Relapse is the greatest challenge in the treatment of addictions, including cigarette smoking. Research on relapse episodes has relied solely on retrospective recall; furthermore, relapse research has failed to integrate data on proximal antecedents of relapse episodes with data on individual differences and on contextual variables that predispose people to relapse. The proposed study uses an innovative method of computerized self-monitoring to study smoking relapse. A hand-held self-monitoring computer (SMC) will be used to characterize the baseline smoking patterns of 252 subjects recruited for smoking cessation treatment. Subjects carry the SMC for 2 weeks and record the situational and affective antecedents of smoking. The SMC also prompts subjects randomly to record data when they are not smoking, thus measuring the base-rates of relevant behaviors outside of smoking situations, and estimating the association between these stimuli and smoking. The computer also obtains daily measures of stress and withdrawal (both subjective and objective [cognitive performance]). Once subjects quit smoking, they use the SMC for 4 weeks to record the antecedents and sequelae of episodes of smoking (""""""""lapses"""""""") and strong temptations; the SMC continues to randomly sample situations and to assess stress and withdrawal. Statistical methods include logistic regression and survival analysis. We will examine proximal antecedents of lapses and temptations in comparison to base-rates in randomly-sampled situations. Other analyses will consider the role of background stress and withdrawal in promoting relapse, and their interactions with proximal precipitating stimuli. The prospective design will assess whether the after-effects of relapse crises (""""""""Abstinence Violation Effects"""""""") promote progression towards relapse. Finally, using data collected at baseline, we will examine whether (as conditioning theories predict) individual differences in smoking patterns are mirrored in individual differences in relapse patterns. The study includes methodological analyses: The effects of prolonged monitoring will be measured in a comparison group that undergoes self-monitoring but does not quit smoking. The validity of retrospective relapse interviews, widely-used to collect relapse data, will be assessed relative to real-time self-monitoring data. The findings of the study are expected to be highly relevant for the design of more effective interventions to prevent relapse to cigarette smoking and other drug addictions.
Showing the most recent 10 out of 29 publications
