Abstract
INTRODUCTION: A recent meta-analysis of Internet interventions for smoking cessation found mixed evidence regarding effectiveness. One explanation may be differential use of non-assigned cessation treatments-including other Internet programs-that either amplify or mask study intervention effects. We examined the impact of non-assigned treatment use on cessation outcomes in The iQUITT Study, a randomized trial of Internet and telephone treatment for smoking cessation. METHODS: Participants were randomized to a basic Internet (BI) comparison condition (N = 675), enhanced Internet (EI: N = 651), or EI plus telephone counseling (EI+P: N = 679). The primary outcome was 30-day point prevalence abstinence (ppa) at 3 and 6 months. Assigned intervention use was assessed with automated tracking data. Assessment of non-assigned treatments included pharmacotherapy, behavioral, alternative, and non-study Internet treatments. Univariate and multivariate logistic regression models examined whether non-assigned treatment use was associated with 30-day ppa. RESULTS: About 70% of participants used at least one non-assigned treatment. A higher rate of non-study Internet treatment among BI participants was the only treatment group difference at both 3 and 6 months. Multivariate models controlling for condition and baseline predictors of non-assigned treatment use showed that high-intensity non-study Internet treatment was positively associated with 30-day ppa at 3 and 6 months, and pharmacotherapy and behavioral treatment use was negatively associated with 30-day ppa at 6 months. CONCLUSIONS: Non-assigned treatment use is an important factor to consider when evaluating Internet cessation interventions. Results highlight methodological issues in selecting a comparison condition. Researchers should report non-assigned treatment use alongside main trial outcomes.