Why Therapists Don’t Improve? … And How They Could
Posted on July 5, 2017 by Dr. Bruce Wampold
How time flies. I have been researching, practicing and supervising psychotherapy for over 35 years. When I began graduate school Hans Eysenck’s claims that psychotherapy was not effective, and likely harmful, was widely disseminated and believed. To say the least, it was not an optimistic time to be in training to become a psychologist. Fortunately, during my first semester Mary Lee Smith and Gene Glass published a meta-analysis in the American Psychologist (one of the first ones ever published) of nearly 400 controlled studies of the out of psychotherapy. The results of that study were profound: Smith and Glass found that psychotherapy—many different types of therapy—were remarkably effective! Approximately 80% of clients receiving psychotherapy will be less distressed than those who did not receive psychotherapy.
Psychotherapy is remarkably effective
Many studies since Smith and Glass’s meta-analysis have confirmed that psychotherapy is remarkably effective. Many clinical trials comparing psychotherapy to psychotropic medications have found that psychotherapy is as effective as medications, but with many advantages over medications. Psychotherapy is longer lasting—that is, there is decreased likelihood of relapse when psychotherapy is finished than when medications are withdrawn. In fact, it seems that medications increase the likelihood of relapse, even when combined with psychotherapy, whereas psychotherapy effects persist long after treatment is terminated. Each additional course of medication is less effective than the previous one whereas additional psychotherapy over time remains effective. And of course, psychotherapy does not produce the disagreeable side effects produced by medications, such as loss of libido, weight gain, and sleep disturbance. As well, the evidence indicates that therapists in practice achieve outcomes comparable to those achieved in clinical trials, where many factors that make therapy challenging are controlled (e.g., comorbidity). Psychotherapy also seems to be efficient—generally, clients use a sufficient amount of therapy to achieve desired outcomes and then terminate.
But, we can—and need to—do better
So much for the good news. There is also some troublesome evidence. Evidence-based medicine uses a statistic called NNT—Number Needed to Treat. NNT is the number of patients who need to receive a particular treatment to have one better outcome than the alternative, which may be another treatment, a placebo, or no treatment. Smaller NNTs indicate more effective treatments. The NNT for psychotherapy is 3, when compared to no treatment. That is, for every three clients receiving psychotherapy, only one would have a better outcome than if they had received no therapy and two clients did not have a better outcome.
This sounds terrible, but one needs to recognize that the NNT for psychotherapy is smaller (ie. superior) than most accepted medical treatments in many areas, including cardiology (eg. statins for heart disease prevention NNT = 60), pulmonology (eg. antibiotics for acute bronchitis NNT = 6), prophylaxis (eg., flu vaccine NNT = 12), pediatrics (eg. antibiotics for acute ear infection—no fewer serious complications, no less re-occurrence of infection, NNT = 20 for reduced pain at 1 to 7 days, and NNT = 9 for harm, mainly diarrhea). Yet, it is disturbing to know that only one in three psychotherapy clients benefit from psychotherapy, which means we can—and need to—do better.
Improvement through evidence-based treatments
Several means to provide better mental health services have been proposed. First, clinical scientists have recommended disseminating evidence-based treatments. Essentially, the claim is that if all therapists provided evidence-based treatments to their clients, more clients would benefit. Several venues have implemented variations of this idea, by incentivizing or mandating particular treatments, but such efforts have generally failed, despite large investments in money. This is not surprising because there is no evidence that any one form of psychotherapy is superior to another.
Improvement through Routine Outcome Monitoring (ROM)
The second means to improving mental health services is to provide therapists with information about patient progress in therapy. This idea, which is known by many names, including "feedback" and "routine outcome monitoring," involves measuring the mental health status of the client every session and providing this information to the therapist (and the client, in some systems). Feedback systems provide the therapist with signals that the client is at risk for failure (deterioration or less than expected progress). Many studies have examined these feedback systems vis-à-vis treatment as usual (treatment without feedback)—there is evidence that such systems improve outcomes, primarily by reducing the rate of failure of cases with signals, but the evidence is not strong. Moreover, there are challenges to implementing feedback systems. But the most troublesome issue is that such systems do not seem to help therapists become better therapists. In research that one of my students conducted, we found that over a period of up to 18 years, the outcomes of therapists, who practiced in an agency using feedback, deteriorated. The effect was small, yet is disturbing that therapists do not improve as they gain experience and see more clients! Hopefully, though, some therapists did improve, which leads to the third way that mental health services might be improved—by focusing on therapists.
Improvement through deliberate practice
There is strong evidence that some therapists consistently achieve better outcomes than other therapists. These therapist effects exist across therapies—that is, there are more effective (and less effective) cognitive behaviour therapists, psychodynamic therapists, emotion-focused therapists and even more effective (and less effective) psychiatrists administering antidepressants. These therapist effects are present regardless of the characteristics of the clients treated. We know some clients are more challenging than others; therapist effects actually are more pronounced for more severely distressed clients. Differences among therapists are important—if the clients of the more poorly performing therapists (say the bottom 5% to 10%) were reassigned to the other 90% to 95% of therapists, dramatically more clients would recover!
The evidence about therapists suggests that if we could assist therapists to improve, the quality of mental health services would increase dramatically. The Calgary Counseling Centre, a community counseling agency in Calgary, has worked diligently to help therapists improve. And it has worked! Over time, outcomes at the Calgary Counseling Center have improved. This improvement was due to the fact that the therapists within the agency improved.
The evidence suggests that if we are to improve the quality of mental health care, individual therapists must continue to improve. Such improvement involves the deliberate practice of particular skills. In my next post I will discuss the skills that characterize effective therapists as well as how therapists can practice these skills to improve, using the principles of deliberate practice.
About the Author
Dr. Wampold is the Patricia L. Wolleat Emeritus Professor of Counseling Psychology at the University of Wisconsin—Madison and Director of the Research Institute at Modum Bad Psychiatric Center in Vikersund, Norway. He is a Fellow of the American Psychological Association, Board Certified in Counseling Psychology by the American Board of Professional Psychology, and is the recipient of the 2007 Distinguished Professional Contributions to Applied Research Award. His current work is summarized in The Great Psychotherapy Debate (with Z. Imel, Routledge, 2015).