The effect of proactively identifying smokers and offering smoking cessation support in primary care populations: a cluster-randomized trial.

Aims: To establish whether proactively identifying all smokers in primary care populations and offering smoking cessation support is effective in increasing long-term abstinence from smoking.

Design: Cluster randomized controlled trial.

Setting: Twenty-four general practices in Nottinghamshire, randomized by practice to active or control intervention.

Participants: All adult patients registered with the practices who returned a questionnaire confirming that they were current smokers (n = 6856).

Intervention: Participants were offered smoking cessation support by letter and those interested in receiving it were contacted and referred into National Health Service (NHS) stop smoking services if required.

Measurements: Validated abstinence from smoking, use of smoking cessation services and number of quit attempts in continuing smokers at 6 months.

Findings: Smokers in the intervention group were more likely than controls to report that they had used local cessation services during the study period [16.6% and 8.9%, respectively, adjusted odds ratio (OR) 2.09, 95% confidence interval (CI) 1.57-2.78], and continuing smokers (in the intervention group) were more likely to have made a quit attempt in the last 6 months (37.4% and 33.3%, respectively, adjusted OR 1.23, 95% CI 1.01-1.51). Validated point prevalence abstinence from smoking at 6 months was higher in the intervention than the control groups (3.5% and 2.5%, respectively) but the difference was not statistically significant (adjusted OR controlling for covariates: 1.64, 95% CI 0.92-2.89).

Conclusions: Proactively identifying smokers who want to quit in primary care populations, and referring them to a cessation service, increased contacts with cessation services and the number of quit attempts. We were unable to detect a significant effect on long-term cessation rates, but the study was not powered to detect the kind of difference that might be expected.