Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Motivational Interviewing and Motivational Enhancement in Addiction Care

Abstract. This fundamental research monograph examines motivational interviewing (MI) and motivational enhancement therapy (MET) as psychological methods for supporting behavior change in addiction care. It reconstructs the development of MI from its origins in problem-drinking treatment through contemporary collaborative methods; analyzes ambivalence, change talk, sustain talk, commitment, autonomy, treatment engagement, and therapist fidelity; and evaluates original randomized trials, process studies, meta-analyses, and clinical guidance. Major evidence includes Project MATCH, NIDA Clinical Trials Network studies, cannabis-treatment experiments, the 2023 Cochrane update, and mechanistic coding research. The synthesis finds context-dependent benefits, especially against weak comparison conditions, with much less consistent superiority over active treatments and continuing uncertainty about durable effects, retention, and causal mediators. It delineates implications for alcohol, cannabis, opioids, adolescents, clinical implementation, and digital fidelity supervision. MI is positioned as an evidence-informed component of accessible, patient-centered care rather than a prerequisite for medical treatment.

Keywords: motivational interviewing; motivational enhancement therapy; addiction care; ambivalence; change talk; sustain talk; treatment engagement; treatment matching; clinical trials; therapeutic fidelity; meta-analysis

Scientific Scope and Research Questions

Motivational interviewing (MI) is a collaborative, goal-oriented method of communication that attends especially to the language of change within an interpersonal relationship. In addiction care its immediate object is the person's engagement with a meaningful behavioral decision: whether to discuss substance use, reduce consumption, accept an assessment, initiate or continue treatment, use overdose-prevention resources, or pursue a personally chosen recovery goal. Motivational enhancement therapy (MET) is a more structured, usually brief clinical application of motivational principles, commonly incorporating personalized assessment feedback and planned follow-up. Neither term denotes a single pharmacological treatment or a complete explanation of the origins of addiction. The operational distinction matters because trials of one session, four-session MET, and multi-component therapy answer different questions Miller and Rollnick (2023) Miller (2023); Schwenker et al. (2023) Schwenker (2023).

The present monograph asks what constitutes competent motivational work, how the method developed, which conversational and motivational mechanisms are plausible, and what experimental comparisons establish about substance use, treatment engagement, and sustained outcomes. It examines alcohol, cannabis, other drugs, adolescents and adults, culturally diverse settings, brief intervention, clinical sequencing, practitioner fidelity, and outcome measurement. The evidence base is evaluated through original trials, therapy-process coding, reviews, meta-analyses, and public clinical guidance. This is a critical narrative scientific synthesis rather than a preregistered systematic review. It therefore reports findings from identified sources without claiming an exhaustive literature search or new pooled effects.

The central inference is comparative: efficacy depends on what MI is compared with, who receives it, who delivers it, which outcome is measured, and when that outcome is observed. An improvement after treatment is not, by itself, evidence that MI caused the improvement; an advantage against no intervention cannot be generalized automatically to superiority against a competent active treatment. The current Cochrane synthesis (2023) Schwenker (2023) makes this distinction particularly important.

Historical Formation: From Problem Drinking to an Interviewing Method

William R. Miller's 1983 paper Motivational Interviewing with Problem Drinkers Miller (1983) marked an early explicit articulation of the method in the context of alcohol problems. Its explanatory vocabulary drew on person-centered counseling, social psychology, attribution, dissonance, and self-efficacy, locating motivation within a conversation rather than treating resistance as a fixed patient trait. The 1983 account belongs to a particular historical stage of MI, including its original language about discrepancy and readiness; present editions place stronger emphasis on partnership, acceptance, compassion, and the person's agency.

Miller's historical review (2023) Miller (2023) traces a progression from clinical observations and skill descriptions to cross-professional application, randomized research, and investigations of therapist and client speech. The collaboration of Miller and Stephen Rollnick systematized the approach for clinical practice and education. Their fourth edition (2023) Miller (2023) organizes the work around engaging, focusing, evoking, and planning; those interrelated tasks provide a contemporary account of the clinical conversation.

Historical claims require separation from current efficacy. The importance of a treatment innovation and its diffusion across services show scientific influence and feasibility; its comparative benefit requires appropriately controlled outcome studies. Similarly, early theories of discrepancy and verbal commitment motivated productive experiments, but their mechanistic validity must be assessed against subsequent process findings rather than inferred from the longevity of MI terminology.

Motivation, Ambivalence, and the Choice of Change Goals

In addiction care motivation is a dynamic configuration of incentives, expectations, capacities, perceived risks, social context, and possible futures. A person may value immediate effects of substance use while also acknowledging conflict with health, finances, relationships, or personal projects. Ambivalence is thus an intelligible co-occurrence of competing evaluations. MI invites a patient to articulate both sides without forcing the clinician's preferred conclusion into the patient's mouth. The purpose of the dialogue is to discover whether and how a desired direction of change becomes personally actionable.

This account supports distinctions between wanting change, believing change is possible, prioritizing change, committing to an intention, initiating action, and maintaining it. A clinician may encounter strong desire alongside low confidence, or confident plans alongside little concern about consequences. These psychological states cannot be reduced to a single readiness score. The foundational Amrhein et al. (2003) Amrhein (2003) study investigated multiple dimensions of client speech—commitment, desire, ability, need, readiness, and reasons—and encouraged more precise tests of the relation between motivational language and subsequent drug use.

Readiness also changes as housing, transportation, withdrawal, physical illness, legal pressure, family support, stigma, and treatment availability change. A patient's failure to attend a clinic may reflect structural obstacles or medical instability rather than a motivational deficit. A scientifically adequate formulation combines the interpersonal process with actual opportunity for change. SAMHSA TIP 35 (2019) Substance Abuse and Mental Health Services Administration (2019) situates motivational counseling inside a wider treatment environment.

The Four Conversational Tasks

Engaging establishes a working relationship in which the patient's experience and priorities can be understood. Focusing identifies the specific behavioral target to be considered; this is essential because an interview can feel supportive while never establishing what decision is at stake. Evoking draws out the patient's own reasons, abilities, needs, values, and commitments concerning that target. Planning assists the person in choosing feasible next actions and supports their continuing ability to act. These tasks often recur rather than progressing in a rigid order Miller and Rollnick (2023) Miller (2023).

In addiction services the focus may need to shift during a single encounter. Someone seeking treatment for alcohol use may first require a discussion of safe withdrawal, and someone using opioids may ask about medication initiation before discussing abstinence. A task-oriented MI formulation permits such shifts while preserving the patient's role in decisions. Engaging is not a substitute for offering information; evoking is not an excuse to omit clear risk information; planning is not equivalent to demanding a written promise. The manner of presenting options can be collaborative even when clinical action is urgent.

The four-task structure also creates observable criteria for practice. An assessor can ask whether there was shared focus, whether patient language was elicited rather than supplied, whether the clinician recognized discord, and whether any plan was voluntarily expressed and realistic. These criteria support both research coding and practical supervision; they are not proof that any individual session will produce a lasting substance-use reduction.

