Africa 2011.

Methods	Multicentre, randomised, double blind, placebo‐controlled trial.
Participants	1099 women without significant obstetric complications expected to give birth vaginally at 6 different hospitals in South Africa, Uganda and Nigeria. Those who delivered via caesarean or had an assisted vaginal birth; those in whom sublingual administration of misoprostol was not possible; those in whom the pregnancy was not viable according to local gestational viability age limits; those who declined, or were too ill or distressed, to give consent; and those not entitled to give informed consent, such as minors were excluded.
Interventions	400 mcg of misoprostol immediately after delivery, in addition to standard active management of the third stage of labour, as currently practiced in the collaborating centres (parenteral oxytocin in Uganda and South Africa, and oxytocin or ergometrine in Nigeria) versus placebo immediately after delivery, in addition to standard active management of the third stage of labour, as currently practiced in the collaborating centres.
Outcomes	The primary outcome was the incidence of 500 mL or more of measured blood loss within 1 hour after the trial medication was administered. Secondary outcomes were the mean measured blood loss 1 hour after the trial medication was administered, a measured blood loss of 1000 mL or more, adverse effects such as shivering (mild and moderate/severe) and pyrexia, manual removal of the placenta, laparotomy, hysterectomy, and maternal morbidity and mortality.
Notes	Active management of third stage labour. Blood collection was started as soon as possible after delivery. A fresh, non‐absorbent sheet was placed under the buttocks of the woman. A low‐ profile, plastic bedpan was positioned below the woman's perineum in a position to collect all subsequent blood loss, for a period of 1 hour, measured by a standard clock. The blood collected in the bedpan and any spilled blood, which was collected from the non‐absorbent sheet and any blood‐soaked small gauze swabs, were transferred to a measuring jar and the volume was measured. Alternatively, the blood was collected into a plastic sheet placed below the woman immediately after delivery.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The random allocation sequence was generated by using computer‐generated random numbers and was stratified by country in blocks of 6‐8."
Allocation concealment (selection bias)	Low risk	"The packs were identical in shape, color, weight, and feel, and contained either 2 tablets of 200 mcg of misoprostol or 2 matching placebo tablets."
Blinding (performance bias and detection bias) All outcomes	Low risk	Double‐blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	4 people were lost to follow‐up due to data not recorded.
Selective reporting (reporting bias)	Low risk	Primary outcomes of the review were reported.
Other bias	Low risk	The 2 groups were similar in terms of baseline characteristics, except for a 2.8‐year difference in mean age.