Methods	RCT with randomisation of individual women.
Participants	Inclusion criteria Women with clinical diagnosis of incomplete miscarriage confirmed by transvaginal ultrasound. Specifically - women with clinical diagnosis of incomplete miscarriage, positive urinary pregnancy test, confirmed by transvaginal ultrasonography (TVS) with evidence of retained product of conception (POC). N = 201 women. Exclusion criteria Women with an intrauterine dimension measuring < 11 cm2 (sagittal plus transverse plane) were considered to have an empty uterus and excluded from randomisation. Also excluded were women with: severe blood loss; sepsis; known allergy to prostaglandins or analogue, history of asthma, clinician thought unsuitable for misoprostol.
Interventions	Intervention: vaginal misoprostol. 800 ug - 2 doses if necessary. N = 96. Comparison: oral misoprostol. 800 ug - 2 doses if necessary. N = 105.
Outcomes	Efficacy; side effects; short-term complications. outcomes assessed at 1 day following treatment and again at 2 weeks.
Notes	Setting: The Chinsese University of Hong Kong. Additional outcomes assessed but not pre-specified in the review: bleeding; pain; fever; drop in Hb.
Risk of bias
Bias	Authors’ judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer-generated set of random numbers in blocks of 5.
Allocation concealment (selection bias)	Low risk	Opaque envelopes labelled serially.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No information given.
Incomplete outcome data (attrition bias) All outcomes	High risk	Of the 201 women randomised, 198 got the treatment allocated, but only 186 were analysed because 12 were lost to follow up - 7.5%. It is unclear whether ITT analysis was undertaken
Selective reporting (reporting bias)	Unclear risk	No obvious outcome reporting bias but authors do not list their outcomes and although only report significant differences in abstract, in paper they report several adverse outcomes with data. We did not assess the trial protocol.
Other bias	Low risk	Significantly more women in oral group had a past history of termination, P < 0. 001, but this was thought to probably not to create important bias