Relational Foundations: Empathy, Acceptance, and Partnership

MI belongs to a family of counseling approaches in which relationship quality has both ethical and methodological significance. Accurate empathic listening may help a patient disclose substance use, uncertainty, and treatment experiences that would otherwise remain hidden. Acceptance and partnership sustain the patient's opportunity to evaluate alternatives. A respectful stance supports autonomy while maintaining the clinician's responsibility to recognize urgent medical risks. This relational dimension is elaborated in the clinical manual Miller and Rollnick (2023) Miller (2023).

Empathy is not interchangeable with agreement. A clinician can accurately reflect the value a patient finds in drinking while also documenting injury, withdrawal risk, or family concerns. Likewise, collaboration does not prevent a clinician from recommending effective treatment. The meaningful scientific question concerns how relational conduct shapes disclosure, perceived agency, conversational participation, and later behavior.

The broader process literature remains mixed. In a synthesis of 36 primary studies, Magill et al. (2018) Magill (2018) found evidence for aspects of a technical pathway linking counselor behaviors, client language, and follow-up behavior, while the relational hypothesis was not supported in that specific aggregated test. This finding does not establish that empathic practice is dispensable; it shows that a plausible clinical foundation may be difficult to validate as an isolated mediator across heterogeneous coded trials.

Change Talk, Sustain Talk, and Commitment Language

Change talk refers to patient statements favoring a particular behavioral change; sustain talk refers to statements favoring persistence of the current behavior. Both forms can appear within a few minutes, particularly when competing incentives are salient. Commitment language is narrower: it concerns intention, decision, or pledge. The useful clinical distinction is between describing a possible advantage of treatment and declaring an action that a person intends to attempt.

The observational analysis by Amrhein et al. (2003) Amrhein (2003) coded videotaped interviews with 84 participants with drug-use problems and studied how expressions of desire, ability, need, reasons, readiness, and commitment related to later outcomes. The work helped orient MI research toward the temporal organization and strength of patient speech, although observational prediction does not alone establish that inducing a specific utterance will change behavior.

A later study by Moyers et al. (2009) Moyers (2009) used sequential behavioral coding and reported support for a pathway from therapist conduct through change talk to drinking outcomes. The 2018 process meta-analysis Magill (2018), however, demonstrated that MI-consistent therapist behavior may elicit both change and sustain language; it also reported stronger consistency for sustain talk as a predictor of worse outcome than for change talk considered independently. Good research therefore treats conversation as a dynamic interaction whose effects cannot be inferred by simply counting affirming statements.

Discrepancy, Self-Efficacy, and the Temporal Structure of Change

A motivational discrepancy arises when a person recognizes a difference between current behavior and an important value, responsibility, or goal. Historically, discrepancy was often described through a dissonance framework. Contemporary MI attempts to explore it in ways that preserve confidence and agency. A patient who already feels shame about substance use may become more disengaged if discrepancy is intensified through criticism. Reflection can instead make room for a sober evaluation of costs, desired benefits, and reasonable alternatives Miller (1983) Miller (1983); Miller and Rollnick (2023) Miller (2023).

Self-efficacy concerns perceived ability to perform actions in an identified situation. Within addiction treatment it should be linked to concrete conditions: refusing a drink at a particular gathering, attending a medication visit, asking for transport, managing craving, or seeking support after renewed use. Confidence without resources can be fragile, while extensive assistance may make a difficult behavior feasible even when initial confidence is low.

A credible causal account must therefore include time. Clinician behavior influences what the patient says during an encounter; statements may reflect or shape intentions; intentions encounter environmental opportunities and constraints; subsequent behavior feeds back into expectations and confidence. Each transition requires empirical testing. The Project MATCH three-year follow-up Project MATCH Research Group (1998a) identified readiness and self-efficacy as important prognostic variables, but the prognostic value of these measures is not itself evidence for a unique MI mechanism.

Motivational Interviewing and Motivational Enhancement Therapy

The terms MI and MET overlap scientifically but should remain operationally distinct. MI describes a conversational method usable within many treatment encounters. MET denotes a more specified treatment package frequently involving structured personal assessment, normative or individualized feedback, discussion of reactions to that feedback, and additional sessions devoted to change planning. In Project MATCH Project MATCH Research Group (1997a), MET consisted of four sessions, compared with twelve sessions each of cognitive-behavioral coping-skills therapy and twelve-step facilitation during a twelve-week treatment period.

The comparative design establishes a useful interpretation limit: an observed effect of MET may reflect its counselor conduct, structured feedback, duration, attention, or their combination. Conversely, a trial of a single MI conversation cannot be taken as a direct test of a four-session manual-guided program. Some research also uses the labels flexibly, so the actual intervention description must take precedence over the title.

For synthesis, the key intervention variables are contact time, number of sessions, individual or group format, voluntary or mandated participation, baseline severity, nature of feedback, trained counselor characteristics, comparison treatment, additional psychosocial services, and presence of medical care. Aggregating these without distinction can produce misleading conclusions about what was tested Schwenker et al. (2023) Schwenker (2023).

Project MATCH: Design, Outcomes, and What the Trial Established

Project MATCH randomized two parallel cohorts of people with alcohol dependence: 952 outpatients and 774 patients entering aftercare following intensive treatment. The three manual-guided approaches were MET, cognitive-behavioral coping skills, and twelve-step facilitation. Investigators asked whether participant characteristics predicted superior outcomes when aligned with a particular treatment. The primary 1997 outcome report Project MATCH Research Group (1997a) found substantial improvements across treatment conditions and comparatively limited, inconsistent evidence that the proposed matching attributes identified one reliably superior therapy.

This landmark study demonstrated that a shorter MET protocol could be evaluated alongside longer established interventions in a major multisite research program. It did not prove that four sessions are universally equivalent to twelve, because treatment content, populations, and service context differed. The later three-year outpatient analysis Project MATCH Research Group (1998a) reported some sustained matching effects, notably involving initial anger, while concluding that differences among the three approaches remained relatively few.

A scientifically responsible interpretation distinguishes treatment main effects from matching effects. The finding that an attribute such as anger may modify a treatment contrast warrants follow-up research; it is not a validated general-purpose algorithm for allocating every patient to MET. The Project MATCH publications also illustrate how intensive trial monitoring, selected therapists, and a common treatment infrastructure can limit automatic generalization to under-resourced routine clinics.

Treatment Matching, Anger, and Individual Differences

Treatment matching proposes that the expected benefit of a specific therapeutic approach varies across recognizable patient characteristics. Potential moderators include psychiatric severity, dependence severity, social networks, readiness, anger, age, language, treatment history, and external pressures. A moderator requires a statistical interaction between a measured baseline characteristic and randomized treatment allocation. Evidence that a characteristic predicts poor outcome in all groups is prognostic evidence and does not establish treatment-specific matching.

