Siddiqi 2013.

Study characteristics
Methods	Design: 3‐group cluster‐randomized controlled trial Setting: Health centers in Pakistan Recruitment: Opportunistic in practice
Participants	1955 patients aged 18 years or older with suspected pulmonary tuberculosis (cough for 3 weeks without any other cause) who also regularly smoked tobacco (≥ 1 cpd), av. age 41, 95% M, av. cpd 16
Interventions	Intervention 1: Behavioral support sessions (BSS) • Participants received 2 structured sessions delivered by directly‐observed therapy facilitators using an educational flipbook (session 1: 30 minutes; session 2: 10 minutes on the quit day) • Directly‐observed therapy facilitators received a 1‐day training program delivered by the research team • Other healthcare professionals received briefing about BSS Intervention 2: BSS+ • Participants received a free 7‐week course of sustained‐release bupropion in addition to BSS • Physicians received training and written guidance on prescribing bupropion Control: • Participants received usual care and a self‐help leaflet on smoking cessation. No further details on the usual care reported • Directly‐observed therapy facilitators received information on trial procedures only
Outcomes	Continuous abstinence at 6m Valiation: Expired CO ≤ 9 ppm
Funding Source	International Development and Research Centre, Canada
Author's declarations of interest	Unable to access this information through the link provided in the paper
Notes	Strategy: Adjunctive counseling + Cost‐free medications + Provider training Level: Patient + Provider Comparison type: Multicomponent vs. standard care; active vs active (isolating effect of cost‐free medications)
Risk of bias
Bias	Authors' judgement	Support for judgement
Sequence Generation	Low risk	QUOTE: "...computer generated random‐number lists to generate the allocation sequence"
Allocation concealment	Unclear risk	No details reported
Blinding of outcome assessors All outcomes	High risk	Smoking status is self‐report and there was different contact between the control arm and the intervention arms
Incomplete outcome data All outcomes	Low risk	At participant level, the overall loss to follow‐up was 5.8% (n = 114/1955 survivors); 8.0% (n = 53/659) in the BSS+ group, 3.1% (n = 20/640) in the BSS group, and 6.3% (n = 41/656) in the control group at 6‐month follow‐up
Recruitment bias (cluster RCTs only)	Low risk	Participants were affiliated with the health centers before randomization
Balanced baseline characteristics? (cluster RCTs only)	Low risk	QUOTE: "The 3 groups were generally similar with respect to the baseline characteristics, although mean age, sex, and smoking type differed slightly"
Adjustment for clustering in analysis? (cluster RCTs only)	Low risk	QUOTE: "...adjusting for cluster effect using an intraclass correlation coefficient of 0.036"