Effect of smoke-free policies in outdoor areas and private places on children's tobacco smoke exposure and respiratory health

Background: Smoke-free policies in outdoor areas and semi-private and private places (eg, cars) might reduce the health harms caused by tobacco smoke exposure (TSE). We aimed to investigate the effect of smoke-free policies covering outdoor areas or semi-private and private places on TSE and respiratory health in children, to inform policy. Methods: In this systematic review and meta-analysis, we searched 13 electronic databases from date of inception to Jan 29, 2021, for published studies that assessed the effects of smoke-free policies in outdoor areas or semi-private or private places on TSE, respiratory health outcomes, or both, in children. Non-randomised and randomised trials, interrupted time series, and controlled before–after studies, without restrictions to the observational period, publication date, or language, were eligible for the main analysis. Two reviewers independently extracted data, including adjusted test statistics from each study using a prespecified form, and assessed risk of bias for effect estimates from each study using the Risk of Bias in Non-Randomised Studies of Interventions tool. Primary outcomes were TSE in places covered by the policy, unplanned hospital attendance for wheezing or asthma, and unplanned hospital attendance for respiratory tract infections, in children younger than 17 years. Random-effects meta-analyses were done when at least two studies evaluated policies that regulated smoking in similar places and reported on the same outcome. This study is registered with PROSPERO, CRD42020190563. Findings: We identified 5745 records and assessed 204 full-text articles for eligibility, of which 11 studies met the inclusion criteria and were included in the qualitative synthesis. Of these studies, seven fit prespecified robustness criteria as recommended by the Cochrane Effective Practice and Organization of Care group, assessing smoke-free cars (n=5), schools (n=1), and a comprehensive policy covering multiple areas (n=1). Risk of bias was low in three studies, moderate in three, and critical in one. In the meta-analysis of ten effect estimates from four studies, smoke-free car policies were associated with an immediate TSE reduction in cars (risk ratio 0·69, 95% CI 0·55–0·87; 161 466 participants); heterogeneity was substantial (I2 80·7%; p