The Project MATCH primary report (1997) Project MATCH Research Group (1997a) and secondary hypotheses paper (1997) Project MATCH Research Group (1997b) evaluated numerous prespecified interactions. Some effects appeared for high anger and dependence or particular network variables, while the overall pattern did not sustain most hoped-for allocation rules. Multiple tests, time-dependent findings, and heterogeneous outcomes complicate interpretation. The long-term analysis Project MATCH Research Group (1998a) offers a longer view but does not remove the need for independent replication.

Clinically meaningful personalization must add feasibility and patient preference to statistical moderation. Selecting treatment solely from a baseline score can ignore medication eligibility, service availability, comorbidity, and goals of care. Future trials can improve precision by prespecifying interactions, testing them in independent samples, describing absolute expected benefits, and studying decisions actually available to services.

The NIDA Clinical Trials Network and Community Clinics

Effectiveness research moves beyond carefully selected specialty settings to ask whether an approach can work in routine care. The National Institute on Drug Abuse Clinical Trials Network compared MI-related interventions with usual intake or counseling practices in community treatment programs. Carroll et al. (2006) Carroll (2006) studied motivational interviewing for engagement and outcomes among people seeking treatment for substance use, bringing the intervention into the operational reality of multisite services.

The subsequent Ball et al. (2007) Ball (2007) study, explicitly titled Site Matters, evaluated MET across community drug-abuse clinics and drew attention to between-site variation. Different staffing, recruitment, local practice, fidelity, and comparator quality can alter the apparent benefit of the same manual. The NIDA study registry for CTN-0004 National Institute on Drug Abuse (2007) specifies retention and reduction of drug use as central outcomes and describes a three-session MET comparison with standard treatment.

These trials show why implementation belongs within the causal model. When staff have little supervision or treatment access is constrained, a counseling conversation may influence initial engagement but remain insufficient to sustain recovery. Site effects also warn against claiming that a positive result from one clinic will transport unchanged across service systems, substances, and populations.

Alcohol Use: Brief Intervention and Continuing Care

Alcohol-related MI spans preventive conversations, early intervention for harmful drinking, specialist dependence treatment, and continuing engagement. A brief encounter in primary care differs from structured therapy for dependence with withdrawal risk. The NICE alcohol-use guideline National Institute for Health and Care Excellence (2011) recommends a motivational intervention during initial assessment and treats individualized goals, risk, comorbidity, and patient preferences as components of care. It also recognizes circumstances where abstinence is clinically appropriate and where harm reduction supports engagement.

An important causal distinction is between reduced consumption and medically safe withdrawal. MI can support an informed decision to seek help, but severe alcohol withdrawal may require clinical assessment and supervised management. Narrative framing that suggests a patient should first prove motivation before receiving withdrawal care misrepresents clinical priorities.

The Project MATCH evidence provides comparative data on psychosocial treatment outcomes. The broader Cochrane update (2023) Schwenker (2023) shows that MI effects relative to no intervention are more favorable than differences relative to treatment as usual or another active intervention. Research needs to specify drinking frequency, quantity, alcohol-related problems, adverse events, functioning, and treatment engagement rather than collapse these outcomes into the phrase success in alcohol treatment.

Cannabis Use and Motivational Enhancement

Cannabis treatment research offers particularly useful contrasts among MET, multi-component behavioral care, and delayed treatment. The Marijuana Treatment Project Research Group (2004) Marijuana Treatment Project Research Group (2004) randomized 450 adults diagnosed under the historical DSM-IV cannabis-dependence category to two-session MET, nine-session combined MET/CBT/case management, or delayed treatment. This design compared differing intensities and components, and it provides evidence that a treatment package containing motivational work can be evaluated on cannabis use and related consequences.

A separate Walker et al. (2006) Walker (2006) adolescent experiment involved 97 young people with frequent cannabis use who were not seeking formal treatment. Both immediate MET and delayed-treatment groups reduced use; the trial did not find a clear between-group advantage at the reported follow-up. A later Walker et al. (2011) Walker (2011) trial further examined MET among adolescents not seeking treatment. Such findings prevent an overly broad claim that the method is consistently superior merely because within-group reductions occur.

The 2021 VA/DoD substance-use guideline VA/DoD Clinical Practice Guideline Working Group (2021) provides a weak recommendation in favor of CBT, MET, or their combination as initial interventions for cannabis use disorder, considering patient preference and availability. This is an intervention-specific clinical judgment, not a claim that all people who use cannabis require treatment or that short counseling reliably treats severe impairment.

Opioid Use Disorder and Access to Medications

Opioid use disorder demands particular attention to the distinction between motivational support and evidence-based medical treatment. Opioid agonist therapy and other clinically indicated medication pathways address risks that conversation alone cannot reverse, including overdose, recurrent use, and discontinuity of care. MI may be offered to explore treatment goals, previous experiences, ambivalence about medication, missed visits, or concerns about stigma. It must not become a required test of readiness before a patient can receive indicated medication or overdose-prevention services.

The Yakovenko et al. (2024) Yakovenko (2024) explicitly states that psychosocial interventions may be offered as adjuncts to opioid agonist therapy and should not be mandatory prerequisites. It also supports harm reduction as part of continuing care. This distinction has immediate implications for causal interpretation: a study of MI-related retention cannot be read as evidence that medication initiation should be delayed pending motivational counseling.

Research on motivational interventions in opioid services must specify whether the outcome is first contact, induction, continuation, retention, substance use, overdose, or functioning. Baseline opioid tolerance, withdrawal, concurrent drug use, housing instability, clinic hours, and medication availability can affect each transition. Counselling may strengthen collaborative decision-making, but its effects are conditional on access to effective services.

Stimulants, Polysubstance Use, and Clinical Heterogeneity

People with stimulant-related or multiple substance-use problems form heterogeneous populations. Substance availability, concurrent alcohol and opioid use, sleep disturbance, affective symptoms, cognitive demands, criminal-legal involvement, and housing insecurity may influence both motivational discussions and outcomes. Trials pooling broad categories of illicit drug use should therefore report substance-specific results when possible and consider whether the effect is driven by particular populations or settings.

The multisite Carroll et al. (2006) Carroll (2006) and Ball et al. (2007) Ball (2007) trials provide information on MI/MET in drug-treatment clinics, but generalization to every substance, age group, and pattern of concurrent use requires separate evidence. For an individual also receiving pharmacological care or contingency management, motivation-based communication should be assessed as a component of a larger system rather than assumed to account for the entire treatment effect.

A crucial scientific problem is whether reported reductions in a target drug coincide with substitution, escalation of another drug, or meaningful improvement in functioning. Repeated assessments and transparent denominators are needed to address these questions. Interventions can also target safer use and continued care even when abstinence is not immediately chosen, while preserving accurate discussion of medical risk.

Adolescents, Emerging Adults, and Educational Settings

Adolescent developmental context changes the meaning of voluntary engagement, consent, family influence, and peer risk. Youth may face school-based referrals, parental pressure, disciplinary consequences, and comparatively limited control over housing and transport. MI can create space for a young person's account of goals and perceived consequences. An ethically sound program must distinguish confidential care from coerced disclosure and align its procedures with relevant consent and safety requirements.

