Table 3. Impact of smoke-free legislation on primary and secondary outcomes.

					Sensitivity analyses
	Primary analysis				Model 1				Model 2
	N	OR	95%CI	P-value	N	OR	95%CI	P-value	N	OR	95%CI	P-value
	Primary outcomes
Low birth weight	9,933,349	0.961	0.949–0.974	<0.001	5,822,837	0.980	0.962–0.998	0.032	9,933,349	0.963	0.951–0.976	<0.001
Stillbirth	9,984,278	0.922	0.881–0.965	<0.001	5,850,909	0.910	0.855–0.970	0.003	9,984,278	0.919	0.878–0.962	<0.001
Neonatal mortality	9,933,349	0.924	0.883–0.966	0.001	5,822,837	0.932	0.877–0.991	0.025	9,933,349	0.924	0.883–0.966	0.001
SIDS	9,933,349	1.018	0.916–1.132	0.735	5,822,837	0.990	0.808–1.213	0.924	9,933,349	1.018	0.916–1.132	0.735
Secondary outcomes
Very low birth weight	9,933,349	1.010	0.978–1.042	0.558
Early neonatal mortality	9,933,349	0.958	0.890–1.032	0.258
Late neonatal mortality	9,911,272	0.863	0.793–0.940	0.001
Post-neonatal mortality	9,904,292	0.954	0.900–1.010	0.106
Infant mortality	9,933,349	0.937	0.904–0.971	<0.001

Odds ratios indicate odds of developing outcome in period after versus period before July 2007, when smoke-free legislation was introduced. Primary models are adjusted for non-linear underlying time trends (via B-splines), month, maternal age, maternal marital status, sex, socioeconomic status, region, and urbanisation level and based on individual-level analysis of complete cases. Mortality models are furthermore adjusted for birth weight. For sensitivity analyses, model 1 is the complete case model with additional adjustment for parity, whereas in model 2 missing data for parity are imputed. OR = odds ratio; CI = confidence interval; SIDS = sudden infant death syndrome.