Drug and Alcohol Findings home page in a new window Druglink page for May/June 2013

Sometimes best to break the rules

Motivational interviewing is the most influential counselling style in the treatment of substance use problems, promising to sidestep the 'resistance' which characterised client responses to more confrontational methods and nudge them towards desired behaviour changes (study 1). Its originator William Miller reframed resistance as neither the manifestation of an inherent character flaw nor a symptom of disease, but a product of interactions with therapists who impose abstinence goals and stigmatising diagnoses. He developed an approach which explicitly avoided these and other deterrent interactions, instead relying on amplification of aspects of the client's ambivalence about their substance use which cannot be resolved without changing in the direction desired by counsellor and (even if not at first) their client.

As well as an empathic, client-centred overall style, and a focus on harnessing ambivalence, this gave counsellors a distinct list of dos and don'ts to follow: ask open questions, reflect back the client's comments, don't argue, don't warn, don't tell the client what they 'must' do, don't explicitly confront, avoid labelling. The don'ts were intended to avoid the traps which provoke clients to defensively dig in their heels, disengage from the therapeutic process, or simply leave. Imagine then the seismic implications of discovering that in certain circumstances, the opposite was the case: it would cause a re-evaluation of the fundamental understandings on which the approach was built.

That's what happened in the course of an analysis of data from a trial of methods for training clinicians and counsellors in motivational interviewing. The analysis was published in 2005 (study 2) following publication of the main findings of the trial (study 3) the year before. First we'll remind ourselves of the important implications of the main findings.

Workshops not enough

First findings from the trial (study 3) had shown that client responses to trainees changed in the desired direction only when workshops had been reinforced by continued expert coaching and feedback on performance.

Importantly, this US study had recruited an unusually diverse (in terms of initial proficiency) set of addiction counsellors and clinicians who had applied for training in motivational interviewing. Unlike other studies, it did not eliminate the less competent or those who failed after training to demonstrate proficiency, exposing the natural variation in performance obscured by more controlled studies. These trainees were randomly allocated to different training programmes. Most basic was merely being given a training video and manual and being told to train yourself; it had little impact.

A second option was a workshop, during which it was stressed that this was not a complete training regimen, but a platform from which trainees could learn by paying attention to and responding to their clients in their everyday work; signs of commitment to change would indicate they were on the right track, resistance would call for a change of direction. These trainees evidenced post-training improvements in counselling proficiency with an actor simulating a client, but the gains were found to have dissipated four months later when they submitted tapes of their work with real clients. At this stage they were little better than untrained counsellors.

Another set of options followed the workshop with one of three forms of continuing support. One took the form of mailed feedback on sample recordings of the trainee's counselling sessions, comparing their detailed proficiency profile with that of expert practitioners. The second instead took the form of six 'coaching' phone calls initiated by the trainer to ask about any problems and help solve them, each incorporating role-play exercises. The third consisted of both forms of continuing input, meaning that counsellors could not only obtain expert guidance on their problems with clients, but also gain from feedback on their sample sessions.

Each post-training support option largely prevented the deterioration in proficiency seen after just the workshop. But only the third, enriched form of continuing support made enough of a difference to what the trainees did for proficiency also to be reflected in increased 'change talk' (thought the main way the therapy promotes real change) and diminished resistance among the clients. It seemed that the workshop's attempt to initiate self-generated learning was insufficient without an external guide to help trainees recognise when clients were or were not responding well and to offer guidance on how best to respond. The implications were that investment in workshops would not improve client outcomes without this kind of continuing support.

Now the shock

Though a disappointment, the inadequacy of one-off workshops was not entirely a surprise. The main way they had improved proficiency was to eliminate some of motivational interviewing's 'don'ts', but these it seemed had quickly crept back in the absence of continued expert coaching, causing clients to disengage with therapy. Or that at least was the presumption. A second analysis (study 2) set out to test whether these and other therapist behaviours really had affected clients in the expected ways.

