Cinciripini 2005.

Methods	Double‐blind parallel‐group placebo‐controlled randomised clinical trial Study period: February 1996 to April 1997
Participants	N = 147 Participants were smokers recruited from the Houston, Texas, metropolitan area via newspaper, radio, and TV advertisements and public service announcements. Inclusion criteria: smoking ≥ 10 cigarettes/d at baseline, between 18 and 75 years old Exclusion criteria: taking smoking cessation treatment; taking psychoactive medication; or having any uncontrolled systemic illness, contraindications for taking venlafaxine or the nicotine patch, current substance abuse, or other psychiatric disorders
Interventions	Venlafaxine (n = 71) Placebo (n = 76) 21 weeks of active venlafaxine or placebo. After a 1‐week no‐medication baseline (3 weeks before the quit date), participants began antidepressant therapy 2 weeks before quitting at an initial dose of 75 mg/d (37.5 mg/d twice daily). The dose was increased up to 150 mg/d during the week just before participants were to quit. During each subsequent week, the dose was raised in 37.5‐mg increments up to a maximum 225 mg/d. Two weeks before the end of treatment, the dose was decreased by 37.5 mg every 2 to 3 days. The medication cycle was completed 18 weeks after quitting. All participants also used the nicotine patch (Prostep, 22 mg) for 6 weeks, beginning on their quit date, and received smoking cessation counseling.
Outcomes	Outcomes: Abstinence was assessed on the quit date and at post quit weeks 1, 3, 6, 18 (end of treatment), 26, and 52. Abstinence was verified in person by expired air carbon monoxide ≤ 10 ppm or by a saliva cotinine sample of < 15 ng/µL at 26 or 52 weeks.
Funding source	Support for this research was provided by grants from the MD Anderson Cancer Center (PRS), the National Cancer Institute (P50CA70907), and the National Institute on Drug Abuse (R01DA1182‐01) to Paul M. Cinciripini. Study medication was provided by Wyeth‐Ayerst Laboratories.
Declaration of interest	Declaration of interest was not reported.
Notes	Analysis of pharmacogenetics was reported in Cinciripini 2005.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomisation was performed by the pharmacy, but details are absent.
Allocation concealment (selection bias)	Low risk	Randomisation was centralised and was performed by a third party (pharmacy).
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants, study staff, and study personnel with direct patient contact were blinded to group assignment.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinding was appropriate and the outcome of the study was objective.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Incomplete outcome data were similar between groups.
Selective reporting (reporting bias)	Unclear risk	No protocol was found.
Other bias	Low risk	Additional analyses of pharmacogenetics in Cinciripini 2004 were conducted in the large majority of the original randomised trial population, so selection bias seems unlikely.