Background:
Telephone-based smoking cessation services ('quitlines') are both effective and cost-effective. Knowledge of modifiable baseline factors in real-life settings with heterogeneous participants is essential for the development and improvement of treatment protocols to assist in telephone-based smoking cessation. The aim was to assess self-efficacy as a predictor for abstinence at the 12-month follow-up at the Swedish National Tobacco Quitline (SNTQ).

Methods:
The data were collected from a randomized controlled trial comparing the effectiveness of proactive and reactive service at the SNTQ. Included were 612 clients calling the SNTQ between February 2009 and September 2010. Outcome measures were self-reported point prevalence and 6-month continuous abstinence at the 12-month follow-up. Plausible predictors of smoking cessation were assessed at the first call and in a baseline questionnaire. Self-efficacy was measured by three questions: (1) the likelihood of being smoking free in 1 year; (2) the ability to handle stress and depressive mood without smoking; and (3) the likelihood of using medication against craving if necessary. The associations between predictors and outcome were subjected to logistic regression analysis.

Results:
Of the three self-efficacy predictors for abstinence at month 12, only the perceived ability to handle stress and depressive mood without smoking remained significant in the adjusted analyses (odds ratio, OR, 1.1 for point prevalence and 1.2 for 6-month continuous abstinence). The overall strongest predictor in the adjusted analyses was smoking status in the week before baseline (OR 2.7 for point prevalence and 3.1 for 6-month continuous abstinence).

Conclusions:
The perceived ability to handle stress and depressive mood without smoking at baseline predicted the subjects' abstinence at the 12-month follow-up. An assessment of/adjustment for stress and depressive mood may be appropriate in future smoking cessation research.