Hilton 2009.

Methods	RCT, stratified by parity and hospital.
Participants	Inclusion criteria: non‐cephalic singleton pregnancies over 37 weeks with normal AFI. Participants split into nulliparous (N = 82) and multiparous (N = 44). N = 126. Exclusion criteria: labour, ruptured membranes, history of third‐trimester bleeding, any preexisting uterine scar, pregnancy‐induced hypertension or gestational diabetes, oligohydramnios, hydramnios, IUGR, macrosomia, maternal hypotension, inability to comprehend the consent form.
Interventions	Intervention:tocolytic: nitroglycerine ‐ nitric oxide donor (A3) parenteral. IV nitroglycerine (10 mL of 100 micrograms/mL). N = 65 (nulliparous = 42, multiparous = 23). Comparison:placebo. N = 61 (nulliparous = 40, multiparous = 21).
Outcomes	ECV success; cephalic presentation at delivery; CS rate; maternal discomfort; headaches; flushing; hypotension; palpitations; fetal heart rate abnormalities.
Notes	Nulliparous group: 4 women excluded after randomisation. In experimental group, 1 excluded for pregnancy‐induced hypertension, and 1 excluded for decreased AFI. Control group: 1 excluded as woman was in labour, and 1 excluded because of cephalic presentation. 1 woman in placebo group lost to follow‐up. Multiparous group: 3 women excluded after randomisation. In experimental group, 2 excluded as presentation was cephalic at the time of version. In placebo group, 1 woman excluded as presentation was cephalic at time of version.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation tables used.
Allocation concealment (selection bias)	Low risk	Using separate randomisation sequences for nulliparous and multiparous women at each hospital site, participants were assigned a study number from sequentially numbered opaque envelopes. The study number was forwarded to the pharmacy, and allocated treatment was provided on the basis of the corresponding study number from randomisation tables kept in the pharmacy. No further details provided on randomisation sequences used. Group of allocation was unknown to obstetrician, nurse, anaesthesiologist and woman.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Treatment was prepared as 10 mL of clear fluid in a 10 mL syringe with 10 mL of 100 micrograms/mL of nitroglycerin for women in the nitroglycerin group, or 10 mL of normal saline for women in the placebo group. Syringes for nitroglycerine and placebo were visually indistinguishable. Group for allocation was unknown to obstetrician, nurse, anaesthesiologist and woman.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Treatment was prepared as 10 mL of clear fluid in a 10 mL syringe with 10 mL of 100 micrograms/mL of nitroglycerin for women in the nitroglycerin group, or 10 mL of normal saline for women in the placebo group. Syringes for nitroglycerine and placebo were visually indistinguishable. Group for allocation was unknown to obstetrician, nurse, anaesthesiologist and woman.
Incomplete outcome data (attrition bias) All outcomes	Low risk	1 woman lost to follow‐up. No apparent exclusion of women after randomisation. 7 women did not undergo ECV, but their outcomes were included in the analysis. Nulliparous group: 2 in nitroglycerine group excluded: 1 had pregnancy‐induced hypertension, 1 had a decreased AFI. Nulliparous group: 2 in placebo group excluded: 1 cephalic, 1 lost to follow‐up. Multiparous group: 2 in nitroglycerine group excluded: had cephalic presentations. Multiparous group: 1 in placebo group excluded: had cephalic presentation. Data on fetal heart rate abnormalities were available for 61 women in nitroglycerine group and 58 in placebo group; for side effects, the numbers were 59 and 58, respectively.
Selective reporting (reporting bias)	Unclear risk	We did not assess the trial protocol.
Other bias	Low risk	Baseline data: similar for maternal age, GA and anterior placenta in nulliparous and multiparous trials.