The New Savoy Partnership Website

Evidence based practice - The state of the art in treating common mental illness

  print button close window

Dr Matthew Patrick,Trust Director, Tavistock and Portman NHS Foundation Trust
Professor Robert Elliott, Professor of Counselling, University of Strathclyde and Emeritus Professor of Psychology, University of Toledo
Professor John Markowitz, Research Psychiatrist, New York State Psychiatric Institute, USA and Clinical Professor of Psychiatry, Weill Medical College of Cornell University, New York
Professor Lars-Goran Ost, Professor of Clinical Psychology, Stockholm University, Sweden
Dr Frank Margison, Medical Director and Consultant Psychiatrist, Manchester Mental Health and Social Care Trust

Biography: Prof. Robert Elliott

Robert Elliott, Ph.D., is Professor of Counselling in the Counselling Unit at the University of Strathclyde, and Professor Emeritus of Psychology at the University of Toledo (Ohio). He received his Ph.D. in Clinical Psychology from UCLA in 1978. He served as President of the Society for Psychotherapy Research (2000-2001) and as co-editor of the journal Psychotherapy Research (1994-1998). He is co-author of Facilitating emotional change (1993, with Leslie Greenberg and Laura Rice), Learning process-experiential psychotherapy (2004, with Jeanne Watson, Rhonda Goldman, and Leslie Greenberg), and Research methods in clinical psychology (2002, 2nd ed., with Chris Barker & Nancy Pistrang), as well as more than 90 journal articles or book chapters. He is a Fellow in the Divisions of Psychotherapy and Humanistic Psychology of the American Psychological Association and is the 2008 recipient of the Carl Rogers Award from the Division of Humanistic Psychology of the American Psychological Association. He is Editor Emeritus of Person-Centered Counseling and Psychotherapies.


Title: Empirical Support for Person-Centred/Experiential Psychotherapies: Meta-analysis Update

Moderator: Robert Elliott (University of Strathclyde)

Co-authors: Beth Freire & Mick Cooper (University of Strathclyde)

Background: For reasons that are not clear, the large empirical literature that supports the Person-Centred/Experiential (PCE) practice is generally not known or reflected in mental health policy, a problematic situation that extends to PCE therapists and counsellors themselves.

Aims: Building on previous meta-analytic studies (e.g., Elliott, Greenberg & Lietaer, 2004), we added another 30 predominantly recent outcomes studies to the large sample previously reported, for a total of roughly 140 quantitative outcome studies on person-centred, process-experiential/emotion-focused, gestalt and related experiential therapies. In this paper, we summarize these studies in terms of the three lines of evidence on the effectiveness of these therapies across a broad range of client presenting problems.

Results: The following results hold across both previous and current replication samples of quantitative therapy outcome research:

  1. Clients in PCE therapies experienced large amounts of pre-post change.
  2. Posttherapy gains were maintained over early and late follow-ups.
  3. In controlled studies, clients experienced large gains relative to untreated groups.
  4. In general, PCE therapies appeared to be statistically and clinically equivalent when compared to non-PCE therapies including CBT.
  5. In the old sample, CBTs did better than person-centred or nondirective therapies, but the difference is small, and may be due to researcher allegiance.
  6. The strongest support for PCE therapies is for couples problems, depression and PTSD/trauma, where they meet standard criteria for Evidence Based Practice.
  7. There is suggestive evidence of effectiveness for severe disorders (schizophrenia, borderline process) and psychosomatic problems.

Implications: These results are consistent with complementary lines of evidence relating empathy to outcome (Bohart et al., 2002), and client treatment preference data. Taken together, the body of evidence clearly indicates that PCE therapies should offered to clients in primary care, NHS, and other mental health settings. Relying on multiple lines of evidence, such as provided in the present study, provides a sound basis for establishing public mental health policy.

Keywords: Outcome research, Person-Centred/Experiential therapies, mental health policy

Biography: Prof. John C. Markowitz

John Markowitz, M.D. is a Clinical Professor of Psychiatry at Weill Medical College of Cornell University, Research Psychiatrist at the New York State Psychiatric Institute, and Adjunct Clinical Professor of Psychiatry at Columbia University. He was graduated from the Columbia University College of Physicians & Surgeons and did his psychiatric residency at the Payne Whitney Clinic of New York Hospital. He began his research career at Cornell under the tutelage of Drs. James Kocsis, the late Samuel Perry, and the late Gerald L.Klerman. In 2001 he shifted from full time employment at Cornell to the New York State Psychiatric Institute. An NIMH Faculty Scholar Award allowed him to train in cognitive behavioral therapy at the Beck Institute in Philadelphia while learning interpersonal psychotherapy (IPT) from Dr. Klerman at Cornell. Dr.Markowitz's research has included funded comparative studies of psychotherapy and pharmacotherapy trials for depressed patients with HIV infection, and dysthymic and other chronically depressed patients. He is also conducting trials of IPT for patients with borderline personality disorder and posttraumatic stress disorder. He has written or co-authored fourteen books and more than 200 peer-reviewed articles and chapters.

