From the Editor

Is CBT overrated? The authors of a new JAMA paper raise this question in a cutting Viewpoint.

In this two-part Reading of the Week series, we look at two papers, both published in JAMA. These Viewpoint pieces make interesting, provocative arguments.

Last week, we looked at conversational agents.

This week, we ask: is CBT really the gold standard for psychotherapy?

University of Giessen’s Falk Leichsenring and Medical School Berlin’s Christiane Steinert consider CBT and the research that has been done in the area. “CBT is usually considered the gold standard for the psychotherapeutic treatment of many or even most mental disorders.” But should it be? Leichsenring and Steinert argue no.

beck_aaron_t-_112798Aaron Beck: Great bowtie, but is his CBT really so great?

In this Reading, we review their paper, and consider their argument.


Therapy and CBT

“Is Cognitive Behavioral Therapy the Gold Standard for Psychotherapy? The Need for Plurality in Treatment and Research

Falk Leichsenring and Christiane Steinert

JAMA, 21 September 2017 Online First


Mental disorders are common and associated with severe impairments and high societal costs, thus representing a significant public health concern. About 75% of patients prefer psychotherapy over medication. For psychotherapy of mental disorders, several approaches are available such as cognitive behavioral therapy (CBT), interpersonal therapy, or psychodynamic therapy. Pointing to the available evidence, CBT is usually considered the gold standard for the psychotherapeutic treatment of many or even most mental disorders. For example, the American Psychological Association’s Division 12 Task Force on Psychological Interventions currently lists CBT as the only treatment with “strong research support” in almost 80% of all mental disorders included in its listing.

For a treatment to be considered the gold standard requires that substantial supporting evidence exists. Recently, however, additional research findings have emerged that question the prominent status of CBT. In this Viewpoint, we review some of the most important findings.

11153Falk Leichsenring

So begins a paper by Leichsenring and Steinert. Note that this paper is open access.

They make six points about CBT:

Limited Study Quality

The authors argue that the quality of the evidence is as important as the quantity. Yet “a recent meta-analysis using criteria of the Cochrane risk of bias tool reported that only 17% (24 of 144) of randomized clinical trials (RCTs) of CBT for anxiety and depressive disorders were of high quality.”

Weak Empirical Tests

The authors argue that, to examine efficacy, a treatment should be compared to more than just a waiting list. Yet 80% of CBT studies for anxiety did just this; the number is 44% for depression. “Being more effective than waiting list controls is not a strong proof of efficacy and may lead to overestimating the efficacy of CBT especially because waiting list controls may even represent a nocebo condition.”

Uncontrolled Researcher Allegiance

The authors argue that “researcher allegiance” – that is, a researcher’s own belief in the superiority of the treatment – can bias papers. For several studies, CBT researchers had “high risk for researcher allegiance.”

“In essence, the treatment conditions against which CBT was compared were designed in a way that essential curative factors were excluded.”

Central Mechanisms of Change Not Corroborated

“Cognitive therapy assumes that improvements in symptoms are achieved through changes in key cognitive processes (eg, negative triad, ie, a negative view of self, others, and the future). In a review based on the available evidence, a prominent CBT researcher concluded that this central assumption of CBT is not correct.”

Limited Efficacy: CBT Is Not a Panacea

“Several meta-analyses reported limited efficacy of CBT.”

“In the few high-quality studies available for depressive and anxiety disorders, CBT was found to be less efficacious than in low-quality studies, mostly reducing the efficacy of CBT in panic disorder and social anxiety disorder. In the high-quality studies, CBT achieved large effect sizes only in comparison with waiting list conditions. Compared with treatment as usual, effect sizes were only small to moderate (0.30-0.45). Thus, the additional gain of CBT over treatment as usual is limited and may eventually even be the result of allegiance effects.” They note problems with panic disorder studies, as well as research in other mental disorders.

No Clear Evidence of Superior Efficacy

“A first-line treatment usually is clearly more effective than other treatments. However, there is no clear evidence that CBT is more effective than other psychotherapies, either for depressive disorders or for anxiety disorders. This is also true for several other mental disorders (eg, personality disorders or specific eating disorders).”

They conclude:

After psychoanalysis had dominated the early years of psychotherapy claiming to be the gold standard, a “CBT-centric” era began and some of its proponents succeeded in presenting CBT not only as the empirically best studied treatment, but also as the most effective psychotherapy: “The most effective contemporary approach is Cognitive Behavior Therapy…” While CBT is beneficial for many patients, and CBT researchers developed and tested treatments often long before other approaches, the evidence suggests that CBT should not be considered the gold standard of psychotherapy.


No form of psychotherapy can presently claim to be the gold standard, suggesting the need for plurality in treatment and research, ie, a variety of different psychotherapeutic approaches. All evidence-based therapies have their strengths, be it a focus on cognitive, emotional, interpersonal, or unconscious processes. Only plurality allows for bridging the gap between the different approaches and for learning from each other to further improve the treatment of patients with mental disorders.

A few thoughts:

  1. This is a well-argued Viewpoint.
  1. The authors raise good points here.
  1. The authors warn of the dangers of replacing the era of psychoanalysis – marked by much self-congratulations and little scientific study – with another era of overenthusiasm. The thoughtfulness of this paper, the fact that it was published in a major journal, and the rich literature critical of CBT that the authors were able to tap, all suggest that psychiatry has learned from the past.
  1. CBT isn’t just found in studies, of course. There are real world experiences, too. None would be bigger than the UK’s IAPT – where CBT is offered free at the point of use.

The National Health Service collects and publishes data on the outcomes.

You can find the data (released monthly) here:

A graphical summary of the latest data:

Figure 2: Percentage of eligible referrals having recovered, reliably improved, and reliably recovered, July 2016 to June 2017, England 


Many benefit from CBT – more than 65% show “reliable improvement.”

  1. Is CBT the “gold standard?” The authors debate this question. To some extent, the answer comes down to interpretation of the term “gold standard.” Readers can draw their own conclusions, but this much is clear: CBT isn’t perfect and – while it helps many people with depression, anxiety, and other mental disorders – it certainly is no panacea.
  1. I’d like to thank Drs. Daniel Gorman and Paula Ravitz for suggesting this selection.
  1. As an aside, earlier this week, Richard Thaler won the Nobel Memorial Prize for Economic Sciences. His work on behavioural economics has implications for health care. Some have tried to use his “nudge” to get better health outcomes. In a past Reading, we looked at nudges to reduce smoking (a New England Journal of Medicine paper). You can find it here:


Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.