McCambridge and Strang (2004) McCambridge (2004) investigated a single-session MI intervention among young people in a cluster-randomized design. Their 2005 follow-up McCambridge (2005) addressed deterioration of effects over time. McCambridge, Slym, and Strang (2008) McCambridge (2008) later randomized 326 students aged 16–19 years using cannabis frequently to MI or drug information and advice, and reported no between-group difference; fidelity was also limited. The sequence demonstrates why long-term comparisons and intervention quality are central.

The adolescent cannabis trials of Walker and colleagues (2006; 2011) Walker (2011) further separate feasibility of attracting participation from a demonstrated reduction attributable to treatment. School settings may facilitate access while creating selection and privacy concerns. Research must document who agreed to participate, who was excluded, and whether follow-up differed by initial motivation or school participation.

Culture, Language, Identity, and Adaptation

Communication is shaped by language, cultural norms concerning authority, family obligations, shame, agency, and help seeking. MI's commitment to eliciting the patient's meaning creates possibilities for adaptation, but the clinician's assumptions and the structure of the service can still constrain what is disclosed or chosen. Cultural responsiveness requires more than translation: examples, assessment feedback, clinical expectations, and available options must be intelligible and respectful within the person's context.

A Carroll et al. (2009) Carroll (2009) multisite randomized trial examined MET with Spanish-speaking substance users in routine treatment services. Such studies address the transportability of interventions across language groups and clinical contexts more directly than assuming that results from English-language samples apply unchanged. The earlier review of Hettema, Steele, and Miller (2005) Hettema (2005) discussed variation in effects by population; this observation supports further confirmatory work rather than a fixed universal ranking of populations by responsiveness.

Research designs should assess linguistic competence, cultural adaptation procedures, demographic representation, interpreter involvement, clinician-client concordance where relevant, measurement equivalence, and access to follow-up services. A culturally adapted intervention can retain its core collaborative method while modifying the manner in which goals and feedback are discussed. The outcome question remains empirical.

Therapeutic Alliance, Discord, and Rupture Repair

Discord can arise when counselor and patient disagree about the focus or meaning of a conversation. It may appear as withdrawal, argument, silence, or a shift away from the proposed topic. An MI-consistent response often examines the clinician's contribution: whether advice was premature, the goal was imposed, or the patient's experience was inadequately understood. The objective is to restore a workable relationship while maintaining accurate clinical information.

Alliance measures and coding of observable behaviors capture partially different constructs. A high patient-rated alliance may coincide with a session that contains little change-focused discussion; competent questioning can coexist with momentary discomfort. Mechanistic models that assume a single relational score fully explains change therefore omit the focus and temporal organization of talk.

The 2018 Magill et al. meta-analysis Magill (2018) assessed relational and technical hypotheses together and found meaningful variation across pathways and moderators. Its conclusions encourage methods that observe therapist behavior, patient speech, and subsequent outcomes without treating any one correlate as a complete causal mechanism. Where a rupture has occurred, repair should be studied as a sequence of interaction and later participation rather than merely a post-session satisfaction rating.

Personalized Feedback and the Interpretation of Norms

Assessment feedback can draw attention to patterns a patient has not explicitly compared with their own goals: frequency and quantity of use, consequences, risky situations, or variation over time. MET often incorporates such feedback into a discussion of the patient's reaction. The data must be accurate, interpretable, and presented in a way that allows questions. Population norms may be useful for context, but group comparisons should not be mistaken for a personal diagnosis or moral judgment.

The intervention mechanism is not simply information transfer. A clinician can deliver identical data through an authoritarian lecture or a collaborative discussion. The latter may support autonomy and elicit the patient's appraisal, while the former may promote disengagement. Experimental studies should separate the content of feedback from the interviewing style and identify whether the patient understood its meaning.

Comparison groups complicate effect estimates. The 2023 Cochrane review Schwenker (2023) distinguishes no intervention, treatment as usual, assessment plus feedback, and other active interventions. A contrast against assessment-plus-feedback is especially informative because both conditions involve attention to the substance-use problem; a benefit is then more plausibly related to the added conversational method, subject to differences in contact and implementation.

Observational Coding of Clinical Conversations

Mechanism research requires an observation system that records what clinicians and patients actually do. Coding schemes can classify open questions, reflections, affirmations, direction-giving, expressions of autonomy, change talk, sustain talk, and sequences of utterances. The choice of coding unit matters: counting sentences, turns, or episodes can yield different estimates, and statements of similar wording may carry different interpersonal functions depending on their context.

The Amrhein et al. (2003) Amrhein (2003) work provided a foundational account of commitment-related speech. Moyers et al. (2009) Moyers (2009) used sequential coding to evaluate a possible mediational chain. Magill et al. (2018) Magill (2018) aggregated evidence across 36 primary process studies and demonstrated that aggregate correlations can support some but not all aspects of a proposed mechanism.

For these measures to be useful, analysts must report the coding manual, training, blind assessment where feasible, inter-rater agreement, handling of repeated utterances and nested data, and the temporal order of coding and outcome measurement. A correlation between therapist skill and patient language is compatible with reciprocal causation: more motivated patients may also elicit particular clinician responses.

Therapist Fidelity and the MITI Framework

A treatment label in a trial protocol does not guarantee delivery of the intended intervention. Fidelity concerns whether the observed conversation approximates the skills and relational style specified by the intervention. The Motivational Interviewing Treatment Integrity framework, developed and revised through process research, offers structured ratings and counts rather than reliance on provider self-description. The MITI 4 methodological article (Moyers et al., 2016) Moyers (2016) reports rationale and preliminary reliability and validity evidence, including coding of clinical samples.

Fidelity research also distinguishes competence from mere adherence. A clinician may ask many open questions without maintaining a collaborative goal, or may appear warm while frequently supplying arguments rather than eliciting patient reasons. Quality evaluation should consider sampled encounters, coder training, contextualized ratings, and change over time. Self-rated proficiency is insufficient as the sole evidence of competence.

The Miller et al. (2004) Miller (2004) randomized training study found that feedback and coaching can enhance proficiency after training. Its importance lies in showing that implementing MI requires observation and skill development, not simply disseminating a manual. A treatment-effect study that does not document fidelity may underestimate the effect of a competently delivered method or mistakenly attribute generic counseling to MI.

Training, Supervision, and Therapist Effects

Training researchers face two distinct outcomes: whether a clinician learns observable MI behaviors and whether patients benefit as a result. The Miller et al. (2004) Miller (2004) trial randomized 140 licensed substance-use professionals across multiple training conditions and used recorded practice to assess later proficiency. Feedback and coaching contributed to performance, demonstrating the value of supported learning beyond didactic exposure.

Therapist effects also appear in large clinical trials. The Project MATCH Research Group (1998) Project MATCH Research Group (1998b) examined differences among 80 therapists across the three treatment conditions and detected therapist-related variation in satisfaction and outcomes. An average treatment effect can therefore hide clinically relevant variability in delivery.

