Assessing the Potential Effects of a Flavored Tobacco Product Ban to Inform FDA Decision-Making: A systematic review, decision analysis and economic evaluation

Introduction: The purpose of this dissertation is to synthesize and translate research on flavored tobacco in the United States in order to inform regulatory decision-making at the U.S. Food and Drug Administration. Three related manuscripts aim to: 1) synthesize existing research examining the use and attitudes/knowledge/perceptions of flavored tobacco in the United States, specifically focusing on how age and nonflavored tobacco use are correlated with flavored tobacco use; 2) estimate the relative health effects, compared to the status quo, that would be expected to result from a) a ban on combustible flavored tobacco products and b) a ban on all flavored tobacco products, and organize existing relevant research that informs the policy decision; and 3) determine acceptable ceiling costs for the policies considered in the second aim, and identify the minimum effectiveness levels at which these interventions would be considered cost effective and cost-saving. Methods: We used a systematic review to achieve the first aim, and conducted a qualitative synthesis for included studies. To achieve the second aim, we developed a decision tree using TreeAge Pro 2014. This model compared three policy options: 1) the banning of all flavored combustible tobacco products, 2) the banning of all flavored tobacco products (inclusive of combustible and smokeless products), and 3) the status quo, in which only flavored cigarettes are banned. To achieve the third aim, we performed a threshold analyses to establish cost and performance standards for the policy options considered in the second aim. Microsoft Excel 2011 was used for this analysis. Results: The results of the systematic review highlight the association between flavored tobacco use and young age, and suggest that flavoring may promote tobacco use in adolescent and young adult populations. The second manuscript, which examined the policy options among a young adult population, found that a ban on all flavored tobacco products would likely produce the most favorable outcome of the policy options considered. This finding is contingent upon the validity of the model’s underlying structural assumptions, as well as assumptions that determined the model’s parameter values. The results of the cost threshold analysis indicated that the cost of either policy option would need to reach well into the multi-billion dollar range in order for society to deem the cost of either of the policies to be unacceptable. The results of the effectiveness threshold analysis found that both policy options would be cost-effective if less than 1% of the expected number of averted adults smokers were actually averted. As with the model developed for the second manuscript, the parameter values used to conduct the threshold analyses were subject to uncertainty. Conclusions: The results of this dissertation speak to the FDA’s request for information and data to support the establishment of a product standard prohibiting flavors in at least some noncigarette tobacco products. The studies described here can be thought of as exploratory analyses that can be built upon and adapted to directly answer questions posed by the FDA. The model presented in the second manuscript, in particular, could be used to address a range of regulatory options, as it provides a general framework for considering the impact of policies intended to reduce the attractiveness of tobacco products, while taking into consideration the varying levels of risk presented by different products. Additional research should be conducted to inform the development of a more complex model that can provide additional insight into the effects of implementing the policies considered in this dissertation, as well as other regulatory options