Gaume J., Magill M., Longabaugh R. et al.
Alcoholism: Clinical and Experimental Research: 2014, 38(7), p. 2138–2147.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Gaume at email@example.com. You could also try this alternative source.
Swiss study of brief alcohol interventions with a representative sample of heavy drinking young men exposed the determining influence on later drinking of the practitioner’s competence in motivational interviewing and how they behave in the session.
Summary Swiss men aged 20 are subject to conscription into the army via an assessment of fitness to serve, offering a chance to test a brief alcohol intervention among a representative sample of young men rather than the more typically studied college or patient populations. The study site was an army intake centre serving the French-speaking region of Switzerland. Conducted along motivational interviewing lines, the alcohol intervention lasted 20 to 30 minutes, during which the intention was to address drinking and its consequences and, if the client agreed, to plan changes in drinking behaviour.
To reveal the influence of the practitioner, this Swiss study of brief alcohol interventions with heavy drinking young men recruited clinicians widely differing in experience of motivational interviewing and left them to their own devices.
Findings show that such counselling does not automatically have any significant effect; impacts depend on the competence of the practitioner and how they behave in the session.
Implications are that frequent superficially motivational remarks fail to connect with clients, while just a few which show a real attempt to understand have a positive impact, and just one which fails to be client-centred can render the session ineffective.
Among conscripts who had regularly consumed at least 60g of alcohol at a sitting (most had), a previous study found that six months later such an intervention had reduced consumption by roughly 20%. However, the featured study was less concerned with the effect of the intervention, than with the degree to which its impacts were affected by counsellor characteristics. For that reason, the intervention was not standardised. Though all had been trained in motivational interviewing, counsellors were free to deliver the intervention as they saw fit.
The 18 counsellors in the study were physicians and psychologists from a local alcohol treatment centre, selected to differ in sex, professional status, clinical experience, and experience of motivational interviewing. The aim was to generate sufficiently wide differences between them for these (if they were going to at all) to affect the outcomes of the interventions. In addition, they were asked to rate their confidence in the effectiveness of brief motivational interventions and in their ability to deliver these, and their counselling sessions were recorded and analysed for the degree to which their comments were consistent with motivational interviewing.
Of the 831 conscripts asked to join the study, 637 agreed. Of these, 431 met the study’s criterion for hazardous drinking – scoring at least four on the three questions of the AUDIT-C screening questionnaire. They were allocated at random either just to undergo baseline assessments for the study (the control group), or to these plus motivational counselling. On joining the study they averaged about 50g or just over six UK units of alcohol on each of the two days a week they drank, and a third were probably dependent, but on average there was little desire to alter drinking patterns. Three months later about 82% could be followed up to assess the impact of the interventions on drinking and the degree to which the counsellor had affected these impacts.
First it was established that, relative to the control group, the intervention had slightly but significantly reduced both average drinking days per week and a composite of this plus measures of the intensity of drinking on those days.
Next the analysts investigated whether on the composite drinking measure, conscripts had reacted differently to different counsellor characteristics. Whether the counsellor was a doctor or a psychologist made no difference to the impact of their counselling, but all the other assessed characteristics were significantly related. The more experienced (either clinically or in motivational interviewing) half of the counsellors reduced drinking relative to the control group, but the less experienced half did not. Similarly, when they were divided into two halves, only counsellors more confident about the effectiveness of the intervention or of their ability to deliver it improved on trends in the non-counselled control group. There were ten female counsellors and they did not improve on control group trends, while the eight men did.
Tested next (this time by dividing sessions into top and bottom halves) were the counsellors’ motivational interviewing skills. In line with expectations, only sessions in which counsellors had been rated as relatively competent in motivational interviewing’s client-centred counselling style, or during which they had avoided comments inconsistent with the approach (these were very rare), improved on drinking trends among the control group. Findings on the degree to which counsellors reflected back the client’s comments were more complex. Though such comments remained in the minority, when a relatively large proportion of reflections conveyed different or additional meanings rather than just more or less echoing the client, drinking was reduced; on average, other sessions did not significantly affect drinking. Contrary to expectations, sessions in which counsellors more often made comments considered consistent with motivational interviewing failed to reduce drinking, while those in the ‘worst’ half on this measure were on average followed by drinking reductions.
These factors interacted such that only the more clinically experienced male counsellors reduced drinking – not women, and not less experienced men. Those who were both experienced and who felt confident in their abilities reduced drinking more than other counsellors. Sessions during which counsellors had been rated as relatively skilled in motivational interviewing were followed by reduced drinking, regardless of whether the counsellor was also relatively experienced in the approach.
In this study where the clients were young male conscripts who generally saw little reason to change their hazardous drinking, a brief intervention based on motivational interviewing produced the best results when delivered by male counsellors with relatively extensive clinical experience, and when the counsellor felt relatively confident in their ability to deliver brief motivational interventions and was relatively committed to the approach.
Best results too came from interventions characterised by relatively high proficiency in client-centred counselling and motivational interviewing. Proficiency tended to come with experience, but it was the skills which were important, not experience in itself. Even one comment inconsistent with motivational interviewing, such as unsolicited advising or confrontation, seemed to nullify the intervention. The quality and the exact combination of skills seemed to matter more than the quantity. Using a high number of open questions and simple reflections without eventually showing support or in-depth understanding through more meaningful reflections might not be enough to change alcohol-related belief systems.
