Skip to main content
. 2016 May 5;2016(5):CD005336. doi: 10.1002/14651858.CD005336.pub4

Carroll 2006.

Methods Randomised controlled trial. Treatment referrals to a substance abuse treatment unit
Participants 136 individuals were referred from the office of adult probation to a substance abuse treatment unit and were randomised
Most participants were male (88%, 94%, 94% and 82% in Groups 1, 2, 3 and 4, respectively) and were in their early twenties (average 21.0, 21.5, 21.1 and 21.2 years), and most were African American (52%, 77%, 53%, 58%). Participants were typically employed (54%, 53%, 33%, 54%) and had completed at least high school (51%, 53%, 56%, 48%). Diagnosis of anxiety, depressive or personality disorder was common (81%, 86%, 75%, 69%)
Participants began to use cannabis at an average age of 14 years (14.4, 14.4, 14.9, 14.7) and used cannabis every second day (average 13.8, 13.7, 12.4 and 12.5 days per 28 days)
Participants consumed alcohol only a few days per month (average 1.9, 1.7, 4.1, 3.3 days). Use of tobacco and other illicit substances was not reported, although participants were excluded if they reported a "physical dependence on alcohol or opioids"
Interventions Group 1: 8‐session MET/CBT over 8 weeks + up to $340 CM for treatment adherence + up to $540 for continuous abstinence (69.7% of participants completed treatment as intended; n = 33)
Group 2: 8‐session DC and option for self help groups over 8 weeks + up to $340 CM for treatment adherence + up to $540 for continuous abstinence (63.7% of participants completed treatment as intended; n = 34)
Group 3: 8‐session MET/CBT over 8 weeks (66.7% completed treatment as intended; n = 36)
Group 4: 8‐session DC and option for self help groups over 8 weeks (39.4% completed treatment as intended; n = 33)
Groups 1 and 4 shared the goal of cannabis abstinence, and Groups 2 and 3 shared the goal of cannabis reduction. Participant reimbursement for follow‐up assessments was not reported. Staff went through intensive training including demonstration of competence. Intervention fidelity was ensured through supervision and videotaping sessions and use of the Yale Adherence and Competence Scale
Outcomes Proportion of smoking days; duration of longest abstinence in days (self report and urinalysis); proportion with clinical improvement (defined as completing treatment and submitting ≥ 1 negative urine); other substance use reported on ASI
Notes Follow‐up was provided at end of treatment, then at 3 and 6 months
Follow‐up rates at final assessment:
  • Group 1: n = 27, 81.8%; Group 2: n = 24, 70.6%; Group 3: n = 27, 75.0%; Group 4: n = 30, 90.9%


Intention‐to‐treat analysis approach was used
Study was funded by the National Institute on Drug Abuse. Study authors reported no declarations of interest
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomisation process not explained
Allocation concealment (selection bias) Low risk Centrally located; otherwise does not refer to concealment procedures
Blinding of participants and personnel (performance bias) 
 All outcomes High risk Participant and personnel blinding was not possible because of the type of intervention provided
Blinding of outcome assessment (detection bias) 
 Subjective outcomes Unclear risk Blinding of assessors was not described, although staff were highly trained
Blinding of outcome assessment (detection bias) 
 Objective outcomes Low risk Urine collected during the trial to establish length of abstinence
Incomplete outcome data (attrition bias) 
 All outcomes Low risk Follow‐up rates were moderate, and no group differences were reported. ITT was used
Selective reporting (reporting bias) Low risk With the minor exception that ASI scores were not reported, all pre‐specified outcomes were reported
Other bias Low risk Compliance with CM was a little low. Outside treatments accessed during the trial period were not assessed, but this seems unlikely given the intensity of treatment. No other bias was found