This analysis used the post-training, real-client audiotapes from the parent study to relate therapist behaviour to the degree to which clients cooperated with therapy and opened up emotionally and by disclosing personal information – responses which overlap with therapeutic alliance and signify active engagement in therapy.

The first surprise was that client engagement was unrelated to the frequency with which the therapist made statements compatible with motivational interviewing's ethos such as asking open questions – the approach's 'dos'. But engagement was strongly related to embodying the overall spirit of motivational interviewing and to more general social skills not confined to motivational therapists, including empathy, warmth, supporting the client's autonomy, and coming across as 'genuine', an amalgam of seeming open, honest and trustworthy.

This last quality, being genuine, was difficult for raters to agree on from the audiotapes (videos might have helped), but still was about as strongly related to engagement as the other qualities. It also seemed to account for a twist in the findings with potentially far-reaching implications.

Just as the number of 'dos' were unrelated to engagement, so too in the initial analysis were the number of 'don'ts' – instances of the counsellor doing things (such as warning the client) incompatible with motivational interviewing – another surprise. In theory, confronting clients, warning or directing them, and imposing advice, should have provoked clients to resist therapy.

The biggest surprise came when the counsellor's general social skills were thrown in to the analytic pot. Now the frequency of don'ts was significantly and quite strongly related to client engagement, but in the opposite direction to that expected: the more the counsellors 'broke the rules', the better their clients engaged. Moreover, when socially skilled counsellors acted in these ways, they actually enhanced the effect their skills had on client engagement.

The interpretation was that within (and only within) the kind of empathic, caring context these socially skilled counsellors were able to create, doing things such as warning and expressing uncalled for advice and concern deepened the client's engagement with therapy. Socially skilled counsellors tended to avoid these risky manoeuvres, but also had the wherewithal to carry them off without alienating their clients – in fact, the reverse.

To the authors, 'genuineness' seemed the explanation for this conundrum. Therapists who honestly and openly expressed the concerns they were feeling and gave advice they felt the client needed without holding their tongues, or trying to manipulate the client into doing the expressing for them, would have rated higher on being genuine, and perhaps also come across this way to the clients. This quality has long been recognised (study 4) as one of the keys to effective therapy.

Despite intuitively 'making sense', the report's results came from a single study, and even if replicated, should not be taken to give the green light to extreme negative responses contraindicated in motivational interviewing like shaming and sarcasm, indicative less of good social skills and a caring attitude than of the lack of them. Motivational interviewing's insights such as rolling with resistance and avoiding confrontation retain their validity as general principles, but should not be seen as unbreakable rules. And though it might be expected, we do not know if deepened client engagement in this study translated in to stronger commitment to curbing substance use and then in actual change. For example, one component of engagement was expressing emotion, yet this is not always related to better post-therapy outcomes.

It just isn't natural

The findings of this study can better be understood in the light of an evaluation (study 5) of a two-day motivational interviewing workshop for probation staff in Oregon, who gave glowing accounts of the improvements in their understanding of and proficiency in motivational interviewing, a view they sustained over the subsequent four months. Their views were corroborated at the end of the workshop by a paper-and-pen evaluation of how they would respond to sample client statements.

The disappointment came when these in-theory assessments were checked against ratings of audiotapes of how the therapists actually behaved at three stages: before the workshop with an offender client; at the end with someone acting as a client; and with a real offender client four months later. Especially when the raters were assessing overall adherence to motivational principles rather than specific techniques, improvements were slight and left trainees far short of expert practice, largely because they were unable to suppress their previous interactional styles. On one dimension which attempted to reflect how 'genuine' the probation officers were, things had even got worse.