Program-level quality assurance should measure a sample of actual encounters over time, provide nonpunitive supervision, track staffing changes, and distinguish implementation failure from absence of efficacy. These procedures should remain proportionate to resources and privacy protections. A clinician who delivers MI proficiently in training may change practice under workload, administrative pressure, remote sessions, or different patient populations. The existence of therapist variability is a scientific argument for evaluation, not an invitation to assign every poor outcome to an individual counselor.

A 2026 Frontier: Human and Machine-Assisted Fidelity Feedback

Digital tools can support transcription, organization of coded conversations, training feedback, and supervisory review. They also introduce possible error in recognizing speaker turns, informal speech, multilingual expression, sarcasm, and the local context of a patient's statement. An algorithm that labels a sentence as change talk has not necessarily established the intention behind that utterance, and a high classification score does not establish improved clinical outcomes.

A 2026 article by West and colleagues West (2026), published online in September 2026, investigated sustained clinician MI skills with human and machine-learning-assisted feedback. The available abstract reports maintained or improved skills with feedback and considerable session-to-session variability, based on an observational feedback program rather than a randomized clinical comparison of machine-assisted and nonassisted care. The study reported author interests in the commercial development of automated therapy feedback. These are implementation and training observations; they are not evidence that software alone can conduct motivational therapy with equivalent effects or safely replace clinical judgment.

Artificial systems may become useful in the scientific architecture of supervision, particularly for flagging encounters for human review and evaluating temporal patterns at scale. Research should prospectively assess annotation agreement, bias across languages and populations, secure data handling, the effect of automated feedback on clinician behavior, and independent patient outcomes. The Artificial Era thus supplies new instruments for studying relational clinical processes while preserving clear distinctions among machine classification, human clinical accountability, and patient benefit.

Research Design and Comparator Selection

Randomization improves the interpretation of treatment differences by balancing measured and unmeasured baseline factors in expectation. Yet randomization alone does not resolve intervention heterogeneity or define the most informative comparison. A waitlist controls for passage of time less completely than an attention-matched active intervention. Treatment-as-usual may vary across sites and over time. Assessment-plus-feedback is distinct from minimal contact, and an established psychological therapy can be a demanding comparator.

A complete MI trial therefore describes the randomization unit, allocation procedures, treatment content, provider selection, number and length of sessions, fidelity assessment, baseline severity, concurrent services, follow-up, attrition, and analysis population. Cluster-randomized studies require attention to within-cluster correlation; multisite trials should consider site-level differences. The McCambridge and Strang (2004) McCambridge (2004) research and the NIDA CTN community trials National Institute on Drug Abuse (2007) illustrate these design choices.

Claims of causal mediation demand stronger assumptions than simple group differences. Therapist behavior and patient language are themselves post-randomization variables and may be influenced by unmeasured patient characteristics. Sequential coding helps establish order, but an observed path is not identical to an experimentally manipulated mechanism. Carefully designed dismantling, factorial, or longitudinal experiments can improve mechanistic inference.

Measurement: Substance Use, Functioning, Retention, and Harm

Substance-use treatment outcomes operate at several levels. Consumption may be measured in days used, quantity per occasion, heavy-use days, biochemical indicators, or timeline-based reports. Treatment engagement may be measured as initial contact, sessions attended, medication initiation, or retention over a defined interval. Clinically meaningful outcomes also include injury, overdose, alcohol-related problems, social functioning, employment, quality of life, and patient-defined goals. These outcomes are related but cannot be treated as interchangeable.

Self-report is often indispensable because biological tests provide limited observation windows or detect exposure rather than impairment. Repeated self-report may be vulnerable to recall errors and social desirability, particularly under legal or family pressure. Biochemical measurements can add information but require substance-specific interpretation. Trials should define the observation period and avoid using an undifferentiated term such as abstinence when the outcome is only a short assessment window.

In Schwenker et al. (2023) Schwenker (2023), three main outcomes were extent of substance use, readiness to change, and treatment retention at different follow-up periods. That review found insufficient evidence to treat changes in readiness or retention as reliable universal endpoints. Future work should incorporate safety events, equitable access, functioning, and transparent reporting of nonresponse and attrition, alongside consumption outcomes.

Evidence Synthesis: The 2023 Cochrane Update

The most important broad contemporary synthesis for the present question is Schwenker et al. (2023) Schwenker (2023), which updated earlier Cochrane work and included 93 studies with 22,776 participants. Eligible comparisons involved individual face-to-face MI or MET against no intervention, treatment as usual, assessment and feedback, or another active intervention. The review searched through dates in 2021 and November 2022 and evaluated risk of bias and GRADE certainty. Thus its 2023 publication date does not imply inclusion of every study published during 2023–2026.

Against no intervention, the synthesis reported a post-intervention standardized mean difference of 0.48 (95% CI 0.07–0.89; six studies, 471 participants), with substantial heterogeneity (I² = 75%) and low certainty. At short-term follow-up the effect was smaller (SMD 0.20; 95% CI 0.12–0.28; 19 studies, 3,351 participants) and evidence certainty was very low. At the reported long-term comparison against no intervention, the estimate of 0.12 had a confidence interval that included zero (95% CI −0.00–0.25; nine studies, 1,525 participants). These estimates are specific to the review's operationalization and follow-up intervals.

The review reported little or no convincing superiority to treatment as usual or other active interventions, as well as uncertainty concerning readiness and retention. Sixty-nine included studies were judged to have high risk of bias in at least one domain. The proper conclusion is calibrated: MI has evidence of benefit in some contrasts, but comparative and long-term superiority cannot be assumed. Evidence quality and comparator identity are essential components of any summary of effectiveness.

Meta-Analytic History and Apparent Differences Across Reviews

Earlier reviews often reported broader favorable estimates, partly because they included many behavior-change targets and different control conditions. Hettema, Steele, and Miller (2005) Hettema (2005) surveyed 72 clinical trials across target problems and described heterogeneity in effect size by practitioner, population, and context. Lundahl, Kunz, Brownell, Tollefson, and Burke (2010) Lundahl (2010) synthesized 119 studies and reported an average small benefit against weak comparison conditions (g = 0.28), while the average contrast against specific treatments (g = 0.09) was not statistically significant.

The difference between these reviews and the 2023 Cochrane update Schwenker (2023) should be interpreted through scope, comparators, study inclusion rules, follow-up timing, and risk-of-bias criteria, not as a simple historical reversal. A meta-analysis of diverse health behaviors answers a different question from a review focused on individual face-to-face MI for substance use. Nor do standardized mean differences permit a direct conversion into an expected number of patients who will achieve abstinence.

A critical synthesis gives priority to the estimand: what is the intervention's added effect in a defined population, relative to a defined available alternative, on a specified outcome at a specified time? Statements such as MI works without those conditions discard the information that determines whether a result transports to actual clinical decisions.

