Estimating the Effect of the Comprehensive Smoke-Free Legislation on Neonatal and Infant Mortality in Middle-Income Countries Using a Synthetic Control Method

Márta K. Radó , Erasmus University Medical Centre, Rotterdam
Frank Van Lenthe, Erasmus Medical Centre, Rotterdam
Aziz Sheikh, The University of Edinburgh
Jasper V. Been, Erasmus University Medical Centre, Rotterdam

BACKGROUND: Increasing evidence supports the positive effects of smoke-free legislation on infant health. However, such studies are almost entirely restricted to high-income countries. We estimated the impact of the comprehensive smoke-free legislation on neonatal and infant mortality in middle-income countries. METHODS: We obtained country-level panel data from the World Bank and Penn World datasets covering the 2001-2017 period. We used the synthetic control method, which reduces the probability of biased estimations. For each country with smoke-free legislation, we constructed a synthetic control country that was the combination of ‘control’ middle-income countries without smoke-free legislation but with similar temporal trends in the outcome and confounding variables in the pre-legislation period. The legislation effects for the post-legislation periods were derived from the difference between temporal outcome patterns of countries with smoke-free legislation and that of their synthetic control country. We compared the legislation effects to the placebo effects obtained by fictitiously assuming that smoke-free legislation occurred at random in the observed countries. We aggregated the country-specific effects to estimate the average legislation effect. FINDINGS: Neonatal and infant mortality rates decreased substantively following smoke-free legislation in a number of countries relative to their synthetic control country. INTERPRETATION: Implementation of comprehensive smoke-free legislation was associated with reductions in infant mortality in some middle-income countries. Expanding smoke-free policies to middle-income countries along with addressing the local barriers and facilitators of this legislation could avert substantial numbers of early-life deaths. The synthetic control method is a promising approach to assessing policies in population health research.

See extended abstract

 Presented in Session 47. Linking Policies, Health and Mortality