Lerman 2002.

Methods	Double‐blind randomised placebo‐controlled clinical trial of bupropion HCL (brand name Zyban) in adult male and female smokers Study period: June 1999 to March 2002
Participants	N = 555 allocated and received intervention (Lerman 2006). Numbers included in published pharmacogenetic analyses include ˜ 414 (Lerman, 2006), ˜ 412 (Conti, 2008), and ˜ 416 (Bergen, 2013), which refer to European ancestry individuals only, but data made available for this review include non‐Hispanic white and non‐Hispanic black, total ˜ 494. Participants were smokers seeking treatment and were recruited through advertisements for a free smoking cessation programme at Georgetown University and at the State University of New York at Buffalo. Inclusion criteria: age ≥ 18 years, reported smoking ≥ 10 cigarettes/d for the previous 12 months, who provided informed consent for both genotyping and treatment Exclusion criteria: planning a pregnancy, pregnant, lactating; seizure disorder, history of head trauma or prior seizure, family history of a seizure disorder; brain (or CNS) tumour; history of or current diagnosis of bulimia or anorexia nervosa; diabetes treated with oral hypoglycaemics or insulin; excessive use of alcohol or alcoholism; current addiction to opiates, cocaine, or stimulants; use of other drugs containing bupropion (e.g. Wellbutrin, Wellbutrin SR); allergy to bupropion; currently taking particular medications (e.g. monoamine oxidase inhibitor, antipsychotics, antidepressants, theophylline, systemic steroids, over‐the‐counter stimulants and anorectics); recently taking an MAOI (< 14 days); or recent discontinuation of a benzodiazepine
Interventions	Bupropion and 7 sessions of in‐person behavioural group counseling (N = 285 allocated, 229 received allocation), and Placebo and 7 sessions of in‐person behavioural group counseling (N = 70 allocated, 211 received allocation) Participants received 10 weeks of pills (active or placebo) initiated on the first day of counseling. Bupropion treatment was standard (150 mg/d for 3 days, then 300 mg/d). The target quit day was day of the third counseling session.
Outcomes	Primary outcomes: continuous abstinence measured at end of treatment and at 6 months after cessation Secondary outcomes: short‐term quit rates using 7‐day and 30‐day point prevalence 83% and 86% of participants reporting abstinence at end of treatment and at 6 months provided a CO sample for verification.
Funding source	Study medication for the bupropion trial was provided by GlaxoSmithKline. This research was supported by grants from the National Cancer Institute and the National Institutes of Drug Abuse, P50CA/DA84718 and RO1CA 63562.
Declaration of interest	Dr Berrettini has consulted for GlaxoSmithKline.
Notes	Lerman 2002 described an interim analysis of mediators on cessation and was the first published description of this trial. Lerman 2006 is the definitive description of the trial and the first published analysis of pharmacogenetics.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation was computer generated by a senior data manager.
Allocation concealment (selection bias)	Low risk	Allocation was concealed from counselors and study assistants.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blind
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcomes were assessed via a timeline follow‐back method through phone interviews with study assistants.
Incomplete outcome data (attrition bias) All outcomes	High risk	As described in Lerman 2006, 17% withdrew after allocation but before intervention (not included in ITT analysis), and 23% of the remainder were lost to follow‐up or discontinued treatment but were included in the ITT analysis. No significant differences in losses were reported at either stage by arm. However, as abstinence outcomes were 27% at EOT and 22% at 6 months, attrition proportion seems high.
Selective reporting (reporting bias)	Low risk	Conti 2008 reported on continuous abstinence.
Other bias	Low risk	Genotype completion rate was very high overall, so selection bias by treatment group is low.