The effectiveness of therapy

By Martin E.P. Seligman, PhD
APA President

My main campaign promise, to see that the funding to launch major studies on the effectiveness of therapy is made available, has been fulfilled. The National Institute of Mental Health (NIMH) has agreed to fund such studies on a massive basis. Effectiveness research promises to be a significant aid in the fight against profit-driven health schemes for the treatment of mental illness in the next century.

The distinction between an efficacy study and an effectiveness study is crucial. An efficacy study is a laboratory distillation of treatment. Efficacy studies consist of a brief and fixed duration of therapy, manualized treatment, a single simple disorder and random assignment of clinic volunteers to type of treatment. Efficacy studies have generally shown that under such well-controlled conditions a brief course of specific psychotherapy or of medication works reasonably well for a variety of disorders. Such studies have good internal validity, but their findings can generalize poorly to treatment as it is actually done in the real world. Effectiveness studies such as the well-known Consumer Reports study, on the other hand, look at the outcome of treatment as it is actually delivered in real life. Treatment is not necessarily brief or of fixed duration. It is determined by the clinician?s judgment, the patient?s decision, progress toward goals, and life changes. There is no manual to follow, and the therapist can adjust tactics to accord with the patient?s progress. Disorders are complex, and patients choose a type of therapy they believe in and choose a therapist they believe will help them.

The results of effectiveness studies differ in several crucial ways from the results of efficacy studies: First, there seems to be substantially more patient improvement in effectiveness studies. Second, there seems to be little specificity of the type of therapy or drug for any disorder in effectiveness studies, but high specificity in efficacy studies. Third, effectiveness studies robustly document that the longer the therapy, the more improvement on every outcome variable. Fourth, effectiveness studies show that health-care plans that limit duration of therapy hurt the outcome of therapy.

It is these discrepancies that have led the National Institute of Mental Health under the leadership of Steve Hyman, MD, Grayson Norquist, MD, and Barry Lebowitz, PhD, to sponsor a major new program of outcome research. No dollar figure has yet been announced, but there is reason to believe that at least $20 million a year will be spent. Managed care is now deeply cutting the quality of mental health services in order to make money, and there is no adequate data to bridle such cuts. Efficacy studies, suited only to testing brief and inexpensive treatments, can only bolster the use of brief and inexpensive treatments. The existence of a definitive body of effectiveness research, however, is likely to complement efforts currently under way to ensure that the health-care marketplace considers quality as well as cost issues. This is not a short-term fix for the present health-care crisis. This is the long-term solution.

What will happen next? In the next year we will see a call for proposals for effectiveness research on psychotherapy and drugs. Fierce competition for the research dollars will ensue, and NIMH?s standards will be very high. The outcome of that competition will be crucial to the future of psychotherapy. It is often the case that the question asked determines the answer given. If drug researchers get these grants, it is likely that 10 years from now we will see guidelines that tell us that drugs are the treatments of choice. If managed-care companies and their researchers get these grants, this will further fuel their effort to recommend the least expensive and briefest of treatments.

If, however, our science and practice communities now come together to do this research, the outcome is likely to be optimal for our providers and for our patients. The Science and Practice directorates are working together to communicate with NIMH about possible collaborative efforts.

At the initiative of Bob Brown of the Board of Professional Affairs, there was a recent meeting of Science and Practice Directorate staff, several members of BPA, the Committee for the Advancement of Professional Practice (Ron Fox, chair) and the Board of Scientific Affairs (Paul Sackett, chair). They agreed that a planning meeting this spring?followed by a larger working group meeting in the fall including APA and NIMH staff, practition-ers and methodologists?would be useful to develop these initiatives.

All of the following can now be targets of discussion between APA and NIMH: well-funded statewide practice research networks, conferences to broaden and deepen the assessment of outcomes beyond mere symptoms of disorder, summer workshops to train a new cohort of researchers in effectiveness methodology and in effectiveness-efficacy hybrid designs, collaborations between psychoanalytic therapists and outcome scientists, the investigation of doctoral-level training in the outcome of treatment and of psychologists in primary care, effectiveness of psychological care in the relief of physical illness, and the role of large-scale preventive programs on the mental health of the nation.

This is a splendid opportunity for science and practice to work together, the best opportunity that I have seen in my lifetime.