Rogers 1998.

Methods	RCT with randomisation of individual women in balanced blocks, with allocation to 1 of 2 delivery postures within each arm
Participants	UK hospital setting. High‐income country Inclusion criteria: 1512 women at low risk of PPH giving birth at study hospital (including water births) Exclusion criteria: placenta praevia, previous PPH, APH after 20 weeks’ gestation, Hb < 10 g/dL or MCV < 75 fL, non‐cephalic presentation, multiple pregnancy, intrauterine death, epidural anaesthesia, parity > 5, uterine fibroid, oxytocin augmentation infusion, anticoagulation therapy, intended instrumental or OVB, duration of gestation < 32 weeks, (plus any other contraindication, in clinician’s view)
Interventions	Intervention: active management of 3rd stage (N = 748) 2 arms: active management ‐ upright position (N = 374); active management ‐ supine position (N = 374) prophylactic oxytocin (19.5%) or "oxytocin + ergometrine" (75%) as soon as possible after birth of anterior shoulder (within 2 min of birth). Number of units/mL not given, nor reason for the difference immediate cord clamping and cutting delivery of the placenta by CCT or maternal effort Control: expectant management of 3rd stage (N = 764) 2 arms: expectant management ‐ upright position (N = 381); expectant management ‐ supine position (N = 383) no prophylactic administration of uterotonic drug no cord clamping until after pulsation ceased delivery of placenta within 1 h by maternal effort Data included in comparisons 1 and 2
Outcomes	Pre‐specified outcomes: PPH (> 500 mL) as assessed/estimated by the attending MW (used for power calculation); severe PPH 1000 mL), blood transfusion, iron tablets postnatally, Hb at 24‐48 h P/N, self‐completed questionnaire on maternal fatigue and depression at 6 weeks P/N, nausea, vomiting, headache, hypertension, manual removal of placenta, ERPC, neonatal outcomes, views of mothers and staff
Notes	Actual management used in the active arm: 699 (93.4%) had full active management; 95% given prophylactic uterotonic before delivery of placenta; 93% cord clamped before pulsation ceased; 46% CCT; 44% upright Actual management used in the expectant arm: 488 (63.9%) had full expectant management; 21% received oxytocic for treatment, and 2.5% prophylactically; cord left unclamped till pulsation ceased 70%; placenta delivered by CCT 12%; 43% upright The setting is described as one where the midwives were “similarly confident” in active and expectant management. However, the questionnaire administered to 92 of the 153 midwives prior to the study commencement showed that, whereas 84% felt “very confident” of active management, only 41% were “very confident” of expectant management. Maternal mean Hb levels were reported with SEs and so we calculated SDs Dates of study: June 1993‐December 1995 Funding sources: “The study was supported by a grant from the Public Health and Operational Research Committee of the Anglia and Oxford Regional Health Authority. The NPEU is supported by the Department of Health”. Declarations of interest: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Variable sized balanced blocks “...randomisation envelopes were prepared in advance..” in an external academic unit ‐ National Perinatal Epidemiology Unit, Oxford
Allocation concealment (selection bias)	Low risk	Sequentially numbered, opaque, sealed envelopes stored on the ward. Entry to the study occurred when an envelope was opened.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not possible to blind participants and personnel
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Partly blinded. The technicians who did antenatal and postnatal blood tests were unaware of allocation
Incomplete outcome data (attrition bias) All outcomes	Low risk	Data available on 1507 out of 1512 at discharge (< 0.5% attrition, approximately equal losses in both groups) At 6 weeks' follow‐up < 5% attrition
Selective reporting (reporting bias)	High risk	No study protocol found. A large number of outcomes reported but not noted in methods of paper
Other bias	High risk	Initial power calculation suggested a sample size of 2000. Interim analysis showed a higher PPH rate than expected, so sample size was revised to 1500 and the study stopped earlier than expected Similar at baseline 93.4% of active management group received active management. Only 63.9% of women in expectant management group received expectant management Also: MWs not confident re 1 arm, so observer bias the hypothesis was phrased in favour of active management lowering PPH rate and other complications, rather than being a null hypothesis setting is described as one where the midwives were “similarly confident” in active and expectant management. However, the questionnaire administered to 92 of the 153 midwives prior to the study commencement showed that, whereas 84% felt “very confident” of active management, only 41% were “very confident” of expectant management