By four months later even the post-workshop boost in use of specific techniques had eroded. Clinching this negative picture was the fact that, compared to pre-workshop tapes, their clients too did not evidence greater commitment to change versus resistance. It seems likely that the natural way a parole officer relates to offenders is quite far removed from motivational interviewing, and reversion to type was the dominant trend. Being trained to go against the grain simply meant raters felt officers were less genuine in their interactions with clients after than before the workshops. Told about this finding, the trainees explained that this new approach felt unnatural. It does not take much imagination to realise that within the undeniably unequal and coercive context of the criminal justice system, adopting an 'It's up to you' stance might feel like a false position, and also feel false to outsiders and offenders.

Another possibly related finding is that motivational interviewing has actually worked best without a manual for the therapist to follow. This was the conclusion of a review by Findings (study 1) which was then confirmed by a synthesis of the research co-authored by William Miller (study 6). He and his colleagues found that of all the differences between motivational approaches including duration, how many motivational-style principles and techniques were said to have been deployed, and therapist training and support, only one was related to outcomes – whether the study had made its therapists follow a manual. Unexpectedly, the relationship was in the 'wrong' direction: manualised therapy had less impact.

Hole in road

The authors cautioned that undocumented differences between the studies where a manual was or was not used could have created a false impression. However, their comments suggest they believe the effect is real and important: "counselors sometimes attend such training in the hope of learning a few tricks to make clients do what they want them to do. MI is nothing of the sort. Rather, it is a complex clinical style for eliciting the client's own values and motivations for change. It is far more about listening than telling, about evoking rather than instilling".

Had they had the featured report's findings to hand, they might have added that the quality of being genuine can suffer from drilling in 'tricks' and in unnaturally withholding normal caring responses, but also that contravening motivational interviewing's tenets is risky unless done by a socially skilled therapist who by doing so conveys rather than contravenes the empathic concern at the heart of good therapy.

Not really a surprise?

In a way, these findings should not be a surprise. Everyone knows the difference between warning and directive advice which conveys and comes from concern for one's welfare and respect for one as an equal, and that which comes from and conveys accusation, denigration, and an attempt to exert control. We also know that the former is likely to be listened to and deepen our relationship with the carer, while the latter signifies an alternative agenda rather than common pursuit of the recipient's welfare. In certain circumstances, entirely avoiding directive advice and warnings can seem as uncaring and unnatural as suggesting to a pedestrian heading blindly towards a deep pit that they consider the pros and cons of moving forward, but in the end it is up to them; the natural and caring response is to shout, 'Stop'. And when that person is actually subject to your direction, to talk to them as if they were not will fail to convince and risks counterproductively sounding false.

For more on these and similar studies see these background notes to an earlier Findings review.


Major source studies Click blue titles to download the Findings analysis or access original documents


1 The motivational hello. Ashton M. Drug and Alcohol Findings: 2005, 13, p. 23–30.

2 How does motivational interviewing work? Therapist skill predicts client involvement within motivational interviewing sessions. Moyers T.B., Miller W.R., Hendrickson S.M.L. Journal of Consulting and Clinical Psychology: 2005, 73(4), p. 590–598.

3 A randomized trial of methods to help clinicians learn motivational interviewing. Moyers T.B., Miller W.R., Hendrickson S.M.L. Journal of Consulting and Clinical Psychology: 2004, 72(6), p. 1050–1062.

4 Evidence-based psychotherapy relationships: Congruence/genuineness. Kolden G.G., Klein M.H., Wang C-C. et al. Psychotherapy: 2011, 48(1), p. 65–71.

5 A small study of training in motivational interviewing: Does one workshop change clinician and client behavior? Miller W.R., Mount K.A. Behavioural and Cognitive Psychotherapy:: 2001, 29(4),p. 457–471.

6 Motivational interviewing.Hettema J., Steele J, Miller W.R. Annual Review of Clinical Psychology:: 2005, 1, p. 91–111.

L10 Web Stats Reporter 3.15 LevelTen Hit Counter - Free PHP Web Analytics Script
LevelTen dallas web development firm - website design, flash, graphics & marketing