Go to Home Page

The Next Step in Client-Therapist Matching

Using Prescriptive Psychotherapy

The Major Problems with Client Assessment in TMatch

The first major problem with client assessment for Prescriptive Psychotherapy matching suggestions in TMatch were that, except for Resistance, the assessment questions were created from scratch, without any studies of validity or reliability. Such studies could have lead to successive refinements. The second problem was that threshold levels of each client characteristic that triggered matching recommendations were based on guesses. For example, it was arbitrarily assumed that the level of Resistance that triggered matching suggestions for Resistance were triggered by any Resistance level slightly over the center of the scale used (the Therapeutic Reactance Scale). The third problem was that there was no information available as to what level of a client characteristic would trigger what level of therapist characteristic. Again, in TMatch this was based on guesses. The solution used were logarithmic formulas (base e), which were chosen to give proportionally more weight to lower scores above matching thresholds.

The Next Step for Client Assessment

Beutler and associates have created several elegant assessment instruments in connection with their Prescriptive Psychotherapy systems that would be extremely useful in connection with matching clients to therapists. Many of these instruments are part of their Systematic Treatment Selection (STS) software, which assesses clients on the four areas of client characteristics listed above, and then creates appropriate treatment suggestions for therapists. This client assessment would be extremely helpful for use in client-therapist matching based on Prescriptive Psychotherapy principles, since this matching starts from the same place as the treatment matching, which is assessing clients. The first version of TMatch was based on self-assessments. In this connection, there is an STS client self-rating questionnaire, which theoretically would work very well as a starting point for client-therapist matching. An additional part of STS is a Clinician Rating Form, which would be useful if clients were assessed by clinicians first and then matched to appropriate therapists. Although I have a copy from Dr. Beutler of the questions on both of these forms, they are not completely useful without in addition having the method of rating these questions onto the appropriate scales, and in addition knowing what levels on each scale that trigger treatment recommendations. For example, there are many questions for determining coping style on both of these instruments, but there are no instructions for turning the answers into numerical coping style ratings. Even if these instructions were available, without instructions for what levels of what coping styles triggered exactly what treatment suggestions, I would have to guess or try to recreate from scratch what Beutler and associates have already worked out through years of study. If this information was available, these client assessments would be an ideal starting point. However, if a client-therapist matching method were to include other criteria besides Prescriptive Psychotherapy recommendations, such as client preferences, either of these assessment instruments would probably be too long, and would have to be shortened by some method.

The Major Problems with Therapist Assessment

Therapist assessment for matching is a more difficult problem. Prescriptive Psychotherapy and STS make treatment suggestions, not therapist suggestions. An underlying assumption is that therapists have the abilities to modify their treatments according to these suggestions. Therapist matching in TMatch assumes that therapists have different styles and abilities, and different therapists will match better with clients based on the client ratings on these dimensions. In TMatch, therapist assessments were created from scratch, without any validity assessment or refinement. In addition, as explained above, levels of therapist characteristics that matched client characteristics were based on guesses, not on available statistical information. Of course, the major problem with therapist assessment, which has been mentioned many times on this web site, is that recommendations for treatments are only translatable into therapist matching recommendations if therapists are reliably different in treatment styles and therapy emphasis corresponding to these recommendations, and these differences can be reliably and validly assessed.

The Next Step for Therapist Assessment

Beutler and his associates have two instruments that are used in connection with STS that might be excellent starting points in creating an instrument that could be used for therapist matching. These are the STS Therapy Rating Scale, and the Prescriptive Therapy Adherence Scale. These instruments assess how well therapists followed treatment suggestions. They have excellent questions, and could be rewritten as therapist self-assessment instruments for matching based on self-assessments, such as was done in TMatch. Alternatively, therapists could be matched based on observer ratings using these instruments with much less change. These two systems could even be combined, with therapists self-assessments being tested by observer ratings. There is another endeavor by Beutler and his associates that would be useful for matching. For clinics using STS, they have an algorithm that assesses therapists relative successes with clients with different profiles, and successfully predicts based on this which therapists will match best with which clients in the future. These therapists would be an excellent pool for creating an effective therapist assessment instrument, and for determining which therapist characteristics are appropriate for assessment in this context.


For more information, or to give feedback, email Kenneth Frankel, Ph.D.