Kottke 1989.

Study characteristics
Methods	Design: 3‐group cluster‐randomized controlled trial Setting: Primary care, USA Recruitment: Providers were recruited through mailing with brochure. Participants were recruited in practice
Participants	66 providers, 15% F, av.age 40, av. cpd 19 1653 patients smoked
Interventions	Intervention 1: physicians received a 6‐hour workshop on smoking cessation and smoking cessation manuals Intervention 2: physicians received smoking cessation manuals (same as the one given for those in intervention 1) to hand out to people who smoked 3. Control: no assistance. No further details reported
Outcomes	Abstinence (undefined) at 12m Validation: Blood cotinine levels (cut‐off not reported) Quit attempts Measures of provider implementation: Ask, Advise, Assist‐Quit date, Assist‐Self‐help, Arrange
Funding Source	This study was supported in part by National Institutes of Health grant CA38361, National Institute of Drug Abuse grant DA04066, and a National Institute of Drug Abuse Research Scientist Award, DA00109
Author's declarations of interest	Not reported
Notes	Strategy: Provider training Level: Provider Comparison type: Single component versus standard care
Risk of bias
Bias	Authors' judgement	Support for judgement
Sequence Generation	Unclear risk	No details reported
Allocation concealment	High risk	QUOTE: "After the randomization had been initiated, it became apparent that some physicians had given home addresses while others had given work addresses. This had prevented the investigators from recognizing all cases in which multiple physicians from the same group had responded to the recruitment letter. To prevent contamination from having physicians of the same practice in different trial groups, all physicians in the same practice were either moved to the most intense level of intervention to which any of them had been originally randomized or, if not yet randomized at the time this problem was discovered, added to the group to which their partner(s) had been randomized"
Blinding of outcome assessors All outcomes	Low risk	Smoking status biochemically validated
Incomplete outcome data All outcomes	Low risk	Attrition rates were under 50% and similar across the groups. At physician level, there was no loss to follow‐up (n = 0/66) at 1 year. At participant level, the overall loss to follow‐up was 13.0% (n = 215/1653); 13.2% (n = 87/660) in the workshop intervention group, 12.5% (n = 74/593) in the materials group, and 13.3% (n = 53/400) in the control group were lost to follow‐up at 1 year.
Recruitment bias (cluster RCTs only)	Low risk	Participants were affiliated with the practice before randomization
Balanced baseline characteristics? (cluster RCTs only)	Low risk	QUOTE: "Neither the mean age of the physicians, nor the patient load on the physician differed significantly among the three groups"; "While a higher proportion of the patients of physicians in the no‐assistance group had at least some education beyond high school (51.8% vs 42.1% for patients of physicians in the workshop group and 42.9% for patients of physicians in the materials group [P<.001]), the distributions for the other variables did not differ significantly among the patients in the three groups"
Adjustment for clustering in analysis? (cluster RCTs only)	Low risk	QUOTE: "The physician was the unit of analysis...multivariate regression was used to adjust for potentially confounding effects of differences among the groups of doctors and their patients"