Process Meta-Analysis: Supported and Disputed Mechanisms

MI process research attempts to identify how therapist behavior might alter later behavior through patient speech and relationship quality. The Magill et al. (2018) Magill (2018) meta-analysis covered 36 primary studies, 40 effect sizes, and 3,025 participants. The review reported a correlation between MI-consistent skills and change talk (r = .55), alongside a correlation with sustain talk (r = .40). MI-inconsistent clinician behavior was associated with more sustain talk; the aggregate technical model received some support.

When change and sustain language were modeled separately, sustain talk was more consistently associated with adverse follow-up behavior than change talk was with improvement. The authors reported that the relational hypothesis was not supported in their aggregated analysis, although relational and intrapersonal moderators partially explained heterogeneity. This complex result is important precisely because it tempers the assumption that producing more change talk is sufficient to alter consumption.

The Moyers et al. (2009) Moyers (2009) sequential coding study provides complementary evidence for a mediational chain. Together, these findings motivate experimental manipulation of clinician responses, stronger temporal measurement, and careful control of baseline patient characteristics. A mechanism remains a scientifically productive model when some links are supported and others unsettled; conceptual clarity is improved by describing those limits explicitly.

Clinical Integration With CBT and Other Behavioral Treatments

MI may prepare the conditions for a patient to consider behavioral strategies, while cognitive-behavioral interventions directly train coping, stimulus control, alternative reinforcement, and relapse-related responses. Combined treatment has been evaluated particularly in cannabis-related programs. The Marijuana Treatment Project (2004) Marijuana Treatment Project Research Group (2004) compared brief MET with a longer package that incorporated MET, CBT, and case management. Because treatment intensity and components differed, observed package effects cannot be attributed exclusively to either MI or CBT.

Integrating MI with cognitive-behavioral work requires a clinically coherent sequence but does not require a standardized ritual of motivation certification. A patient may be ready for practical coping strategies early; another may want to understand options before selecting them. The role of MI is to support deliberation and agency while helping the person use effective interventions.

The VA/DoD 2021 guideline VA/DoD Clinical Practice Guideline Working Group (2021) suggests MET, CBT, or combinations for cannabis use disorder, contingent on patient preferences and available treatment. For alcohol and opioid disorders, the appropriate combination may also involve medication, supervised withdrawal, or other clinical services. Experimental evaluations should compare realistic integrated pathways rather than present MI as a universal stand-alone replacement.

Engagement, Dropout, and Continuity of Care

Early discontinuation is a major treatment challenge, but patient attendance is shaped by more than motivation. Clinic schedules, cost, travel, stigma, occupational obligations, childcare, unstable housing, withdrawal, and previous treatment experiences can interrupt care. A short MI encounter may help a patient clarify the value of continuing treatment; reliable access and institutional responsiveness determine whether follow-through is possible.

The NIDA CTN-0004 design National Institute on Drug Abuse (2007) treated retention as an important outcome. The Cochrane review (2023) Schwenker (2023), however, found evidence for MI's effect on retention to be very uncertain. These observations should be held together: retention is a legitimate research target, but the existence of an intervention aimed at retention does not confirm its efficacy.

Services can examine engagement pathways by recording initial interest, referral completion, first appointment, repeat visits, medication continuity, and re-entry following interruption. Each point may involve different mechanisms. A person who returns after renewed substance use may be demonstrating engagement rather than treatment failure. Research should avoid allowing an administrative attendance metric to override patient safety, functioning, or access to clinically beneficial care.

Ethics, Trauma, Stigma, and Risk Communication

Addiction care often involves experiences of shame, trauma, discrimination, and uneven institutional power. MI's collaborative language can improve the conditions in which patients discuss these experiences, but trauma-informed practice requires attention to actual safety, confidentiality, clinical competence, and the limits of questioning. A patient need not disclose traumatic details for motivational work to proceed. Respect for autonomy also demands accurate representation of treatment choices and their risks.

In severe intoxication, withdrawal, imminent danger, or acute psychiatric instability, immediate assessment and appropriate clinical intervention take priority. Collaborative communication can accompany urgent care, yet the method is not a substitute for assessing capacity, medical emergencies, or mandated safeguarding duties. The NICE guidance National Institute for Health and Care Excellence (2011) illustrates how motivational intervention is situated alongside risk assessment and assisted withdrawal; the Yakovenko et al. (2024) Yakovenko (2024) separates optional psychosocial support from access to opioid agonist treatment.

Stigma may also operate inside research measurement. Questions framed around moral failure can suppress disclosure, and institutional pressure can make self-reported commitment unreliable. Ethically defensible studies describe consent, privacy, consequences of participation, appropriate referral pathways, and protection from service denial. Clinical risks and the patient's long-term interests remain central even when the scientific target is a conversational process.

Digital and Remote Motivational Practice

Telephone, video, secure messaging, and digitally supported sessions can extend access while changing the interactional cues available to clinicians. Silence, hesitation, conversational overlap, and body language may be interpreted differently across media. The current MI manual explicitly discusses remote delivery Miller and Rollnick (2023) Miller (2023), indicating that practice architecture is expanding beyond the physical consultation room.

Transporting the effectiveness estimate from an in-person trial to a digital session requires empirical justification. The 2023 Cochrane review Schwenker (2023) selected face-to-face individual interventions, so its pooled estimates should not be represented as definitive evidence for apps, conversational agents, or asynchronous advice. A digital tool might deliver feedback, support reflection, or help schedule care, but these functions vary in fidelity, human oversight, privacy, and available escalation.

The relevant research questions concern modality-specific engagement, patient preference, reliable recognition of risky states, continuity, culturally and linguistically appropriate communication, data protection, and direct comparison with established care. Digital methods should be tested for incremental benefit and harms rather than equated with a clinical relationship by virtue of using therapeutic vocabulary.

Statistical Bias, Missing Data, and Generalizability

Substance-use research faces attrition, selective enrollment, unblinded clinicians, variable self-report, concurrent treatments, and regression toward the mean. Participants volunteering for counseling may change behavior because of assessment, life events, or expectations before the intervention occurs. Analyses restricted to people completing all sessions can exaggerate apparent benefit when completion is related to prognosis. Risk-of-bias assessment must therefore examine allocation, deviations from intended intervention, missing outcomes, measurement, and selective reporting.

The Cochrane update (2023) Schwenker (2023) rated 69 studies at high risk of bias in at least one domain and emphasized uncertainty. Its large overall sample does not eliminate imprecision within particular comparator and time strata, and combining heterogeneous studies does not create one uniform intervention. Cluster allocation, therapist nesting, and site effects further influence precision, as seen in the community trials and Project MATCH therapist analysis Project MATCH Research Group (1998b).

Generalization requires specifying who was eligible, who refused, how clinicians were trained, and what care the control group received. A trial recruiting university students cannot establish effectiveness for people in inpatient withdrawal care. A result in a well-funded specialty program may not reproduce where medication or follow-up services are inaccessible. External validity should be investigated, not presumed.

