A comparison of time-varying covariates in two smoking cessation interventions for cardiac patients

Rilana Prenger, Marcel E. Pieterse, Louise M. A. Braakman-Jansen, Catherine Bolman, Hein de Vries, Loes C. W. Wiggers

The aim of the study was to explore the time-varying contribution of social cognitive determinants of smoking cessation following an intervention on cessation. Secondary analyses were performed on data from two comparable randomized controlled trials on brief smoking cessation interventions for cardiac in- and outpatients. Cox regression with time-varying covariates was applied to examine the predictive cognitions for smoking cessation over time. Both samples showed self-efficacy and intention-to-quit to be strong time-varying indicators of smoking cessation during the full 1-year follow-up period, and during the post-treatment phase in particular. Less consistently, time-varying cons of quitting and social influence were also found to be associated with smoking cessation, depending on the sample and type of intervention. Self-efficacy and intention-to-quit were the major covariates and positively related to smoking cessation over time among cardiac patients, in line with social-cognitive theories. Interestingly, both cognitive constructs appeared to act with some delay. Apparently, smoking cessation is a lengthy process in which the interplay between self-efficacy (and intention indirectly) and quitting behavior will largely determine long-term maintenance of abstinence. The presented time-varying analyses seem a valid and feasible way to underpin trajectories of cognitions in datasets with a limited number of time intervals.