Changes in adolescent smoking with implementation of local smoke-free policies in Indonesia: Quasi-experimental repeat cross-sectional analysis of national surveys of 2007 and 2013

Background: Banning smoking from public places may prevent adolescent smoking, but there is little evidence on impact of smoke-free policies (SFPs) from low and middle-income countries. This study assessed to what extent the adoption of local SPFs in Indonesia between 2007 and 2013 associated with adolescent smoking. Methods: Data on 239,170 adolescents (12–17 years old) were derived from the 2007 and 2013 national health surveys in 497 districts and 33 provinces. This study compared 2013 survey respondents living in districts/provinces that adopted SFPs between 2007 and 2013, with 2007 respondents and 2013 respondents in districts/provinces that did not adopt policies. Multilevel logistic regression analysis assessed whether SFP was associated with daily and non-daily smoking. We controlled for survey year, SFP in 2007, socio-demographics, and district characteristics. Results: Strong district SFPs was significantly associated with lower odds of daily smoking (OR:0.81, 95 %CI:0.69-0.97), but non-significantly with non-daily smoking (OR:0.89, 95 %CI:0.76-1.05). Strong provincial SFPs was not associated with daily smoking (OR:1.02, 95 %CI:0.84–1.25), but was associated with higher odds of non-daily smoking (OR:1.22, 95 %CI:0.99-1.51). Moderately strong SFPs did not consistently show associations in the same direction. For example, moderately strong provincial SFP was associated with higher odds of daily smoking (OR:1.27, 95 %CI:1.11-1.46) and lower odds of non-daily smoking (OR:0.82, 95 %CI:0.72-0.93). Conclusion: We did not detect a consistent short-term effect of district and province-level smoke-free policies on adolescent smoking in Indonesia. Weak implementation and poor compliance may compromise effectiveness, which would call for improvement of SFP implementation in Indonesia.