Utilizing tobacco discrete choice experiments to predict the population impacts of FDA regulatory policies

点击翻译按钮获取中文摘要

PROJECT SUMMARY As the state-of-the-art method to assess consumers’ preferences for new products, discrete choice experiments (DCEs) have been increasingly applied to study the impacts of potential policies on tobacco use behaviors, particularly in the areas of product standards and messaging development (e.g., warnings and campaigns). Compared to observational survey data, DCEs have the advantage of providing nimble evidence on emerging products with low prevalence and predicting the impacts of potential policies on behaviors before they are implemented in the real world. However, their potential in guiding and defending FDA policies has not been fully realized due to several methodological challenges. The first challenge is the lack of consistency in tobacco DCE design and reporting. Researchers often adopt different design features (e.g., whether to include an opt-out option) and use different measures in reporting, such as willingness to pay (WTP) vs. relative importance scores. Therefore, it is important to review current practices, explore how to synthesize existing DCE evidence, and propose guidelines for future DCE design and reporting. The second challenge is hypothetical bias – the bias between stated preferences (SP) or “what people claim that they would do” and revealed preferences (RP) or “what people actually do in the real world.” Recent method development has combined DCEs and RP data (e.g., retail and survey data) to mitigate hypothetical bias, providing us a unique opportunity to revisit existing DCEs and improve their accuracy in policy impact prediction. The overarching goal of this project is to inform the impact of FDA regulatory policies on tobacco use behaviors using existing DCE evidence. We will conduct systematic reviews to synthesize evidence on product standards and messaging development by assessing the ranking of policy importance and WTP. Moreover, we will combine existing DCEs with Nielsen Retail Scanner data and nationally representative surveys to mitigate the hypothetical bias in DCEs, which improves the accuracy of using existing data to predict the impact of potential regulatory policies that have not been implemented yet. The specific aims are to 1) Conduct a systematic review of tobacco DCEs manipulating product attributes. 2) Conduct a systematic review of tobacco DCEs manipulating product messaging. 3) Combine stated-preference DCEs and revealed-preference real-world retail and survey data to predict the impacts of FDA regulatory policies on product market shares and tobacco use. This project will provide the first systematic review of tobacco DCEs in the context of informing FDA regulatory policies, further improving the accuracy of DCE in predicting policy impacts. In summary, this project uses existing data to address priority scientific domains of RFA-OD-21-003, including tobacco product characteristics, communication practices, and tobacco use patterns.