Evaluating A Brief Computerized Intervention for Smoking Cessation Among People With Severe Mental Illness: Do Negative Beliefs About Smoking Facilitate Change?

Background and Purpose: People with serious mental illnesses (SMI) are three times more likely to smoke than those without mental illness. Most individuals with SMI attempt to quit multiple times in their lifetime, however they are often unsuccessful because they underutilize evidence-based smoking cessation treatment (due to lack of availability and access). Studies suggest that motivational interviewing and education increases motivation to seek treatment for smoking cessation in this population. Interactive websites provide a way to educate and motivate a large number of individuals with minimal professional assistance. We used the Smoking Effects Questionnaire in order to evaluate the relationship between participants’ likes and dislikes about smoking while using an electronic decision support system. We then conducted a four-staged mediation analysis to evaluate whether readiness to quit smoking (after using the intervention) mediated the relationship between likes and dislikes and treatment engagement.

Methods: Participants were recruited at a major mental health center in Chicago, IL. One hundred twenty-eight smokers with severe mental illnesses were assessed at baseline for demographics, smoking characteristics, and smoking related readiness to change. They used the decision support system within 2 weeks of the baseline interview. Two months after use of the decision support system, researchers assessed participants’ smoking behaviors, use of evidence-based cessation treatment, clinician contact, and other quitting behaviors. We used logistic and OLS regression models in a statistical package (SAS 9.2) to evaluate 2 potential mediators between attitudes towards smoking (i.e. Smoking Effects Questionnaire) and engagement in treatment by readiness to quit smoking.

Results:  First, two independent models evaluated the relationship between negative effects of smoking and engagement in treatment (OR=1.04, CI=1-1.08, p=.02) and readiness to change (controlling for baseline readiness) b=.03, t(128)=2.09, p=.03); both models were significant. Second, readiness to change after treatment predicted engagement in smoking cessation treatment (OR=1.77, CI=1.09-2.87, p=.02), while controlling for baseline readiness to change. Finally, the last model predicted engagement in treatment with three covariates: negative effects of smoking, readiness to change at baseline and after the intervention. Only readiness to change after the use of treatment was significantly related to engagement in treatment (OR=1.71, CI=1.05-2.78, p=.03), which indicates that readiness to change after the intervention fully mediates the model.

Conclusions: In the process of using a computerized intervention to engage and education smokers with severe mental illness to quit with an evidence-based treatment, participants responded to a validated questionnaire (the Smoking Effects Questionnaire) that asked about the positives and negatives of smoking. At the end of the intervention, participants who had more strongly endorsed negatives were more likely to want to quit smoking and were more likely to engage in quit smoking treatment.