikely to cause bias.
Selective reporting (reporting bias)	Unclear risk	Pre-specified outcomes reported on, but the trial protocol not assessed
Other bias	Low risk	No apparent biases from other sources. Baseline data showed no statistically significant differences between the groups