Download PowerPoint presentation (PDF file)


1. Interpersonal psychotherapy (IPT) is one of the most tested empirically based psychotherapies. A time-limited, diagnosis-targeted, manualized treatment, IPT defines psychiatric diagnoses as treatable illnesses that are not the patient’s fault. It focuses on the relationship between recent life events and affects. Repeated randomized controlled trials have demonstrated its efficacy in treating patients with major depressive disorder and with bulimia, hence IPT has been included in treatment guidelines for those disorders. More preliminary evidence suggests its possible benefit for other psychiatric diagnoses. IPT has not shown efficacy in four prior trials for substance use disorders.

2. I am currently involved in studies testing adaptations of IPT for patients with posttraumatic stress disorder (PTSD), borderline personality disorder, and major depression comorbid with breast cancer. Other recent studies have evaluated the monotherapy and combined IPT and nefazodone for major depression (in the Netherlands), IPT for depressed inpatients (in Germany), an adaptation of IPT as an adjunct to pharmacotherapy for bipolar disorder (in the US), and IPT for depressed medically ill patients (in Canada).

3. Although studies of differential therapeutics are few, IPT may have advantages over cognitive behavioral therapy for patients with greater depressive severity (cf., the NIMH TDCRP study) and patients whose present illness contains meaningful life events or interpersonal stressors. Conversely, CBT may be preferable to IPT for patients reporting no life events, who fall into the IPT category of “interpersonal deficits.”

4. Important future research may include: 1) extending IPT to other disorders (e.g., as a non-exposure-based approach to treating PTSD), 2) more comparative studies with other psychotherapies to determine differential therapeutics, and 3) sequencing studies of when and how to combine IPT with pharmacotherapy.

Biography: Prof. Lars-Göran Öst

Lars-Göran Öst, Ph.D. is professor of clinical psychology at Stockholm University, Sweden, and also has a part-time position at the University of Bergen, Norway. He has been doing research on the psychological treatment of anxiety disorders for the last 30 years, and published randomized controlled trials on cognitive behavior therapy for adults with specific phobias, social phobia, agoraphobia, panic disorder, generalized anxiety disorder and post-traumatic stress disorder. Recently, he has also done RCTs in children and adolescents with specific phobia and social phobia. He has published about 200 articles, book chapters and books. Besides his academic job he also has a small private practice.


Cognitive Behaviour Therapy in anxiety disorders.

Professor Lars-Göran Öst, Department of psychology, Stockholm University, Sweden

Behavioural and cognitive-behavioural treatments for various anxiety disorders have been evaluated in randomized clinical trials (RCTs) for the past 40 years, and there are more than 400 RCTs published during this time. Various authors have described which psychological treatments are empirically supported for the different anxiety disorders.

  • Summarizing these reports I found the following treatments to be empirically supported:
  • Exposure in-vivo for specific phobias, social phobia, panic disorder with agoraphobia.
  • Exposure in imagination for post-traumatic stress disorder.
  • Exposure and response prevention for obsessive compulsive disorder.
  • CBT (cognitive therapy plus exposure) for social phobia, panic disorder with or without agoraphobia, generalised anxiety disorder, PTSD.
  • Applied relaxation for panic disorder without agoraphobia, generalised anxiety disorder.
  • Applied tension for blood-injury phobia.
  • Panic control treatment for panic disorder without agoraphobia.

The meta-analyses published showed average controlled effect sizes (ES) from 0.69 (panic disorder without agoraphobia) to 1.53 (PTSD), and uncontrolled ESs from 0.72 (social phobia) to 2.10 (GAD). However, a problem with meta-analyses is that they lag behind and sometimes new and better treatments haven’t been included in these analyses. To check this possibility I selected the best treatment within each anxiety disorder and now the controlled ESs varied from 1.56 (social phobia) to 2.40 (panic disorder without agoraphobia), and the uncontrolled ESs from 2.26 (social phobia) to 2.86 (specific phobias). On average, the best treatments yielded controlled and uncontrolled ESs that were twice as high as the means from the meta-analyses.

Thus, the efficacy studies show that a number of CBT-treatments are empirically supported and give high ESs, but then another question becomes important: Does CBT work in clinical settings? To answer this question data base searches of effectiveness studies were done up to 2007. The inclusion criteria were that the study must be clinically representative, the patients must fulfill an anxiety diagnosis, and the treatment must be one of the empirically supported CBT-treatments. A total of 67 studies were found and for these the uncontrolled ES was calculated. These were then statistically compared with the mean uncontrolled ES from the most current meta-analysis of efficacy studies for each anxiety diagnosis. The effectiveness studies had mean ESs from 1.05 (GAD) to 2.05 (PTSD) but in no case was the difference between efficacy and effectiveness studies significant, neither at post-treatment, nor at follow-up. This means that CBT works as well in clinical settings as in university departments.

Some of the most important questions for future research are: To improve current treatments and develop new treatments for OCD, GAD, and PTSD; To develop briefer and less expensive treatments without losing clinical efficacy; To match patient characteristics and treatment method; To evaluate early treatment in primary care; How to best disseminate CBT to clinicians; How to best combine different treatments; How to switch to another treatment when the first one has failed.


  • Behaviour Research and Therapy (Elsevier)
  • Behavioural and Cognitive Psychotherapy (Cambridge University Press)
  • Behavior Therapy (Elsevier)Cognitive Behaviour Therapy (Taylor & Francis)
  • Associations:
  • British Association for Behavioural & Cognitive Psychotherapies (
  • European Association for Behavioural & Cognitive Therapies (
  • Association for Behavioral and Cognitive Therapies (