A Meta-Analysis of Research on Motivational Interviewing Treatment

A Meta-Analysis of Research on Motivational Interviewing Treatment Effectiveness (MARMITE) Jennifer Hettema Julie Steele William R. Miller Annual Review of Clinical Psychology Vol 1, 2005 (in press)

funded by a grant from The Robert Wood Johnson Foundation

Adoption Curve for Innovations Source: Everett M. Rogers Diffusion of Innovations

Number of MI Publications Source: www. motivationalinterview. org/library/biblio. html MARMIT

MI Outcome Trials Source: www. motivationalinterview. org/library/biblio. html MARMIT

Inclusion Criteria for MI Trials For within-group effect sizes: n At least one treatment group including MI n At least one post-treatment outcome measure For between-group effect sizes: n At least one control or comparison condition without MI components n Procedure for creating pre-treatment equivalence of groups MARMIT

All studies double-coded for: Methodological quality on 12 dimensions n Other study characteristics n Attributes of the MI intervention(s) n Within-group effect sizes n Between-group effect sizes n MARMIT

Effect sizes were computed: For all reported outcome variables n At all reported follow-up points n For all between-group contrasts n With 95% confidence intervals n Correcting for small sample bias n MARMIT

72 studies included so far: Alcohol (31) One study each: n Drug Abuse (14) Gambling n Smoking (6) Eating Disorders n HIV Risk (5) Relationships n Treatment Compliance (5) n Water purification (4) n Diet and exercise (4) n MARMIT

Types of Comparisons MI vs. Specified Treatment (25) n MI vs. Treatment as Usual (6) n MI vs No Treatment / Placebo (21) n MI added to Specified Treatment (7) n MI added to Treatment as Usual (5) n Mixed Designs (6) n Within-Group Only (2) n MARMIT

Methodological Quality Compared to 361 alcohol treatment trials: n MQS Mean = 10. 76 vs. 10. 68 (ns) n Intervention quality control 78% vs 57% n Multisite trials: 28% vs. 5% n Follow-up > 12 months 18% vs. 51% n Follow-up completion > 70% 45% vs. 75% MARMIT

Outcome (Dependent) Measures Mean of 3. 3 outcome variables per study n Range: 1 to 12 n n To avoid capitalization on change, we computed a combined effect size (d) averaging across all reported outcome variables in each study MARMIT

Specified Characteristics of MI n n n Being collaborative Client centered Nonjudgmental Building trust Reducing resistance Increasing readiness n n n Increasing self-efficacy Reflective listening Increasing discrepancy Eliciting change talk Exploring ambivalence Expressing empathy MARMIT

Specified Characteristics of MI Of 12 possible characteristics of MI, The average number mentioned was 3. 6 Range: 0 -12 MARMIT

Treatment “Dose” of MI Average “dose” of 2 sessions (2. 2 hours) The contrasts in dose varied from: Comparison group 25 hours longer than MI to MI 6 hours longer than no-treatment MARMIT

Quality Control of MI n Average training time: 10 hours (N=13) Manual-guided n Post-training supervision n Fidelity checks n 74% 29% 36% MARMIT

Where was MI tested? n n n n Outpatient clinics (15) Inpatient facilities (11) Educational settings (6) Community organizations (5) G. P. offices (5) Prenatal clinics (3) Emergency rooms (2) n n n n Halfway house (2) EAP Telephone (3) In home (1) Jail (1) Mixed (7) Unspecified (8) MARMIT

Who delivered MI? Paraprofessionals / students (8) n Master’s level (6) n Psychologists (6) n Nurses (3) n Physicians (2) n Dietician (1) n Mixed (22) n MARMIT

Sample Characteristics (N = 14, 267) N = 21 to 952 n Males = 54. 8% n Mean Age = 34 n Ethnic minorities: n Mean = 198 Range = 0 to 100% Range = 16 to 62 43% (N = 37) MARMIT

Some Generalizations n n n Wide variability in effect size across studies, within problem areas (e. g. , for alcohol problems, d varies from 0 to 3. 0) Effects of MI appear early Effects of MI diminish over time, except in additive studies u d =. 77 at post-treatment u d =. 31 at 4 -6 months u d =. 30 at 6 -12 months MARMIT

Effect Size of MI Over Time Controlled Additive Comparative MARMIT

Effect size was not predicted by: Number of MI attributes mentioned n Methodological quality of study, except Use of a manual to guide MI did predict effect size: u Studies not using a manual d =. 65 u Studies using a manual d =. 37 n Demographic characteristics, except: u Anglo/Caucasian samples d =. 39 u Minority samples d =. 79 n MARMIT

Effect size varied with outcome measures Alcohol: n Quantity of drinking n Frequency of drinking n BAC estimates n Negative consequences HIV Risk: n Knowledge n Behavioral Intentions n Sexual risk-taking d =. 30 d =. 31 d =. 22 d =. 08 d = 1. 46 d =. 88 d =. 07 MARMIT

Mean Combined Effect Size by Problem Area (N=72 Clinical Trials) MARMI

Conclusions 1. Robust and enduring effects when MI is added at the beginning of treatment MI increases treatment retention u MI increases treatment adherence u MI increases staff-perceived motivation u MARMIT

Conclusions 2. The effects of motivational interviewing emerge relatively quickly (This is also true of other treatments) Project MATCH Outcomes MARMIT

Conclusions 2 a. The effects of motivational interviewing emerge relatively quickly q This may not be true for certain problem areas or dependent measures where “sleeper” effects occur (e. g. , effects of diet and exercise) MARMIT

Conclusions 3. The between-group effects of motivational interviewing tend to diminish over 12 months This is also true of other treatments u Between-group differences diminish in part because control/comparison groups “catch up” over time u This may not be true of MI’s additive effects with other treatment u MARMIT

Conclusions 4. The effects of MI are highly variable across sites and providers This is also true of other treatments, but may be more true with MI u Provider baseline characteristics do not predict effectiveness with MI u Treatment process variables do u Manuals may not be a good idea u MARMIT