Roski 2003.

Study characteristics
Methods	Design: 3‐group cluster‐randomized controlled trial Setting: 40 clinics of a large multispecialty medical group practice providing primary care services, USA Recruitment: Exit interviews with patients in the clinic
Participants	4813 patients (873 people who smoked) at baseline survey. 4734 patients (863 people who smoked) at follow‐up. Patients were 18 years or older who had visited their provider in the past 30 days, av.age not reported, av. cpd not reported
Interventions	Common component in all groups: printed versions of the smoking cessation guidelines distributed to the practices. Intervention 1: practices received printed versions of the smoking cessation guidelines, financial incentives for reaching preset clinical performance targets Intervention 2: practices received printed versions of the smoking cessation guidelines, financial incentives for reaching preset clinical performance targets combined with access to a centralized registry of people who smoked and intervention system which delivered telephone counseling Control: distribution of printed versions of the smoking cessation guidelines only
Outcomes	7‐day PPA at 6m Validation: None Measures of provider implementation: Ask, Advise, Assist
Funding Source	Supported in part by a grant from the Robert Wood Johnson Foundation (Grant 036023)
Author's declarations of interest	Not reported
Notes	Strategy: Adjunctive counseling + Provider incentive Level: Patient + Practice + System Comparison type: Single component (provider incentives vs. standard care), active vs. active (isolating adjunctive counseling) & multicomponent vs. SC (provider incentives & adjunctive counseling)
Risk of bias
Bias	Authors' judgement	Support for judgement
Sequence Generation	Unclear risk	Randomly allocated by block randomization. No further details reported
Allocation concealment	Unclear risk	No details reported
Blinding of outcome assessors All outcomes	High risk	Smoking status self‐report. Contact between arms was different
Incomplete outcome data All outcomes	Unclear risk	No details reported
Recruitment bias (cluster RCTs only)	Low risk	Participants were affiliated with the practice before randomization
Balanced baseline characteristics? (cluster RCTs only)	Low risk	QUOTE: "At baseline no differences were found between the experimental conditions with respect to identification of tobacco use, provision of advice to quit, and assistance in quitting at the most recent clinic visit"
Adjustment for clustering in analysis? (cluster RCTs only)	Low risk	QUOTE: "Analyses of practice pattern changes (identification, offer of advice to quit) and patient outcomes (quitting) made use of clustered logistic regression"