A study from a prestigious psychology journal recently crossed my desk. It found that therapists who provide cognitive behavior therapy or CBT—including the most experienced therapists—routinely depart from the CBT techniques described in treatment manuals. “Only half of the clinicians claiming to use CBT use an approach that even approximates to CBT,” the authors wrote.1
The finding is not surprising, since treatment manuals do not improve outcomes (read my blog about this), and therapists in the real world naturally adapt their approaches to the needs of individual patients. Their practice methods also evolve over time as they learn through hard-won experience what is helpful to patients and what isn’t.
In fact, studies show that when CBT is effective, it is at least in part because the more skilled practitioners depart from the manuals and use methods that are fundamentally psychodynamic. These include open-ended, unstructured sessions (versus following an agenda from a manual), working with defenses, focusing on the therapy relationship as a window into problematic relationship patterns, and drawing connections between the therapy relationship and other relationships.
So the research finding was no surprise. Something would be seriously amiss if experienced clinicians practiced like beginners, following an instruction manual like a consumer trying to assemble an appliance. What caught my eye was the authors’ conclusion that clinicians should be trained to adhere to CBT interventions “to give patients the best chance of recovery.”
The study did not address therapy outcome, so the authors actually had no idea which therapists were effective or which patients got better. They just declared, with no evidence whatever, that departure from treatment manuals means worse therapy. And this presumption— which flies in the face of actual scientific evidence— slipped right past the “evidence-oriented” reviewers and editors of a prestigious research journal. They probably never gave it a second thought.
The Big Lie
Academic researchers have usurped and appropriated the term “evidence based” to refer to a group of therapies conducted according to instruction manuals (“manualized” therapies). The other things these therapies have in common are that they are typically brief, highly scripted, and almost exclusively identified with CBT. The term “evidence-based therapy” is also, de facto, a code word for “not psychodynamic.”
It seems not to matter that scientific research shows that psychodynamic therapy is at least as effective (see my original research article, The Efficacy of Psychodynamic Psychotherapy or for a popularized version, see Getting to Know Me: What's behind psychoanalysis). Advocates of “evidence-based therapy” often denigrate psychodynamic treatment (or rather, their mistaken stereotypes and caricatures of it). When they use the term “evidence based,” it is often with an implicit wink and a nod and the unspoken message: “Manualized treatment is Science. Psychodynamic treatment is superstition.”
Some explanation is in order, since this is not how things are usually portrayed in textbooks or university classrooms. In past decades, most psychotherapists practiced psychodynamic therapy or were strongly influenced by psychodynamic thought. Psychodynamic therapies aim at enhancing self-knowledge in the context of a deeply personal relationship between therapist and patient.
Psychodynamic or psychoanalytic clinicians in the old days were not especially supportive of empirical research. Many believed therapy required a level of privacy that precluded independent observation. Many also believed that research could not measure crucial treatment benefits like self-awareness, freedom from inner constraints, or more intimate relationships. In contrast, academic researchers routinely conducted controlled research trials comparing manualized CBT to control groups. These manualized forms of CBT were therefore termed “empirically validated” (the preferred term later morphed into “empirically supported” and more recently, “evidence based”).
No research findings ever suggested that manualized CBT was more effective than psychodynamic therapy. It was just more often studied in research settings. There is a world of difference between saying that a treatment has not been extensively researched and saying it has been scientifically discredited. But academic researchers routinely blurred this distinction. A culture developed in academic psychology that promoted a myth that research had proven manualized CBT superior. Some academic researchers—those with little regard for actual scientific evidence—even began saying it was "unethical" to practice psychodynamic therapy because research showed CBT was more effective. The only problem is that research showed nothing of the sort.
This may shed some light on why the authors of the study I described above could so glibly assert that therapists should adhere to CBT treatment manuals "to give patients the best chance of recovery"—and how this falsehood could sail right through the editorial review process of a reputable research journal.
Stay tuned for future posts (here), where I will discuss whether patients who receive “evidence-based therapies” actually get better. The answer may surprise you.
Jonathan Shedler, PhD practices psychology in Denver, CO and online by videoconference. He is a Clinical Associate Professor at the University of Colorado School of Medicine. Dr. Shedler lectures to professional audiences nationally and internationally and provides online clinical consultation and supervision to mental health professionals worldwide.
Visit and "like" my Facebook page to hear about new posts or discuss this one. If you know others interested in this topic, please forward the link (use the email button on this page). See my other blog posts here.
© 2013 by Jonathan Shedler, PhD
1Waller, G., Stringer, H. Meyer, C. (2012). What cognitive behavioral techniques do therapists report using when delivering cognitive behavioral therapy for the eating disorders? Journal of Consulting and Clinical Psychology, 80, 171-175. doi: 10.1037/a0026559