Daly 2016.

Methods	Country: Australia Setting: community well‐child health clinic Type: RCT
Participants	1424 parents of children aged 0 to 4 years attending well‐child health checks
Interventions	Interventions: Arm 1: • Computer‐delivered care ‐ tailored on‐screen information and a printed self‐help report regarding the risks of infant SHS exposure, how to reduce exposure risk, advice on quitting smoking, and contact details of the free quit line • Child health nurse‐delivered care ‐ During the subsequent clinic consultation, nurses provided a brief intervention focussing on risk reduction for the infant and offering NRT to parents/carers who were smokers. Contact details of the quit line were again provided, and nurses discussed the importance of complete home smoking bans, providing advice to address any barriers to their implementation. Arm 2: • Same as above, plus infant urine cotinine measured and results shared with parent, child health nurse, and their GP. A guide to preventing infant SHS exposure and strategies for quitting smoking were also included. Control: • Usual care from child health nurses
Outcomes	Child exposure: Primary outcome: Parent/carer reported infant exposure to SHS, defined as a person smoking in the infant's presence in the past 3 days. At 12‐month follow‐up, if parent/caregiver reported the infant as NOT exposed, this was validated with urine cotinine test. Secondary outcomes: parent/caregiver smoking status and household smoking ban status of the home Target behavioural change: proportion who quit and proportion with complete household smoking ban
Type of intervention	Well‐child (child health check)
Notes	Conflict of interest: unclear Source of funding: Financial Markets Foundation for Children, Community Health and Anti Tuberculosis Association, Centre for Health Research & Psycho‐oncology (CHeRP and infrastructure support from the Hunter Medical Research Institute)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Clinics were randomly assigned to 1 of 2 treatment arms or to a control arm via random number function in SAS statistical software.
Allocation concealment (selection bias)	Unclear risk	Services not blind to study allocation but unclear about allocation concealment
Incomplete outcome data (attrition bias) All outcomes	Low risk	Between 11% and 15% lost to follow‐up or declined to participate at 12‐month follow‐up
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Objective measure (cotinine)
Other bias	High risk	• Variability in quality and consistency of advice given to parents/caregivers to access NRT may bias the effect estimate towards the null. • Exposure misclassification; non‐smoking parents/caregivers had partners who smoked and this was not measured. Furthermore, self‐reported SHS exposure was not validated at baseline assessment. • Not blinded, meaning prone to detection and performance bias