Carroll 2009.

Methods	Multisite RCT (5 sites) in the USA.
Participants	436 Hispanic substance abusers from the USA.
Interventions	MET (n = 214) vs counselling as usual (n =222).
Outcomes	Physiological primary: Percent positive urine specimens. Non‐physiological primary: Days of substance use by week, percent days abstinent from alcohol. Secondary: Treatment retention (days enrolled in treatment at community treatment program through week 16).
Notes	The design paralleled that of Ball 2007b.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Urn allocation.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement.
Blinding (performance bias and detection bias) Patients and providers	Low risk	No blinding, but most outcomes were physiological and also used to validate self‐reports, and not likely to be influenced by lack of blinding.
Blinding (performance bias and detection bias) Assessors	Low risk	Insufficient information to know whether assessors were blinded. But most outcomes were physiological and also used to validate self‐reports, and not likely to be influenced by lack of blinding.
Incomplete outcome data (attrition bias) All outcomes	High risk	28 % lost to follow‐up. Reasons for attrition not described but similar between groups. Not ITT even though they reported an intention to treat sample.
Selective reporting (reporting bias)	Low risk	The published report included all expected outcomes based on the stated hypotheses.
Other bias	Unclear risk	Only self‐reported outcomes. Differences at baseline were not reported.