CBT versus psychodynamic psychotherapy for social anxiety disorder: long live the dodo or time for a change in research emphasis?

shutterstock_72639100

The battle for supremacy between various psychotherapeutic approaches to the support of mental distress has led to many bloody skirmishes but no modality appears to emerge as a clear victor – the Dodo Hypothesis, ‘everyone has won and all must have prizes’, continues to haunt the evidence base which is bedevilled by author allegiance bias (Munder et al 2013) and the limitations of short term trial follow up (Clarkin 2014).

So another trial report steps forward to offer its authors’ interpretation of the varying merits of cognitive behavioural therapy [CBT] and psychodynamic therapy, in this case with regard to the support of individuals diagnosed with social anxiety disorder (Leichsenring et al, 2014).

Methods

This study represents a two year follow-up on the outcomes from a previously reported clinical trial comparing CBT and psychodynamic therapy. Immediate findings from the report had shown a trend towards superiority for CBT that did not reach the level of clinical significance defined by the authors at the outset of the trial.

Aim

In this report, the authors aim to comment on:

…the long-term effects of both CBT and psychodynamic therapy.

Interventions

Participants received either:

  1. CBT based on a model of social anxiety as being the result of internalised negative self-image that is challenged through role playing and other feedback mechanisms to ‘correct’ the aberrant image
  2. Psychodynamic therapy where expectations of personal humiliation through interpersonal relationships are challenged with a view to current relationships and the therapist-client relationship

Both therapies consisted of up to twenty-five 50-minute appointments and there was no difference in the number of sessions provided between treatment arms.

Participants received either CBT or Psychodynamic therapy focussing on social anxiety experiences

Participants received either CBT or Psychodynamic therapy focussing on social anxiety experiences.

Study participants

  • 495 participants were randomised to one of three treatment arms
    • CBT (n=209)
    • Waiting list (n=79)
    • Psychodynamic therapy (n=207)
  • Those completing the study were followed up over 24 months with repeated assessments at 6, 12 and 24 months
    • Completing CBT (n=159)
    • Completing psychodynamic therapy (n=149)
  • A total of 108 participants were available for follow-up assessment at 24 months
    • CBT = 69
    • Psychodynamic therapy = 39
  • Participants lost to follow up at any point in the trial were included through a multiple imputation method; a method that produces more variation in measure than the more traditional last outcome carried forward approach but that is generally accepted as being more accurate
  • Each of the research sites from which participants were recruited were represented by lead therapists from both disciplines in order to minimise potential allegiance bias
  • A central, independent, agency monitored the trial and supervised randomisation

Measures

The primary outcome measures were the dichotomous statements of remission and response defined respectively as:

  1. Reduction in symptom severity to less than a pre-defined threshold value
  2. Symptom severity reduction of 31% (equivalent to a change of ‘much-improved’ on the clinical global impressions scale).

Secondary outcomes included dimensional changes on scales assessing social anxiety, depression severity and personal reporting of interpersonal problems.

No therapy condition achieved superiority over the other throughout the extended two year follow up period

No therapy condition achieved superiority over the other throughout the extended two year follow-up period.

Results

The primary assessment measure of the study related to the rates of remission and response among participants:

CBT Psychodynamic therapy
Remission
Post-therapy 38% [30-45] 28% [21-35]
6 months 44% [34-53] 37% [27-46]
12 months 44% [32-55] 37% [25-48]
24 month 39% [25-42] 38% [22-54]
Response
 Post-therapy 63% [55-71] 58% [49-67]
 6 months 72% [62-81] 65% [55-75]
12 months 70% [58-82] 64% [52-76]
24 months 69% [53-84] 69% [54-84]

Note: Above figures in square brackets are [95% confidence intervals]

Similarly no differences between treatment arms were observed in any of the secondary outcome measures.

Conclusions

The authors concluded:

CBT and psychodynamic therapy were efficacious in treating social anxiety disorder, in both the short- and long-term, when patients showed continuous improvement. Although in the short term, intention-to-treat analyses yielded some statistically significant but small differences in favour of CBT in several outcome measures, no differences in outcome were found in the long-term.

This was a well designed trial that made efforts to minimise researcher allegiance bias at various stages

This was a well designed trial that made efforts to minimise researcher allegiance bias at various stages.

Discussion

In this report the authors demonstrate a statistically, but they argue not clinically, significant outcome in favour of CBT in terms of remission rates immediately post-treatment; any difference between treatment arms is however lost by 24 months follow-up and both treatment arms outperformed the waiting list control; although there are limitations to the use of such control measures.

This appears to be a rigorously conducted trial with efforts to minimise researcher allegiance bias at various steps in the methodology and the results show that the Dodo continues to happily walk along the edge of the cliff, master of all it surveys.

The study is of course not without limitations and is susceptible to the same criticisms levelled at most studies of this nature:

  • While the CBT provided is deemed appropriate, the absence of efforts in the psychodynamic treatment model to identify the origins of the described interpersonal distress can be criticised
  • There is a significant loss of participants to follow up; the multiple imputation statistical model is likely a good correction for this, but ultimately it can never fully account for the loss of participants
  • Outcome measures based on clinical symptom rating can be criticised for their lack of reflection on changes that truly matter to clients
  • The lack of in-depth interpersonal assessment through individual measures means we cannot comment directly on individual preferences regarding therapies

As usual we require more research; although perhaps now is the time to switch the emphasis of research away from partisan declarations of potential therapeutic superiority for varying modalities to recognising that we need to work with individual clients to identify outcomes and therapies that are best suited to their particular needs and circumstances.

Trials continuously fail to demonstrate any difference  between therapeutic approaches. Do we need a change of research emphasis?

Trials continuously fail to demonstrate any difference between therapeutic approaches. Do we need a change of research emphasis?

Links

Leichsenring, F., Salzer, S., Beutel, M. E., Herpertz, S., Hiller, W., Hoyer, J., et al. (2014). Long-term outcome of psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder. The American Journal of Psychiatry, 171(10), 1074–1082. [PubMed]

Shepherd A. Is the Dodo finally dead? The Mental Elf, 14 Apr 2014.

Munder, T., Brütsch, O., Leonhart, R., Gerger, H., & Barth, J. (2013). Researcher allegiance in psychotherapy outcome research: An overview of reviews, 33(4), 501–511. [PubMed]

Clarkin, J. (2014). Raising the bar in the empirical investigation of psychotherapy. The American Journal of Psychiatry, 171(10), 1027–1030. [PubMed]

Leichsenring, F., Salzer, S., Beutel, M. E., Herpertz, S., Hiller, W., Hoyer, J., et al. (2013). Psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder: a multicenter randomized controlled trial. The American Journal of Psychiatry, 170(7), 759–767. [PubMed]

Furukawa, T. A., Noma, H., Caldwell, D. M., Honyashiki, M., Shinohara, K., Imai, H., et al. (2014). Waiting list may be a nocebo condition in psychotherapy trials: a contribution from network meta-analysis. Acta Psychiatrica Scandinavica, 130(3), 181–192. [PubMed]

Andresen, R., Caputi, P., & Oades, L. G. (2010). Do clinical outcome measures assess consumer-defined recovery? Psychiatry Research, 177(3), 309–317. [PubMed]

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+