Kumar 2020.

Study characteristics
Methods	Study design: randomized controlled cross‐over trial Country: India Setting: 0utpatient alcohol treatment Eligibility criteria: individuals with alcohol dependence seeking outpatient treatment and reporting no other substance use Duration of follow‐up: 6 months Informed consent: all participants provided written informed consent Ethical approvals: ethical review obtained. Institution not specified.
Participants	Sample size: 90 (30 naltrexone, 30 acamprosate, 30 baclofen) Description of the target population: Males in outpatient alcohol treatment Age: not reported Sex: 100% male Race/Ethnicity: not reported Marital status: >85% Harmful alcohol use (baseline): mean duration of alcohol intake was 18.58 years (SD = 10.62) in baclofen group, 15.43 years (SD = 8.76) in acamprosate group, and 15.75 years (SD = 10 years) in naltrexone group. Co‐occurring disorders: not described
Interventions	Type: pharmacologic Description (Group 1): Baclofen Duration and frequency Group 1): started with 20 mg/day and increased to an average of 38.11 mg/day  Description (Group 2): Acamprosate Duration and frequency Group 2): started with 333 mg/day and increased to an average of 1021.2 mg/day Description (Group 3): Naltrexone Duration and frequency Group 3): started with 25 mg/day and increased to 50 mg/day Delivery and provider: not described
Outcomes	Primary outcome(s): relapse Primary outcome measurement tool(s): Obsessive Compulsive Drinking Scale Secondary outcome(s): risk of relapse Secondary outcome measurement tool(s): Advance Warning of Relapse (AWARE) Questionnaire Time points: 6 months
Notes	Study funding and conflicts of interest: not reported Linked study records: none
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization of the participants was done using a computerized random table." Pg. 2
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "Our study was not blinded" Pg. 6
Blinding of outcome assessment (detection bias) All outcomes	High risk	"Our study was not blinded" Pg. 6
Incomplete outcome data (attrition bias) All outcomes	High risk	High dropout not equal across conditions (Table 5)
Selective reporting (reporting bias)	Unclear risk	Not described