Advancing Tobacco Research by Integrating Systems Science and Mixture Models

描述（由申请人提供）：在美国，吸烟是导致疾病、残疾和死亡的主要可预防原因，但大约五分之一的成年人吸烟。 在每年尝试戒烟的大约1500万吸烟者中，即使有戒烟辅助工具，绝大多数最终还是会复发。虽然对吸烟依赖的病因学了解很多，但仍有大量工作要做，以有效地帮助吸烟者戒烟，并最终预防与吸烟有关的死亡，最常见的是由于癌症和疾病。戒烟是在各种相互关联的个人和环境因素的背景下发生的，其中许多因素在戒烟后的头几周内迅速变化。我们提出了两个领域的科学研究，以大大改善戒烟的结果。首先，更好地了解戒烟过程中展现的复杂系统动力学将指导临床医生制定干预措施，随着时间的推移适应个人不断变化的需求和对特定治疗的反应。其次，对具有不同特征的个体的差异治疗效果的更全面的科学理解将指导临床医生选择对不同类型的个体最有希望的治疗方法。该项目的总体目标是通过整合新的系统科学方法、时变效应模型和混合模型来进一步推进戒烟科学，并将新方法应用于烟草使用的生态瞬时评估（EMA）数据分析。本项目的具体目标是：（1）建立戒烟经历与戒烟里程碑之间的关系（失效和复发），并检查治疗条件、基线特征和随时间变化的协变量的影响;（2）检查反映基线因素关键组合的潜在个体亚组之间的差异治疗效果;（3）确定潜在的亚组，其特征是在戒烟尝试过程中发生的独特动态过程;（4）促进和促进烟草研究人员对这些创新统计方法的吸收。拟议项目的结果将为构建高效的戒烟干预措施提供信息，这些干预措施（1）适合个人，（2）随着时间的推移适应参与者的反应。重要的是，该项目的总体影响远远超出了拟议的分析集;该项目将通过向烟草研究人员快速、有计划地传播重要的新分析方法和设计考虑因素，以持续、有力的方式加快戒烟研究的步伐。 公共卫生相关性：临床试验数据收集的最新进展是了解戒烟尝试复杂动态的关键。该项目将采用创新的系统科学分析方法和混合模型，从戒烟试验中获得密集的纵向数据。该项目将有助于制定针对个人的吸烟干预措施，随着时间的推移适应个人的需要，并最终减少与吸烟有关的发病率和死亡率。

DESCRIPTION (provided by applicant): Smoking is the leading preventable cause of disease, disability, and death in the United States, but approximately one-fifth of adults smoke cigarettes. Among the approximately 15 million smokers who make a quit attempt every year, the great majority eventually relapse even with smoking cessation aids. While much is known about the etiology of smoking dependence, substantial work remains to effectively help smokers quit and ultimately prevent smoking-related death, most commonly due to cancer, and disease. Smoking cessation occurs within the context of a wide variety of interrelated individual and environmental factors, many of which change rapidly during the first few weeks after quitting. We propose two areas of scientific inquiry to substantially improve smoking cessation outcomes. First, a better understanding of the complex system dynamics that unfold during the smoking cessation process will guide clinicians in the development of interventions that adapt over time to individuals' changing needs and response to particular treatments. Second, a more thorough scientific understanding of differential treatment effects for individuals with different profiles t baseline will guide clinicians in selection of treatments that hold the most promise for different types of individuals. The overall goal of this project is to further the science of smoking cessation by integrating a novel systems-science approach, time-varying effect models, and mixture models, and apply the new approach to analysis of ecological momentary assessment (EMA) data on tobacco use. The specific aims of this project are (1) To establish the relation between the experience of withdrawal over time and survival to smoking cessation milestones (lapse and relapse), and examine the impact of treatment condition, baseline characteristics, and time-varying covariates; (2) To examine differential treatment effects across latent subgroups of individuals reflecting key combinations of baseline factors; (3) To identify latent subgroups characterized by unique dynamic processes occurring during a smoking cessation attempt; and (4) To promote and facilitate uptake of these innovative statistical approaches by tobacco researchers. Results from the proposed project will inform the construction of highly effective smoking cessation interventions that (1) are tailored to the individual and (2) adapt to participant response over time. Importantly, the overall impact of this project extends far beyond the proposed set of analyses; this project will accelerate the pace of smoking cessation research in a sustained, powerful way through rapid, programmatic dissemination of important new analytic methods and design considerations to tobacco researchers. PUBLIC HEALTH RELEVANCE: Recent advances in data collection for clinical trials hold the key to understanding the complex dynamics of smoking quit attempts. This project will apply an innovative systems-science analytic approach and mixture models to intensive longitudinal data from a smoking cessation trial. This project will enable the construction of smoking interventions that are tailored to individuals, adapt to individuals' needs over time, and ultimately reduce smoking-related morbidity and mortality.