Generalized estimating equations for modeling cluster randomized trial data on smoking cessation among tuberculosis patients.

• Published in: PLoS ONE
• Main Authors: Vasantha Mahalingam, Ratnakar Singh, Ramesh Kumar Santhanakrishnan, Adhin Bhaskar, Ponnuraja Chinnaiyan
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2025-01-01
• Online Access: https://doi.org/10.1371/journal.pone.0333992

There is a paucity of studies applying Generalized Estimating Equations (GEE) for longitudinal analysis of smoking cessation outcomes within the framework of a cluster randomized trial, especially among tuberculosis (TB) patients. In this study, a GEE model which accounts for repeated measures and cluster-level effects was implemented to identify factors associated with smoking cessation among TB patients. The data included 375 TB patients who were smokers and given TB treatment during 2013-2016 in Kanchipuram and Villupuram districts under a cluster randomized trial. GEE modeling provided robust, population-averaged estimates while accounting for intra-cluster correlation, confirming the sustained impact of these interventions. The model demonstrated that smoking cessation interventions, when integrated with TB treatment, had an impact on cessation outcomes in these populations.