Cochrane (The same research group that found AA ineffective [rebuttals here and here], declared stimulants an effective treatment for cocaine addiction and provided oxygen for breathless headlines about the effectiveness of naltrexone for alcohol dependence [rebuttal here].) has done a meta-analysis of motivational interviewing (MI) and found it pretty underwhelming:
We included 59 studies with a total of 13,342 participants. Compared to no treatment control MI showed a significant effect on substance use which was strongest at post-intervention SMD 0.79, (95% CI 0.48 to 1.09) and weaker at short SMD 0.17 (95% CI 0.09 to 0.26], and medium follow-up SMD 0.15 (95% CI 0.04 to 0.25]). For long follow-up, the effect was not significant SMD 0.06 (95% CI-0.16 to 0.28). There were no significant differences between MI and treatment as usual for either follow-up post-intervention, short and medium follow up. MI did better than assessment and feedback for medium follow-up SMD 0.38 (95% CI 0.10 to 0.66). For short follow-up, there was no significant effect . For other active intervention there were no significant effects for either follow-up.
This seems to support my personal view that MI can be an effective tool for treatment induction, but is not an effective treatment for substance dependence by itself. The meta-analysis included studies of abuse and dependence, it’d be interesting to see if there were any differences for those populations.