Dickson‐Spillmann 2013.

Methods	Setting: Swiss Institute for Research in Public Health and Addictions, which is associated with the University of Zurich, Switzerland. Study period: April 2011 to February 2012. Recruitment: through advertisements in online and print newspapers. Sample size calculation: described for initial planned simple RCT and then redone as authors realized the study is a cluster‐RCT.
Participants	Number of participants: 223 smokers (hypnosis: 116; relaxation: 107). Inclusion criteria: current smokers prepared to stop smoking; smoking ≥ 5 cigarettes per day; aged 18 to 65 years; not using other cessation aids; understand and speak German; no history of psychotic disorders; not be intoxicated before or during the intervention. Exclusion criteria: using other cessation aids; history of psychotic disorders; intoxicated before or during the intervention. Overall demographics: women (47%); mean age = 37.5 (SD 11.8); Swiss (86.1%) years; mean FTND = 4.7.
Interventions	a) Hypnosis b) Relaxation Both groups received a 40‐minute psychoeducation session before their intervention, a 20‐minute debriefing after the intervention, and a CD for use at home. Both interventions comprised a single 40‐minute session of group treatment. Participants were required to pay CHF 40 to participate in the study.
Outcomes	Definition of smoking cessation: biochemically validated (saliva cotinine < 5 ng/mL undertaken by post) 30‐day point prevalence of smoking abstinence at a 6‐month follow up. Adverse events: collected at 2‐week telephone follow‐up interview and at 6‐month postal follow‐up questionnaire.
Notes	Funding: Swiss Tobacco Prevention Fund. Author declaration of interest: "The authors declare that they have no competing interests". The study was initially planned as an individually‐randomized RCT but analysis was changed to that suitable for a cluster‐RCT when it was realized that the unit of recruitment was clusters (e.g. groups of work colleagues). Cluster randomization was accounted for in the analysis.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random sequence of 20 sessions generated using an online programme with ratio of 1:1 for number of sessions per intervention; one additional last session was randomly allocated by the same programme. Does not state if a block size was used or not
Allocation concealment (selection bias)	Low risk	Therapist blinded until delivery of intervention; therapist was informed by text immediately prior as to which intervention should be delivered.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Biochemically validated so differential misreport judged unlikely
Incomplete outcome data (attrition bias) All outcomes	Low risk	Those with absent saliva sample (n = 1), and those with saliva cotinine higher than usual for occasional smokers (n = 3), were classed as smokers. At 6 months, LTFU for the hypnosis and relaxation groups was 14.7% and 17.1%, respectively.
Selective reporting (reporting bias)	Unclear risk	Primary outcome variable reported and information provided on numbers without biochemical validation. However, 2‐week, but not 6‐month, AE data reported and reason for this omission is unclear.