Guidelines and Translation Into Clinical Services

Clinical guidelines synthesize benefits, harms, feasibility, values, and resources across interventions. They are therefore relevant evidence about recommended practice but should not be confused with primary causal trials. NICE CG115 National Institute for Health and Care Excellence (2011) includes motivational intervention in initial alcohol assessment and promotes individualized treatment goals. SAMHSA TIP 35 Substance Abuse and Mental Health Services Administration (2019) offers a detailed framework for motivational strategies within substance-use treatment. VA/DoD 2021 guidance VA/DoD Clinical Practice Guideline Working Group (2021) addresses specific indications such as cannabis use disorder.

The principal implementation lesson is proportionate use: MI can enrich assessment and treatment conversations, especially when goals are uncertain or patient and clinician priorities differ, but the appropriate therapeutic pathway depends on diagnosis, severity, risk, preferences, and available services. An institution should not use motivational readiness as an administrative barrier to effective medical treatment.

Different guidelines also operate in different jurisdictions and populations. Recommendations made for veterans, United Kingdom services, or Canadian opioid treatment should be interpreted with attention to the covered population, medical system, publication year, and intervention defined. This monograph compares their scientific implications rather than issuing individual treatment instructions.

Unresolved Scientific Questions and Research Priorities

The first unresolved question is mechanistic specificity: how much of MI's effect arises from the counselor's relationship, technical evocation of patient language, assessment feedback, expectancy, ordinary attention, or subsequent access to effective services? Existing coding studies support some pathways, but reciprocal causality, uncontrolled confounding, and imperfect measurement remain obstacles. Experimental manipulation of defined clinician behaviors and pre-registered mediation models would test these alternatives more decisively.

A second challenge is treatment-effect heterogeneity. Research should identify prospective moderators and replicate them independently, with sufficient statistical power and clinically meaningful effect thresholds. The original Project MATCH findings Project MATCH Research Group (1997a) illustrate the attraction and difficulty of such matching. Trials also need broader geographic, cultural, linguistic, developmental, and severity coverage, together with transparent reporting of treatment history and comorbidity.

A third priority is longitudinal, patient-centered outcome measurement. Abstinence and reduced use may matter, but so do mortality, overdose, functioning, autonomy, quality of life, treatment access, and patient-defined changes. Future pragmatic designs should compare MI with realistic alternatives, test whether feedback and coaching maintain fidelity, evaluate digital and machine-assisted supervision without overstated claims, and determine when MI adds measurable benefit to pharmacological and behavioral treatment pathways.

Critical Scientific Synthesis

The body of evidence supports a bounded account of motivational interviewing and enhancement. MI is a teachable and observable conversational method, and MET is a structured clinical application whose trial versions vary in duration and components. Landmark randomized studies have shown that these approaches can be used in alcohol and drug services; the broad 2023 Cochrane review Schwenker (2023) supports some advantages against no intervention and more limited or uncertain advantages over active comparators. Effectiveness is a property of a treatment contrast under specified conditions, not an intrinsic constant attached to an intervention name.

The mechanism literature offers a similarly conditional picture. The clinician influences conversation; patient language sometimes predicts later behavior; proportions of change and sustain talk have informative associations with outcome. But these associations do not justify a simple formula in which provoking positive statements guarantees abstinence. The 2018 process meta-analysis Magill (2018) and Moyers et al. (2009) Moyers (2009) show how the empirical case is both substantive and incomplete.

The clinical place of MI is best understood as support for shared direction, understandable risk communication, and practical engagement across a wider system of addiction care. In particular, treatments with direct evidence for medically important outcomes must remain accessible irrespective of an individual's willingness to undergo counseling. The scientific contribution of MI research lies in treating motivation as a dynamic interpersonal process that can be observed, trained, experimentally compared, and integrated with the realities of clinical access.

Conclusion

Motivational interviewing and motivational enhancement therapy constitute major, experimentally studied approaches to communication and behavior change in addiction care. Their historical development, structured conversational tasks, patient-language research, and multisite clinical trials establish a coherent field with a substantial evidence base. The most informative contemporary synthesis finds benefits that depend on comparator, follow-up interval, clinical population, and study quality. Treatment effect sizes and causal mechanisms must therefore be interpreted in their actual methodological context.

MI can be used to explore ambivalence, support patient choice, clarify values and goals, and help connect decisions with available actions. MET adds a more structured therapeutic program frequently involving feedback and planned sessions. Neither method provides a universal stand-alone treatment for all addictions, and neither should be made a prerequisite for timely medically indicated care. The future of the field depends on stronger mechanistic experiments, independent tests of moderators, better fidelity measurement, broader cultural and clinical inclusion, long-term patient-centered outcomes, and careful evaluation of new digital supervision technologies.

The established conclusion is an integrated one: addiction-care motivation develops within conversation, personal history, opportunities, treatment relationships, and institutional access. Scientific evaluation must continue to examine this complete sequence. The aim of motivational work is a better-informed, more actionable partnership in change, with clinical effectiveness measured by meaningful outcomes rather than by the persuasive force of a counseling script.

