Background: Smoking cessation aids are under-prescribed in primary care. General practitioners (GPs) might prescribe more smoking cessation aids if they focused on the choice between treatments as the default, rather than using motivational interviewing that focuses on whether to make a quit attempt.

Objectives: Explore the self-reported adoption and impressions of shared-decision making (SDM) for smoking cessation treatments using a decision aid amongst GPs who received intervention or control trainings.

Design: Mixed quantitative-qualitative convergent analysis of self-reported behaviours from GPs in a pragmatic, cluster-randomized trial conducted between 2021-2023 in Switzerland and France.

Methods: The intervention training encouraged GPs to offer a choice of smoking cessation aids as the default choice using a decision aid (sustained intervention). The control training presented cessation aids without encouraging GPs to change how they discuss smoking cessation. Implementation was measured using questionnaires before and after the trainings, then at 6- and 12-month follow-up. Self-reported behaviours were quantitatively analysed using logistic regression adjusting for baseline responses. Experiences using SDM and default choice were explored through semi-structured interviews.

Results: A total of 43 GPs completed questionnaires (26 women (60%), age range 30-59 years) and 12 an interview. The intervention group reported more sustained interventions (OR 8.2 (CI 1.9-50)), use of the decision aid (OR 19.7 (CI 4.1-135)), and offer of treatments as the default choice than the control group (OR 6.4 (CI 1.4-40)) at 12-month follow-up. Small sample size bias was addressed by using Firth's methodology. In interviews, GPs in the intervention group highlighted the utility of the decision aid and the importance of adapting the default approach to the context of individual patients.

Conclusion: GPs in the intervention group reported more often initiating treatment discussions with patients who smoke using a decision aid, but not more prescriptions. The default approach was not appropriate for all patients.