Windsor 1988.

Methods	Study design: Randomized controlled trial Setting: University worksite, USA Recruitment: Employees volunteering for a quit‐smoking programme
Participants	378 smokers av. age 37, av. cpd 23 ‐ 27
Interventions	Therapist: health educator All groups received a 10‐min session of brief advice 1. + S‐H manuals 2. + S‐H and another session of counselling (20 ‐ 30 mins) with skills training, buddy selection and a contract 3. as 1, with monetary rewards for cessation 4. as 2, with monetary rewards for cessation
Outcomes	Abstinence at 1 yr (sustained at 6 wks, 6 m, 1 yr, no more than 2 cigs in period) Validation: saliva thiocyanate < 100μg/ml at all follow‐ups
Notes	There was no apparent effect of monetary incentives so this arm is collapsed. 4 and 2 vs 3 and 1. Number of quitters estimated from graphs
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated assignment
Allocation concealment (selection bias)	Low risk	Sealed numbered envelopes opened after informed consent and baseline questionnaire
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Biochemical validation of abstinence
Incomplete outcome data (attrition bias) All outcomes	Low risk	37 lost to follow‐up, included in ITT analysis