References

  • Amrhein, P. C., Miller, W. R., Yahne, C. E., Palmer, M., & Fulcher, L. (2003). Client commitment language during motivational interviewing predicts drug use outcomes. Journal of Consulting and Clinical Psychology, 71(5), 862–878. https://doi.org/10.1037/0022-006X.71.5.862
  • Ball, S. A., Martino, S., Nich, C., Frankforter, T. L., Van Horn, D., Crits-Christoph, P., Woody, G. E., Obert, J. L., Farentinos, C., & Carroll, K. M. (2007). Site matters: Multisite randomized trial of motivational enhancement therapy in community drug abuse clinics. Journal of Consulting and Clinical Psychology, 75(4), 556–567. https://doi.org/10.1037/0022-006X.75.4.556
  • Carroll, K. M., Ball, S. A., Nich, C., Martino, S., Frankforter, T. L., Farentinos, C., Kunkel, L. E., Mikulich-Gilbertson, S. K., Morgenstern, J., Obert, J. L., Polcin, D., Snead, N., & Woody, G. E. (2006). Motivational interviewing to improve treatment engagement and outcome in individuals seeking treatment for substance abuse: A multisite effectiveness study. Drug and Alcohol Dependence, 81(3), 301–312. https://doi.org/10.1016/j.drugalcdep.2005.08.002
  • Carroll, K. M., Martino, S., Ball, S. A., Nich, C., Frankforter, T. L., Añez, L. M., Paris, M., Suarez-Morales, L., Szapocznik, J., Miller, W. R., Rosa, C., Matthews, J., & Farentinos, C. (2009). A multisite randomized effectiveness trial of motivational enhancement therapy for Spanish-speaking substance users. Journal of Consulting and Clinical Psychology, 77(5), 993–999. https://doi.org/10.1037/a0016489
  • Hettema, J., Steele, J., & Miller, W. R. (2005). Motivational interviewing. Annual Review of Clinical Psychology, 1, 91–111. https://doi.org/10.1146/annurev.clinpsy.1.102803.143833
  • Lundahl, B. W., Kunz, C., Brownell, C., Tollefson, D., & Burke, B. L. (2010). A meta-analysis of motivational interviewing: Twenty-five years of empirical studies. Research on Social Work Practice, 20(2), 137–160. https://doi.org/10.1177/1049731509347850
  • Magill, M., Apodaca, T. R., Borsari, B., Gaume, J., Hoadley, A., Gordon, R. E. F., Tonigan, J. S., & Moyers, T. (2018). A meta-analysis of motivational interviewing process: Technical, relational, and conditional process models of change. Journal of Consulting and Clinical Psychology, 86(2), 140–157. https://doi.org/10.1037/ccp0000250
  • Marijuana Treatment Project Research Group. (2004). Brief treatments for cannabis dependence: Findings from a randomized multisite trial. Journal of Consulting and Clinical Psychology, 72(3), 455–466. https://doi.org/10.1037/0022-006X.72.3.455
  • McCambridge, J., & Strang, J. (2004). The efficacy of single-session motivational interviewing in reducing drug consumption and perceptions of drug-related risk and harm among young people: Results from a multi-site cluster randomized trial. Addiction, 99(1), 39–52. https://doi.org/10.1111/j.1360-0443.2004.00564.x
  • McCambridge, J., & Strang, J. (2005). Deterioration over time in effect of motivational interviewing in reducing drug consumption and related risk among young people. Addiction, 100(4), 470–478. https://doi.org/10.1111/j.1360-0443.2005.01013.x
  • McCambridge, J., Slym, R. L., & Strang, J. (2008). Randomized controlled trial of motivational interviewing compared with drug information and advice for early intervention among young cannabis users. Addiction, 103(11), 1809–1818. https://doi.org/10.1111/j.1360-0443.2008.02331.x
  • Miller, W. R. (1983). Motivational interviewing with problem drinkers. Behavioural Psychotherapy, 11(2), 147–172. https://doi.org/10.1017/S0141347300006583
  • Miller, W. R. (2023). The evolution of motivational interviewing. Behavioural and Cognitive Psychotherapy, 51(6), 616–632. https://doi.org/10.1017/S1352465822000431
  • Miller, W. R., & Rollnick, S. (2023). Motivational interviewing: Helping people change and grow (4th ed.). Guilford Press. https://www.guilford.com/books/Motivational-Interviewing/Miller-Rollnick/9781462552795/contents
  • Miller, W. R., Yahne, C. E., Moyers, T. B., Martinez, J., & Pirritano, M. (2004). A randomized trial of methods to help clinicians learn motivational interviewing. Journal of Consulting and Clinical Psychology, 72(6), 1050–1062. https://doi.org/10.1037/0022-006X.72.6.1050
  • Moyers, T. B., Martin, T., Houck, J. M., Christopher, P. J., & Tonigan, J. S. (2009). From in-session behaviors to drinking outcomes: A causal chain for motivational interviewing. Journal of Consulting and Clinical Psychology, 77(6), 1113–1124. https://doi.org/10.1037/a0017189
  • Moyers, T. B., Rowell, L. N., Manuel, J. K., Ernst, D., & Houck, J. M. (2016). The Motivational Interviewing Treatment Integrity Code (MITI 4): Rationale, preliminary reliability and validity. Journal of Substance Abuse Treatment, 65, 36–42. https://doi.org/10.1016/j.jsat.2016.01.001
  • National Institute for Health and Care Excellence. (2011). Alcohol-use disorders: Diagnosis, assessment and management of harmful drinking and alcohol dependence (CG115). https://www.nice.org.uk/guidance/cg115/chapter/Recommendations
  • National Institute on Drug Abuse. (2007). NIDA-CTN-0004: MET to improve treatment engagement and outcome in subjects seeking treatment for substance abuse [Data registry]. https://datashare.nida.nih.gov/study/nida-ctn-0004
  • Project MATCH Research Group. (1997a). Matching alcoholism treatments to client heterogeneity: Project MATCH posttreatment drinking outcomes. Journal of Studies on Alcohol, 58(1), 7–29. https://pubmed.ncbi.nlm.nih.gov/8979210/
  • Project MATCH Research Group. (1997b). Project MATCH secondary a priori hypotheses. Addiction, 92(12), 1671–1698. https://doi.org/10.1111/j.1360-0443.1997.tb02889.x
  • Project MATCH Research Group. (1998a). Matching alcoholism treatments to client heterogeneity: Project MATCH three-year drinking outcomes. Alcoholism: Clinical and Experimental Research, 22(6), 1300–1311. https://doi.org/10.1111/j.1530-0277.1998.tb03912.x
  • Project MATCH Research Group. (1998b). Therapist effects in three treatments for alcohol problems. Psychotherapy Research, 8(4), 455–474. https://doi.org/10.1080/10503309812331332527
  • Schwenker, R., Dietrich, C. E., Hirpa, S., Nothacker, M., Smedslund, G., Frese, T., & Unverzagt, S. (2023). Motivational interviewing for substance use reduction. Cochrane Database of Systematic Reviews, 2023(12), CD008063. https://doi.org/10.1002/14651858.CD008063.pub3
  • Substance Abuse and Mental Health Services Administration. (2019). Enhancing motivation for change in substance use disorder treatment (Treatment Improvement Protocol No. 35). U.S. Department of Health and Human Services. https://www.ncbi.nlm.nih.gov/sites/books/NBK571071/
  • VA/DoD Clinical Practice Guideline Working Group. (2021). VA/DoD clinical practice guideline for the management of substance use disorders. U.S. Department of Veterans Affairs and Department of Defense. https://healthquality.va.gov/HEALTHQUALITY/guidelines/mh/sud/index.asp
  • Walker, D. D., Roffman, R. A., Stephens, R. S., Berghuis, J., & Kim, W. (2006). Motivational enhancement therapy for adolescent marijuana users: A preliminary randomized controlled trial. Journal of Consulting and Clinical Psychology, 74(3), 628–632. https://doi.org/10.1037/0022-006X.74.3.628
  • Walker, D. D., Stephens, R., Roffman, R., DeMarce, J., Lozano, B., Towe, S., & Berg, B. (2011). Randomized controlled trial of motivational enhancement therapy with nontreatment-seeking adolescent cannabis users: A further test of the teen marijuana check-up. Psychology of Addictive Behaviors, 25(3), 474–484. https://doi.org/10.1037/a0024076
  • West, K. M., Axford, K., Carver, D., Coleman, J. J., Baldwin, S. A., Atkins, D. C., & Imel, Z. (2026). Sustaining therapist motivational interviewing skills with a feedback system augmented by machine learning. Journal of Substance Use & Addiction Treatment. Advance online publication. https://doi.org/10.1016/j.josat.2026.210108
  • Yakovenko, I., Mukaneza, Y., Germé, K., Belliveau, J., Fraleigh, R., Bach, P., Poulin, G., Selby, P., Goyer, M.-È., Brothers, T. D., Rehm, J., Hodgins, D. C., Stewart, S. H., Wood, E., & Bruneau, J. (2024). Management of opioid use disorder: 2024 update to the national clinical practice guideline. CMAJ, 196(38), E1280–E1290. https://doi.org/10.1503/cmaj.241173