Commentary on Witkiewitz et al. (2010): Evidence for matching—at last
Article first published online: 9 JUL 2010
© 2010 The Author. Journal compilation © 2010 Society for the Study of Addiction
Volume 105, Issue 8, pages 1414–1415, August 2010
How to Cite
TOBER, G. (2010), Commentary on Witkiewitz et al. (2010): Evidence for matching—at last. Addiction, 105: 1414–1415. doi: 10.1111/j.1360-0443.2010.03026.x
- Issue published online: 9 JUL 2010
- Article first published online: 9 JUL 2010
- Addiction client-treatment matching;
- Motivational Enhancement Therapy;
- motivational interviewing;
- Project MATCH;
Witkiewitz and her colleagues  confirm three things that clinicians know: motivation matters, matching treatment styles to motivational state matters and dependence complicates things. The experienced clinician recognizes that people who are at different points in reaching a decision to change need different interventions; that discussion of behaviour change interventions with people who have not made a decision to change is likely to elicit resistance, and resistance to change is likely to result in treatment non-adherence. The disciplined application of motivational interviewing has assisted the experienced clinician to prepare people for decisions about behaviour change which naturally leads on to behaviour change therapies. These clinical observations have been supported by empirical evidence [2–4]. Why is it that a rigorously controlled trial such as Project MATCH (Matching Alcoholism Treatment to Client Heterogeneity) has not, to date, supported these things we know? Why, some would ask, does it take the controlled trial approach to research so long to catch up with clinical observation and experience?
Orford  has suggested that this may be due to asking the wrong questions in the wrong way about psychological treatments of addiction. However, in their justification for analysing Project MATCH data differently, Witkiewitz et al.  suggest three other possible reasons for the initial, counter-intuitive findings. The advantage of their approach is to demonstrate that in fact the data were there, they simply needed a different method of analysis. The initial method of analysis might have lacked sufficient power to detect differences because of the chosen method of measuring motivation and the exclusion of people with medium levels of motivation to change. Alternatively, they suggest that the matching hypothesis relating to motivation may have been too simplistic, that interactions between multiple client characteristics need to be examined in the prediction of outcomes. Their third suggestion was that both Project MATCH  and UKATT (United Kingdom Alcohol Treatment Trial)  analytical techniques did not take account of the heterogeneity of drinking outcomes in order to identify discrete patterns of growth. They comment that the obscuring of important relationships can result in null findings.
An earlier attempt to make sense of the lack of support for the matching hypothesis in the original analysis raised the question of the extent of difference between the three Project MATCH treatments . These treatments were intended to be distinct in focus and method ; in fact, the difference could be construed as one of emphasis rather than essence: the weight of attention in motivational enhancement therapy (MET) was on motivation and motivational change, culminating in a behaviour change plan. The emphasis in CBT was on behaviour change, albeit that the first session directly targeted motivation. This was also the case in the 12-Step facilitation treatment. In the UKATT we argued that the differences between the two treatments compared were greater still, in that MET had an individual focus and Social Behaviour and Network Therapy (SBNT) a social network focus, and we proposed explicitly that the source of motivational change was targeted as an internal process in the first treatment and as being located in the external network in the second treatment. In other words, we asserted that motivational change was brought about by a different set of processes. This may be part of the problem: the assumption that only motivational interviewing targets motivational change. If we were right in UKATT, then the recommendations made by Witkiewitz could go further. A better method of assessing and documenting motivational change is required, as is a move away from the assumption that only MET and motivational interviewing target motivational change directly.
In an earlier paper reanalysing Project MATCH data, Witkiewitz  found partial support for the self-efficacy matching hypothesis (individuals low in self-efficacy at baseline would have better outcomes when assigned to CBT compared to MET); in the current reanalysis those individuals who are low on baseline motivation but high on baseline severity of dependence do better with CBT than MET. Implicit in the psychological definition of dependence is the belief in an inability to control consumption and the belief in the inability to abstain from alcohol (low self-efficacy by another name?). The practice of motivational interviewing has not addressed the relationship between motivation and dependence ; rather, it has focused upon the degree of ambivalence about the behaviour and the discrepancy between current behaviour and future goals, cognitions with only a tenuous and partially understood relationship with dependence. Witkiewitz then draws conclusions for a stepped-care approach wherein the cognitive and behavioural tasks facing people with addiction problems complicated by high severity of dependence are addressed either simultaneously or consecutively. This is not new, but gives further empirical evidence for the UK Department of Health promotion of a stepped-care approach to treatment which addresses motivation to change and severity of dependence .
One can only hope that Witkiewitz and colleagues' findings will silence the misinterpretation of previous Project MATCH and UKATT findings that ‘it doesn't matter what you do’. In disputing this erroneous interpretation, Carroll & Rounsaville  have highlighted the difficulties of implementing evidence-based treatments and matching them to client characteristics. Process rating of the delivery of treatments in Project MATCH revealed manual and protocol adherence on the part of the therapists, and this was attributed to centrally provided training that ensured competence to practise from the outset and supervision throughout the study to maximise consistency of delivery and prevent therapist drift . The method was replicated successfully in the UK context . Subsequent studies of treatment as usual, where there has been an absence of such quality control measures, have revealed a lack of protocol adherence or very much of any therapeutic behaviour on the part of the practitioner [16,17]. Experienced clinicians, having breathed a sigh of relief at the findings of Witkiewitz, are given ammunition to address the hard task of implementation of these empirically validated therapies and the respective matching criteria.
- 9Therapies for matching: selection, development, implementation and costs. In: BaborT., Del BocaF., editors. Treatment Matching in Alcoholism. Cambridge: Cambridge University Press; 2003, p. 42–61., , , ,
- 11Motivation and barriers to change. In: ToberG., RaistrickD., editors. Motivational Dialogue: Preparing Addiction Professionals for Motivational Interviewing Practice. London: Routledge; 2007, p. 16–33.
- 12Review of the Effectiveness of Treatment for Alcohol Problems. London: Department of Health; 2007., ,
- 17Treatment as usual. In: MacGregorS., editor. Responding to Drug Misuse: Research and Policy Priorities in Health and Social Care. London: Routledge; 2010, p. 40–52., , ,