Methods	4-armed cluster-randomised trial (2×2) to support women to stop smoking in pregnancy and breastfeed postpartum Study conducted in the lower North Island, New Zealand, with recruitment from June 1999 to September 2000
Participants	Inclusion criteria: The midwifery team was the unit of randomisation, which were stratified by locality and randomised into 1 of 4 groups. All midwives in selected localities in the lower north island were invited to take part. Midwives asked all pregnant women who had smoked at the time they conceived to take part in the study Exclusion criteria: Not further specified. Recruitment: 93/121 (77%) midwives invited (from 62 midwifery teams), agreed to participate, and were randomised into 1 of 4 study arms (C = 23,I1 = 22,I2 = 22, I3 = 26). 61 midwives recruited women to the study (76%). 46/349 (13%) women approached declined to take part in the study, 6 were ineligible, and 297 were recruited (C=60, I1=60, I2=69, I3=108) Baseline characteristics: Partner smoking (C = 50%, I1 = 47%, I2 = 62%, I3 = 49%). Mean age: C = 24.9, I1 = 26.1, I2 = 27.3, I3 = 25.1. Maori: C = 42%. I1 = 36%. I2 = 20%, I3 = 27%. Over 50% in receipt of community services card. Progress+ coding: Low SES.
Interventions	Intervention developed with provider input and detailed discussion of provider views included Control: ‘Usual’ maternity care from a midwife, which ranged from asking about smoking, giving advice to quit and to providing more detailed smoking-cessation advice Intervention 1 (smoking education): Midwife training to implement education and support for smoking cessation and reduction Intervention 2 (breastfeeding): Midwife training and support to implement education and support for breastfeeding for women who smoked Intervention 3 (combined): Midwife training to implement smoking education and breastfeeding programmes Smoking education included motivational interviewing provided by a midwife (who was allocated an extra funded visit and given 4 hours training with a counsellor), flip-chart, video-tape Main intervention strategy: Counselling (single intervention) compared to usual care. Groups 1 and 3 compared to groups 2 and 4 in this review Intensity: Frequency (C = 0, I = 2), Duration (C = 0, I = 2). Usual care intensity: F = 1, D = 1 Intervention provided by existing staff (midwives): Effectiveness study
Outcomes	Biochemically validated smoking cessation at 28 and 36 weeks’ gestation* (late pregnancy), and 6 weeks and 4 months postpartum* (0-5 months postpartum). Smoking reduction outcomes of self-reported ‘cut down a little’ or ‘cut down significantly’ are not included in this review as outcomes unclear Breastfeeding outcomes also reported.
Notes	Design effect for clustering reported, so outcome figures used
Risk of bias
Bias	Authors’ judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random sequence generation using excel for each stratum.
Allocation concealment (selection bias)	Low risk	Group allocation by external statistician.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Missing data for most outcomes, 28% attrition for 4 month postnatal follow-up. Only women who moved from the area were excluded from analysis in this review
Selective reporting (reporting bias)	Unclear risk	Smoking status only reported.
Other bias	Low risk	No other bias detected.
Biochemical validation of smoking abstinence (detection bias)	High risk	Serum cotinine samples provided by 108 women. 17/19 self-reported non-smokers had cotinine levels consistent with non-smoking, but outcomes not adjusted for misclassification. 15 ng/mL cut-off level
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not possible to blind midwives to allocation group. Women were not aware of midwife group allocation
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not reported.
Incomplete implementation	Unclear risk	There were problems with some midwives not recruiting any women to the study, but the degree of implementation among those women recruited is not reported
Equal baseline characteristics in study arms	High risk	When compared with control group, women in the smoking group were older and less likely to be Maori. Also the number of women recruited to the combined group was much larger than the other groups, which suggests potential issues with recruitment
Contamination of control group	Unclear risk	Cluster-study design to avoid contamination.