It should be remembered however that reverse causality cannot be excluded – the possibility that it was client behaviours which influenced therapist behaviours, not the other way round. Also these results were found in a distinct population of generally ‘binge’ drinking young men and may not be replicated in other contexts.
commentary Creating a clinically relevant ‘real-world’ scenario, this study was unique in deliberately recruiting clinicians of widely differing experience and then leaving them to their own devices. Its findings show that counselling young adults about their heavy drinking does not automatically have any significant effect. Though all the counsellors had been trained in the intervention’s motivational approach and generally adhered to its client-centred style, it only worked when delivered by what for those drinkers were the right counsellors – more experienced practitioners confident of their motivational interviewing abilities and in the efficacy of the approach, and/or who were especially proficient – an amalgam of demonstrating acceptance of and empathy with the client and embodying the collaborative spirit of the approach.
Micro-analysis of the sessions paints a picture of any number of perhaps superficially positive or affirming remarks failing to connect with clients, while just a few which showed the counsellor really was trying to understand them made a positive impact. On the debit side, just a single remark which demonstrated that the counsellor was not on and by the client’s side, but pursuing their own ‘Doctor knows best’ agenda, rendered the session ineffective.
The implications of this study are explored further below and in the Alcohol Treatment Matrix cell concerned with the influence of the practitioner in screening and brief interventions. In interpreting its findings it is important to bear in mind that it was difficult to tease out the impact of counsellor characteristics which interacted and/or co-varied, and also to identify which influences were generalisable to other contexts. For example, men had better outcomes than women, but this may have been due to the all-male caseload, and in any event, men only did better when they were also relatively highly clinically experienced. Longer experience was in turn related to being a doctor rather than a psychologist, and nearly all the women were psychologists and the men doctors.
One implication of the findings is that in itself experience is not as important as the competence and perhaps too the confidence which can (but not always) come with it. Though inexperienced in motivational interviewing, counsellors who are either ‘naturally’ skilled, or far enough along this road to have rapidly absorbed the approach, will on average reduce drinking. In contrast, though they may be experienced, counsellors who have not developed a corresponding level of competence will be ineffective.
These findings are in line with other studies which suggest that recruiting clinicians who take naturally to a client-centred approach, but have not been trained in motivational interviewing, would be better than training less naturally adept clinicians. In another Swiss study, but this time of emergency department patients, counsellor qualifications, experience and training were equalised, yet still therapists varied widely in their outcomes, from an average 18 UK units (each about 8g alcohol) more per drinking week to an average nine-unit reduction.
Variability in outcomes achieved by brief interventionists is not uncommon. What makes the difference is it seems not necessarily the approach itself, nor the therapist’s strengths and weaknesses, but the combination of approach and therapist; some who do well with a more directive approach do badly when they attempt the client-centred style of motivational interviewing.
In the featured study, so-called ‘complex reflections’ – the times when the counsellor reflected back the client’s own feelings or comments, but with a spin which extended or deepened their meaning – seemed particularly important. When these formed a very small portion of all the reflections, the brief intervention made no difference to drinking; when a larger but still small portion, drinking was reduced. In surprising contrast, simply accreting more of the other responses considered compatible with motivational interviewing actually seemed counterproductive. Such comments were common, occurring on average once or twice per minute of the sessions.
The other side of the equation was counsellor comments incompatible with motivational interviewing. These were very uncommon – usually one or none per session – but when they happened, that session was no more effective at moderating drinking than no counselling at all.
Surprising as they are, these are not isolated findings. In the Swiss emergency department study referred to above, the brief interventionist who most often used recommended motivational interviewing techniques ended with the worst drinking outcomes. Another study of brief alcohol counselling of Swiss army conscripts investigated what generates comments indicative of intention to reduce drinking, considered the main way motivational interviewing affects behaviour. Top of the list were complex reflections, while accreting more of the other motivational-style comments such as open questions, simple reflections, and affirming and supporting, were like the proverbial water off a duck’s back, not moving the conversation in any particular direction. Explicitly non-motivational comments, like confrontation and directing or advising the client, counterproductively prompted comments indicative of intention not to change one’s drinking. An omnibus analysis of two Swiss army conscript trials and a US trial confirmed the negative impact of these kind of comments.
The counterproductive impact of confrontation with heavy drinkers not seeking treatment emerged early in a seminal study led by motivational interviewing’s originator, William Miller. For him and his colleagues, confrontation is one of those counsellor behaviours considered incompatible with the client-centred core of motivational interviewing. What they share is the non-collaborative stance of someone who knows best and is therefore in a position to confront, warn, direct, or advise the client.
Thanks for their comments on this entry in draft to research author Jacques Gaume of the Alcohol Treatment Center in Lausanne, Switzerland. Commentators bear no responsibility for the text including the interpretations and any remaining errors.
Last revised 17 December 2015. First uploaded 11 December 2015
REVIEW 2005 The motivational hallo
REVIEW 2018 Congruence/genuineness: a meta-analysis