Hofmeyr 2011.

Methods	2‐arm placebo‐controlled randomised trial
Participants	1103 women were randomised in a hospital setting in South Africa, Uganda, and Nigeria. The population comprised women of any parity, unspecified whether singleton or multiple pregnancy, at both high and low risk for PPH, who delivered by vaginal delivery. Exclusion criteria comprised women undergoing caesarean section or instrumental delivery, or those who declined participation or were unable to consent, were too ill or distressed to participate or with a not viable pregnancy.
Interventions	400 mcg plus 10 IU of misoprostol plus oxytocin administered sublingually plus IM versus 10 IU of oxytocin administered IM
Outcomes	The study recorded the following outcomes :PPH at 500; PPH at 1000; manual removal of placenta; death; blood loss (mL); fever; shivering.
Notes	Contact with study authors for additional information: yes. Additional data from authors: yes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random numbers and was stratified by country in blocks of 6–8.
Allocation concealment (selection bias)	Low risk	Quote: "The trial medication was provided, and the study drug packs were prepared, by Gynuity Health Projects. When a participant enrolled, the researcher took the next study drug pack from the dispenser and immediately wrote the woman's name both on the pack and in the participant number list, which was kept separate from the case record forms. Enrolment took place when the pack was removed from the pack dispenser. The pack could not be used for another woman or returned to the dispenser."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The study was "double‐blind". Quote: "The packs were identical in shape, colour, weight, and feel, and contained either 2 tablets of 200 mcg of misoprostol (HRA Pharma, Paris, France) or 2 matching placebo tablets".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Assessors were blinded to treatment allocations.
Objective assessment of blood loss	Low risk	Similarly to the study team of Gulmezoglu 2001, investigators appraised blood loss by collection with a fresh non‐absorbent sheet and low plastic “fracture” bedpan placed under the mother from as soon as possible after delivery until 1 hour postpartum. Investigators considered that quote: "longer‐term blood loss measurement is more difficult to standardise". They transferred the blood collected in the sheet and the bedpan (together with any soaked small gauze swabs) to a measuring jar to ascertain the volume. Alternatively, they collected blood with a plastic sheet placed under the mother immediately after delivery. If bleeding continued beyond 1 hour, investigators restarted collection and measurement until bleeding subsided. Attempts were made to minimise any losses on the drapes and gowns of delivery attendants. In addition, quote: "the placental interstices also contain maternal blood (about 9% of placental weight)." Because overestimations (amniotic fluid) and underestimations (blood loss) were likely to be distributed equally between the 2 study groups, and most would have occurred before the onset of measurement, the data were not corrected.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "Data for the primary outcome were not available for 4 of the 1103 women".
Selective reporting (reporting bias)	High risk	The prospectively registered protocol of the study (ClinicalTrials.gov NCT 00124540) lists some secondary outcomes different to those included the study report (≥ 1000 mL within the first hour only, transfusion, Hb < 8 g/dL 24 hours after delivery).
Intention to treat analysis	Low risk	All those who were enrolled and randomly allocated to treatment were included in the analysis, in the groups to which they were randomised.
Funding source	Low risk	The study was supported by funding from Gynuity Health Projects through a grant from the Bill and Melinda Gates Foundation (public funding).