Joseph 2014.

Methods	Country: USA Setting: community (well‐child clinic) Type: observational, quasi‐experimental (historical control)
Participants	Parents who smoke who have children aged 12 and 24 months; 40 parent‐child couples for intervention group and 40 for control group
Interventions	Intervention: Children had serum cotinine measured with lead screening. Lab results were sent to providers and parents. The letter included an explanation that cotinine came from tobacco exposure, and that the normal value was zero. One week later, the tobacco counsellor proactively telephoned to explain the lab result, to describe potential sources of tobacco smoke exposure, including third‐hand smoke, and to convey what is known about the potential health effects of exposure for their child. Counsellor used motivational interviewing and cognitive‐behavioural therapy to engage the parent in a smoking cessation attempt. All parents were encouraged to institute a strict home and car no‐smoking policy, regardless of whether they wanted to stop smoking. If parent wanted to stop smoking, counsellors offered an 8‐session weekly telephone intervention based on an evidence‐based telephone smoking cessation protocol. While no prescription or over‐the‐counter medicine was offered, counsellors did describe them as options and facilitated access where requested. Control: historical group that received usual care
Outcomes	Child exposure: outcomes assessed 8 weeks after initial call, including receipt of tobacco treatment, quit attempts, 7‐day point prevalent abstinence, and current home and car smoking policies Target behaviour change: receipt of tobacco treatment, parent quit attempts, 7‐day point prevalence abstinence
Type of intervention	Well‐child (child health check)
Notes	Conflict of interest: unclear Source of funding: National Cancer Institute (R21CA137014)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Not randomised
Allocation concealment (selection bias)	High risk	Allocation concealment not possible, as the study was not randomised
Incomplete outcome data (attrition bias) All outcomes	Low risk	95% followed up
Blinding of outcome assessment (detection bias) All outcomes	High risk	Not blinded
Other bias	High risk	• Recall bias as data ascertained historically in the comparison group • Misclassification, as smoking status not biochemically validated in control group