A systematic review and meta-analysis of comprehensive community initiatives on tobacco use and related harms

Multi-system approaches that coordinate initiatives across whole communities are a prioritised response to the known global burden attributed to tobacco use, including smokeless tobacco and second-hand smoke exposure. However, the impact of these comprehensive community initiatives (or CCI) is unclear, and moderators of impact are rarely considered. We explored CCIs that address tobacco use or harms reported to be related to tobacco use including approaches, effectiveness and factors that strengthen impact. A systematic review and meta-analysis (PROSPERO ID CRD420251044005). Embase, Medline, CINAHL, PsycINFO, and the Cochrane Database of Systematic Reviews were searched. Reference lists from included articles were hand-searched. A meta-analysis and/or qualitative synthesis were completed on any measure of tobacco use or related harms including potential moderators of impact (including community type, level of engagement, initiative combinations, duration, recency, evaluation quality, and length of follow-up). The systematic review included 40 studies (from 15,117 articles), with 39 assessing smoking, five assessing smokeless tobacco, five assessing second-hand smoke and two assessing tobacco-related harms to health. Meta-analysis included 25 outcome estimates from 17 studies (comparing CCI with no or minimal intervention) across three tobacco outcomes: any smoking in the past month (12 studies, n = 161,245) relative risk (RR) = 0.88 (95% confidence interval [CI] = 0.82–0.94), any at least weekly smoking (four studies, n = 4,569) RR = 0.69 (95% CI = 0.46–1.03), and any use in lifetime (six studies, n = 9,677) RR = 0.92 (95% CI = 0.84–1.00). Data were insufficient for meta-analysis of additional smoking (e.g., frequency) or smokeless use, second-hand smoke or related harms outcomes. The pooled effect sizes indicated significant impact on recent smoking; however, this was in the context of substantial heterogeneity (I2 = 62.0–78.4%; τ 2 = 0.0064–0.1115) across low-quality evaluations that were conducted more than ten years ago. Subgroup and narrative analysis provided support for CCIs that involve policies relating to cigarette price supported by media campaign and school-based and parenting-related activities. Conclusions on the impact of CCI on tobacco use remain tentative though all estimates indicated positive effect. Notably, CCIs were most effective at reducing the proportion of people who already smoke.