Making Change Visible: The Possibilities in Assessing Mental Health Counseling Outcomes

Article excerpt

Most professions have obvious measures of success: Businesses have sales, sports teams have fans and victories, lawyers have courtroom victories, engineers and architects have constructions, farmers have food products, mechanics have working cars, doctors have saved lives, and priests have congregations and alms. Mental health counseling is nearly unique in that its product is comparatively invisible. Even education has an observable product--student performance and grades. Perhaps it is only natural that people are increasingly demanding to see evidence of success in the counseling field. In particular, mental health funding sources and managed care organizations are demanding counselor accountability as a requisite for funding (Brown, Dreis, & Nace, 1999; Kelly, 1996; Lambert, Strupp, & Horowitz, 1997; Whiston, 1996; Wylie, 1994). The purpose of this article is to highlight advances made by psychotherapy researchers in their quest to make counseling outcomes more visible. I also delineate technical problems in ensuring accurate outcome assessment, highlight areas of consensus among outcome researchers to safeguard against those complications in the most practical way, and address the prospect of predicting outcome success using current methods.

The Drive Toward Credibility Through Outcome Research

The desire to make counseling success explicit has led to cost-effectiveness research to link counseling to increased work productivity (Zhang, Rost, Fortney, & Smith, 1999) or decreased physical medical care (Buchanan, Gardenswartz, & Seligman, 1999; Rainer, 1996). Other counseling researchers have tried to develop a set battery of measurements that could show both the need and the success of psychotherapy (Strupp, Horowitz, & Lambert, 1997; Waskow & Parloff, 1975) in a standardized way and perhaps at least approach the kind of standards of measurement that exist in physical health treatment (e.g., a normal body temperature of 98.6[degrees] Fahrenheit). In addition to having accepted methods to measure success, advantages accruing to increased standardization of outcome research (e.g., method of data collection, instrumentation, research design) would allow study comparisons and would hasten the development of a credible and solid knowledge base about effective psychotherapy (Hill & Lambert, 2004).

Developing Signposts About the Rate of Client Change

Knowing about treatment course would conceivably be helpful to the counselor, the client, and the payer. All parties could anticipate their respective levels of investment. Questions of how long and for whom counseling would be helpful might be answered. The research to date already provides a general, albeit tentative, answer regarding expected treatment course. For example, Finch, Lambert, and Schaalje (2001) examined more than 11,000 clients across a number of treatment sites and found a similar decelerating growth curve. The cumulating evidence suggests that although the largest gain occurs early in treatment, more serious levels of disturbance have better results as the benefits of treatment continue to accumulate.

Howard, Kopta, Krause, and Orlinsky (1986) summarized the dose--effect of psychotherapy, using data collected from 2,400 clients during a 30-year period. Results of their probit analysis revealed that 29% to 38% of the clients measurably improved after 1-3 psychotherapy sessions; after 4-7 sessions, 48%-58% improved; after 8-16 sessions, 56%-68% improved. About 75% of clients improved after 6 months of weekly treatment, and 85% improved after 1 year of weekly treatment. A more refined analysis revealed that 50% of the clients diagnosed with anxiety or depression had improved by Sessions 8-13. In contrast, clients diagnosed as being borderline psychotic did not reach similar levels of improvement until Sessions 13-52.

As a general rule of thumb, evidence has suggested a relationship between initial self-reported disturbance level and treatment course. …