OP96 Simulating the impacts of potential spatial restrictions on tobacco retail availability in megacities of China, a case study of Shanghai

Background China is the world’s largest manufacturer of tobacco, with the largest smoking population worldwide. Over 1 million deaths (2010) and $45.28 billion in medical expenditures (2015) are attributed to smoking. Increasing international evidence suggests regulating tobacco retail availability (TRA) could provide new opportunities for effective tobacco control. The effectiveness of different policy options designed to reduce TRA is unknown in China. Using the case study of Shanghai, the biggest mega-city in China with a developed tobacco retail market, our study aims to simulate the potential impacts of introducing spatial restrictions on the availability of tobacco, including differences by neighbourhood deprivation and urbanity. Methods A total of 19413 locations of tobacco retailers in Shanghai were mapped (2019) as the baseline. Under four major restriction types (capping, ban of sales, minimum spacing, and school buffer), our study simulated the potential impacts of 12 simulation scenarios on TRA reduction and the social inequality in TRA. Their impacts on TRA reduction were measured by the percentage reduction in population-weighted Kernel Density Estimation (KDE) across neighbourhoods against the baseline using the median. After testing the statistical differences of TRA across neighbourhood deprivation quintiles using the Kruskal-Wallis test, the impact on equity was examined using the effect sizes of the association between TRA and neighbourhood deprivation. All analyses were further stratified by 3 levels of urbanity (urban centre vs urban area vs suburb). Results As anticipated, all simulation scenarios have the potential to reduce TRA, with the percentage reduction ranging from 9.54% to 81.16%. Stricter restrictions were associated with a greater reduction in TRA for each restriction type. ‘500m-minimum spacing’ was the scenario that resulted in the greatest reduction in TRA. Apart from capping scenarios, most simulation scenarios reduced TRA inequitably across neighbourhood deprivation quintiles, with TRA statistically differing by neighbourhood deprivation quintile (p < 0.05). The most effective scenario for TRA reduction did not necessarily reduce the social inequality in TRA. Compared with the baseline, ‘500m-minimum spacing’ greatly increased the inequality in TRA although it was the scenario with the greatest TRA reduction. Additionally, the potential impacts of simulation scenarios varied by urbanity level. Conclusion Although spatial restrictions offer potential new policy opportunities addressing tobacco availability, they may also increase social inequality. When developing comprehensive tobacco retail regulations for effective tobacco control, policymakers may consider the various impacts of spatial restrictions.