Best Practices : Substance Abuse Treatment and Rehabilitation

2. Matching Clients with Treatments and Therapists

The notion that some clients may do better in some treatment and with some therapists than others has been of concern for some time. The Institute of Medicine report (1990) and a variety of studies have suggested that this is the case.

One relevant study was done at the Donwood Institute in Toronto by Dr. John McLachlan (McLachlan, 1972). It involved alcoholics who had been treated by therapists who were more or less "directive" in their treatment styles. The most directive therapists tended to have clear rules, to be very consistent and to run highly structured treatment sessions. Less directive therapists were more "laid back" and encouraged more self-expression and autonomy. By chance, some patients assigned to different types of therapists differed in what McLachlan called "conceptual level". Basically, this referred to differences in personality and thinking styles resulting in differences in the need for structure and guidance.

McLachlan followed up patients treated by different therapists and found that those whose conceptual levels matched therapists' styles did better than others. Thus, patients with a need for structure did best with directive therapists and vice versa. On the other hand, "mismatched" patients did less well with both types of therapists.

A variety of other positive matches has been found involving different therapies and the following client characteristics: gender, ability to recognize problem drinking situations, beliefs about alcoholism, family history of alcoholism, sociopathy, locus of control and self-image (see Mattson, 1994 for a review). However, most of these matches have not been consistently replicated.

The only large-scale study of client-treatment matching was Project MATCH (1997). This multi-site United States study was designed to test whether different types of alcoholics respond differently to different type of treatment. The treatment types selected for study were: 12-step facilitation where clients were encouraged to join AA; cognitive behavioural therapy, based on social learning theory; and motivational enhancement therapy, based on motivational psychology. Patients were randomly assigned to treatment and, in ex post analyses a variety of hypotheses was tested concerning interactions between treatment types and severity of alcohol involvement, cognitive impairment, psychiatric severity, conceptual level, gender, meaning-seeking, motivational readiness to change, social support for drinking versus abstinence, sociopathy and type of alcoholism.

The Project MATCH Research Group (1997) reported that only one hypothetical "match" was clearly supported by the data. This proposed that clients with low psychiatric severity would do best in the 12-step facilitation condition. The results showed that these clients had more abstinent days during six and twelve months of follow-up than those treated with cognitive behavioural therapy. However, the extent to which clients in either condition became involved in AA during the follow-up period has not been reported.

One explanation for the failure of Project MATCH to find more interactions is that there were "ceiling" effects due to client selection and the use of three quite powerful interventions. The overall results of the study were quite impressive and the majority of clients in each condition showed significant and sustained reductions in alcohol use over the follow-up period. The selection of socially stable and research-compliant clients, together with the high quality of the three types of treatment, certainly contributed to these results and may have left little room for matching effects to show up.

Project MATCH is considered by some to have laid the matching hypothesis to rest, but others remain unconvinced (e.g. Glaser, 1997). Among other concerns, the study has been criticized for not really "matching" clients to treatment because clients had no say in the treatment they received. Also, important matches involving therapist characteristics, or pharmacological treatment, were not tested. Of course, the study focused only on people with drinking problems and not on those with other drug problems.

Large-scale studies that addressed the issue were conducted during the 1970s by Sells and Simpson (Sells, 1974; Sells and Simpson, 1976), and in the 1980s by Hubbard et al. (1984). In general, these do not provide strong evidence for the benefits of matching drug users to treatment, but suggest that, controlling for client characteristics, treatment types in common use in the United States (methadone, therapeutic community, drug-free outpatient) may be equally effective.

Overall, the evidence in favour of matching clients to treatment methods is currently rather weak. However, this does not mean that one treatment will suit all. Clients require an individualized, flexible approach to address a variety of needs. Some clients need services for mental health problems, others require help with employment and other social problems, and some will need temporary or longer-term shelter. Attention to these problems is essential if those involved are to achieve and maintain improvements in substance use behaviours.

Guidelines for the selection of appropriate types and levels of care should therefore be developed and evaluated. As indicated in Profile - Substance Abuse - Treatment and Rehabilitation in Canada, such guidelines are being developed in a number of provinces/territories.

Best Practice Guideline (No. 7)

Although the literature does not yet provide strong evidence by which to match clients to specific treatment interventions, it does not mean that all clients require the same types of services. A variety of flexible and individualized services is required and guidelines for the selection of appropriate services are needed.