Better understanding the process of smoking relapse is key to reducing the high proportion of cessation attempts, more than 90%, lasting less than one year. Even modest reductions in relapse rates could have a dramatic impact, as 45% of smokers make a quit attempt each year. Even though there is growing evidence that multiple levels, from genetic factors to social policies, impact health behaviors, there has yet to be a multi- level analysis of factors impacting smoking relapse. The purpose of this study is to apply multi-level modeling to data from a nationwide sample in order to determine factors at the individual, neighborhood, and regional level that impact the likelihood of relapse. To achieve this goal, the project will merge publically available data from multiple public and non-profit sources with outcome and demographic data from the Technology Enhanced Quitline (TEQ) study, a randomized trial testing the use of an automated telephone system to prevent smoking relapse in users of tobacco quitline services. Data will be merged using the open-source geographic information systems (GIS) software suites GRASS and QuantumGIS. This combined dataset will be analyzed to determine 1) the complex interplay of factors across multiple levels impact participants'smoking behaviors, controlling for intervention exposure, and 2) statistical interactions between multi-level factors and intervention efficacy to determine if consideration of such factors may be useful in targeting similar interventions in the future. Non-linear mixed models will be used to account for the hierarchical structure of the data. These models will provide fixed effects of covariates, and random effects (with a compound symmetry covariance structure) to account for correlated data nested within multiple levels. The project brings together an interdisciplinary team including a geographer, an epidemiologist, a biostatistician, and a nurse researcher specializing in tobacco control and consulting expertise from experts in multilevel modeling and tobacco policy. This study will provide deeper understanding of the impact and interplay of multilevel factors on smoking relapse. Such insights will guide policymakers and facilitate the translation of cessation interventions by moving beyond a reductionist approach and illuminating the regional and neighborhood contexts that affect cessation outcomes, while secondarily providing a framework for applying multilevel modeling to other cancer control problems.
The purpose of this study is to identify environmental and social factors contributing to smoking relapse, which remains the most common outcome of cessation attempts. Identifying the neighborhood and regional factors associated with relapse will lead to new approaches for the treatment of nicotine dependence: helping policymakers to target limited funding for cessation more efficiently and to create a policy environment that supports smokers who are trying to quit.