Grobman 2018.

Study characteristics
Methods	RCT
Participants	Number of women randomised: 6106 Setting: 41 hospitals participating in the Maternal‐Fetal Medicine Units Network of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, USA Study date: women recruited March 2014 through August 2017 Inclusion criteria Low risk, defined as absence of any condition considered to be a maternal or fetal indication for birth before 405 weeks Nulliparous Singleton pregnancy GA: 380 weeks to 386 weeks Reliable information on the length of gestation Vertex presentation with no contraindication to vaginal birth and no caesarean birth planned Exclusion criteria GA at date of first ultrasound > 206 weeks Plan for IOL prior to 405 weeks Plan for caesarean birth or contraindication to labour Breech presentation Signs of labour Fetal demise or known major fetal anomaly Heparin or low‐molecular weight heparin use during the current pregnancy Placenta praevia, accreta, vasa praevia Active vaginal bleeding greater than bloody show ROM Cerclage in current pregnancy Known oligohydramnios, defined as amniotic fluid index < 5 cm or maximal vertical pocket < 2 cm Fetal growth restriction, defined as EFW < 10th percentile Known HIV positivity (due to modified delivery plan) Major maternal medical illness associated with increased risk for adverse pregnancy outcome (e.g. any diabetes mellitus, lupus, any hypertensive disorder, cardiac disease, renal insufficiency) Refusal of blood products Participation in another interventional study that influenced management of labour at delivery or perinatal morbidity or mortality Birth planned elsewhere at a non‐Network site State of cervix: favourable and unfavourable (at the time of randomisation, 63% of the participants had an unfavourable modified Bishop score (i.e. a score < 5))
Interventions	Induction group (n = 3062): women were assigned to undergo IOL at 390 weeks to 394 weeks; with no specific induction protocol mandated. Supplementary Appendix for the manuscript indicated that oxytocin was used, and for women with an unfavourable cervix, cervical ripening was used (actual method left to the discretion of the woman's obstetric provider), in conjunction with or followed by oxytocin; mechanical ripening using a Foley catheter without saline infusion and without concurrent oxytocin was permitted; although not mandated, it was suggested that women should be allowed at least 12 hours in the latent phase after completion of any ripening, ROM, and use of oxytocin before the induction was considered 'failed' (and subsequent caesarean birth) versus EM group (n = 3044): women were asked to forgo elective birth before 405 weeks and to have birth initiated no later than 422 weeks. A specific induction protocol was not mandated for women who underwent induction in either group.
Outcomes	Mother: caesarean section; operative vaginal birth; perineal trauma (third‐ or fourth‐degree perineal laceration); postpartum haemorrhage; breastfeeding (at 4‐8 weeks after birth); length of maternal hospital stay (days, categories including < 2, 2, 3, 4, > 4) Baby: perinatal death (antenatal stillbirth and neonatal death); stillbirth; neonatal death; admission to NICU (intermediate or intensive care unit); neonatal convulsions (seizures); neonatal hypoxic‐ischaemic encephalopathy; pneumonia (confirmed); meconium aspiration syndrome; Apgar score (≤ 3 at 5 minutes); length of neonatal hospital stay (days, categories including < 2, 2, 3, 4, > 4)
Notes	Funding: Eunice Kennedy Shriver National Institute of Child Health and Human Development. Declarations of interest: "Dr. Silver reports receiving consulting fees from Gestavision. No other potential conflict of interest rel avant to this article was reported."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Prepared by an independent data coordinating centre, using the simple urn method, with stratification according to clinical site
Allocation concealment (selection bias)	Low risk	Randomisation sequence prepared by an independent data coordinating centre, and an internet‐based randomisation system was used.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not feasible
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	For neonatal primary outcomes requiring review, "Reviewers were unaware of the trial‐group assignments". Unclear whether any other outcomes were assessed blind, though discussion reported that "because masking was not feasible, ascertainment bias is possible".
Incomplete outcome data (attrition bias) All outcomes	Low risk	3059/3062 women assigned to induction were included in analyses (1 lost to follow‐up; 2 withdrew consent); 3037/3044 women assigned to control included in analyses (2 lost to follow‐up; 5 withdrew consent)
Selective reporting (reporting bias)	Low risk	Comprehensive range of outcomes specified in protocol and reported
Other bias	Low risk	Note: authors noted report that 184/3059 women (6%) in the induction group, and 140/3037 (4.6%) women in the control group did not birth 'per protocol'.