Ashare 2019.

Study characteristics
Methods	Country: USA Setting: not explicitly reported, likely at home and in hospital Aim: to evaluate the safety and efficacy of varenicline for SC among people living with HIV Study design: parallel placebo‐controlled phase 3 RCT Dates conducted: October 2012‐September 2018
Participants	179 adult smokers all with a confirmed HIV diagnosis, currently smoking and motivated to quit. All treated with antiretroviral therapy with HIV loads < 1000 copies/mL and CD4+ counts > 200 cells/mm3. 32% female, mean age 48.6, average CPD at baseline 11.5
Interventions	Varenicline 1 mg x 2/day Placebo inactive tablets, same regimen All participants received 6 sessions over 9 weeks of interactive behavioural support in person or by telephone. Treatment period was 12 weeks.
Outcomes	Primary: 7‐day PPA, CO‐confirmed, at weeks 12 and 24 Secondary: CA and time to relapse
Notes	New for 2022 update Funding source: "National Institute on Drug Abuse, Penn Center for AIDS Research, Penn Mental Health AIDS Research Center. Pfizer provided medication and placebo free of charge" Declaration of interests: "Dr. Schnoll has provided consultation to Pfizer, GlaxoSmithKline, and Curaleaf. Dr. Gross serves on a Pfizer Data and Safety Monitoring Board for a drug unrelated to smoking or HIV. Dr. Ashare has an investigator‐initiated grant from Novo Nordisk for a drug unrelated to the current study."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Eligible participants were randomized 1:1 by a computer‐generated protocol provided by the study statistician to the University of Pennsylvania’s Investigational Drug Service (IDS)"
Allocation concealment (selection bias)	Low risk	Randomisation conducted separately by the University of Pennsylvania’s Investigational Drug Service (IDS), and all other study personnel were blinded. Therefore, concealment was likely adequate.
Blinding (performance bias and detection bias) All outcomes	Low risk	Abstinence was biochemically validated and "All participants and study personnel, aside from IDS [Investigational Drug Service], were blinded from treatment arm allocation throughout the trial."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition rates were similar between study arms.
Selective reporting (reporting bias)	Low risk	Prespecified outcomes from trial registry entry all reported in published results