McCambridge 2004.

Methods	Cluster‐randomised controlled trial
Participants	Number of participants: 179 City and country: London, England Type of setting: Urban School setting: Alternative campus (further education training) Gender: 46% female, 54% male Mean age: 18.0 years Inclusion criteria: 16 or older, attending FET, weekly or more use of cannabis Exclusion criteria: Younger than 16, older than 19; less than weekly use cannabis; literacy (low levels); not English speaking
Interventions	Number of adolescents allocated to each group: 97 allocated to experimental condition, 82 allocated to control condition Brief intervention: Motivational intervention versus information and advice‐giving Dosage: 1 session Type of delivery: Face‐to‐face (individual) Timing: 1 hour
Outcomes	Follow‐up at 3 months and 12 months (2005 study) Measures: Severity of Dependence Scale, The Drug Attitudes Scale Primary outcomes: Frequency cannabis use Quantity cannabis use Cannabis use mean dependence score Frequency alcohol use Quantity alcohol use Alcohol use mean dependence score Frequency alcohol use Quantity alcohol use Alcohol use mean dependence score Quantity methamphetamine tablets used Secondary outcomes: Cannabis‐Interactional Problems score Cannabis Problems score Alcohol‐Interactional Problems score
Notes	Only alcohol, cannabis frequency outcomes were measured at 12 months' follow‐up Funding: Action on Addiction for 12 12 months' follow‐up assessments. Conflict of interest: Information not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation was non‐computerised and consisted of a colleague not involved in the study allocating clusters randomly to either the intervention or control condition. Stratification by college was applied in order to control for local variations in drug use
Allocation concealment (selection bias)	Low risk	Complete concealment was mentioned by the authors
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding is not possible for the type of intervention
Blinding of outcome assessment (detection bias) All outcomes	Low risk	As 1 interventionist was the study principal investigator, a second independent interviewer who was blind to study condition was employed to conduct 3 months' follow‐ups, and an additional interviewer who was blind to initial group allocation was employed for 12 months' follow‐ups
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition analyses conducted, and no difference was found between groups. Various factors associated with attrition in both groups were identified and controlled for in the analysis. In addition, follow‐up rates were provided for 3 months' follow‐up (experimental group: 92.4%; control group: 86.3%) and 12 months' follow‐up (experimental group: 80%; control group: 82%)
Selective reporting (reporting bias)	High risk	All outcomes discussed and reported on at 3 months' follow‐up, although at 12 months' follow‐up there was some unplanned deterioration of the intervention effect, so certain outcomes were not reported on
Other bias	Low risk	No other sources of bias identified