Tan 1989.

Methods	RCT.
Participants	Inclusion criteria: breech presentation beyond 33 weeks' gestation without contraindication to ECV. N = 90. Exclusion criteria: signs of growth restriction, vaginal bleeding in the third trimester, toxaemia of pregnancy, labour, polyhydramnios, placenta praevia, previous CS scar, contracted pelvis, fetal malformation and uterine malformation.
Interventions	Intervention 1:tocolytic: salbutamol ‐ beta stimulant (A1) ‐ parenteral. Group 2 received an IVI of salbutamol until maternal heart rate exceeded 100 beats per minute for 30 minutes. N = 30 (nulliparous 17, multiparous 13). Intervention 2:tocolytic: salbutamol ‐ beta stimulant (A1) ‐ oral. Group 1 received salbutamol 4 mg orally 3 times a day for at least 1 day. N = 30 (nulliparous 16, multiparous 14). Comparison:placebo. Group 3 received no salbutamol. N = 30 (nulliparous 17, multiparous 13). Groups 1 and 3 received dummy IV lines
Outcomes	Immediate ECV success.
Notes	Singapore. This study compared 2 different routes of administration (oral and IV) of a tocolytic drug to facilitate ECV. So it provides data only for tocolysis vs placebo, and the different routes of administration are considered in subgroups.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information provided.
Allocation concealment (selection bias)	High risk	2 stacks of randomised cards divided according to parity with each stack further subdivided by a colour code for gestation A or B.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	States that clinicians were blinded to treatment and dummy IVs were inserted. Clinicians did not know parity or gestation. Women's status unclear, but clinicians more likely to be able to influence outcomes.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	States that clinicians were blinded to treatment and dummy IVs were inserted. Clinicians did not know parity or gestation. Women's status unclear, but clinicians more likely to be able to influence outcomes.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No exclusions apparent after randomisation and no loss to follow‐up.
Selective reporting (reporting bias)	Unclear risk	We did not assess the trial protocol.
Other bias	Unclear risk	The 3 groups were similar in placental site, abdominal girth, maternal weight, fetal birthweight and stratification of parity and gestation across groups Failed ECV: Time taken was significantly longer (10.5 + 4.9 vs 5.6 + 3.9; P value < 0.001) and onset of labour was significantly earlier (17.6 + 9.8 vs 25.2 + 14.9 days; P value < 0.02), implying that longer manipulation hastened the onset of labour by 70%.