Predictors of smoking relapse in women with cardiovascular disease in a 30-month study: extended analysis.

PURPOSE: To identify predictors of long-term cessation after intervention for woman hospitalized with cardiovascular diseases (CVD). METHODS: A randomized clinical trial (RCT) with a 30-month follow-up was conducted, recruiting 277 woman smokers with CVD from 10 hospitals. The intervention consisted of inpatient and outpatient counseling on smoking cessation and pharmacological protocol. Psychophysiological factors and time since quitting associated with relapse in the literature were assessed in our prediction model. RESULTS: The nine independent variables for smoking relapse included time since quitting, group assignment, the interaction between time and group assignment, serious quit attempts and five and psychophysiological factors. Lower self-efficacy at baseline was a significant predictor of relapse (OR = 0.98, 95%CI = 0.97,0.99). Another predictor was the interaction between time and group assignment and the finding showed that for the usual care group, increase in time since quitting provided a significant protective factor (OR = 0.69, 95%CI = 0.60,0.79). CONCLUSIONS: Cessation programs should include content on self-efficacy to help prevent relapse. The interaction between time and group warrants further investigation for its prediction for relapse.