Resnicow 1997.

Methods	Setting: predominantly African American community in USA Recruitment: in healthcare, church, and public housing settings; presented as 'health promotion' ‐ not smoking cessation
Participants	650 smokers who completed follow‐up interviews recruited in treatment channels and 504 in control channels (attrition similar between groups) Average age 45 years, average cpd 16
Interventions	∙ Self‐help kit including Kick It guide, video, and aids; bimonthly mailings and single booster telephone call ∙ Health education materials not exclusively addressing smoking, and a cholesterol education video
Outcomes	Point prevalence at 6 months Validation: none
Notes	Less than a third of intervention group received telephone call Post hoc analysis reported significantly higher quit rates amongst call than no call group Multi‐variate analysis controlling for intracluster correlation gives OR of quitting in treatment group as 1.36 (95% CI 0.87 to 2.11) compared to OR 1.42 (95% CI 0.98 to 2.04) from figures used in meta‐analysis
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Cluster‐randomised; stratified by type of site before recruitment of smokers; method of sequence generation not reported
Allocation concealment (selection bias)	High risk	Allocation known at time of recruitment; unclear whether this introduced high risk of bias; all participants received smoking cessation materials
Blinding (performance bias and detection bias) All outcomes	High risk	No biochemical validation and differential levels of contact between groups (including additional phone call); differential misreport judged possible
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition similar between treatment (7.5%) and control (6.8%) conditions Non‐respondents did not differ on baseline characteristics; not included in meta‐analysis denominators