No evidence that behavioural therapies are any better than other psychological therapies for depression

Depression is a big problem.  In fact, it’s the third leading cause of disease burden worldwide (WHO, 2004 – as cited in Shinohara et al, 2013) and the largest source of nonfatal disease burden in the world (Ustun, 2004 – as cited in Shinohara et al, 2013).   What’s more, the number of people affected by it is predicted to increase over the next two decades (WHO, 2008 – as cited in Shinohara et al, 2013).

NICE recommends both pharmacological and psychological interventions for depression, either independently or in combination.  Surveys tell us that there is a patient preference for psychological interventions (Churchill, 2000; Riedel-Heller, 2005 – as cited in Shinohara et al, 2013), which means it would be helpful to know exactly which elements of the psychological intervention are the most effective so that we can maximise this efficacy where possible.

People have attempted to answer this research question before: in 1996, Jacobson demonstrated that the behavioural elements of CBT were as effective alone as the whole CBT package.

NICE guidelines recognise the value of behavioural approaches and recommend techniques such as behavioural activation for moderate depression.  However, with so many varied psychological approaches available, clear evidence about how behavioural therapies stack up against other psychological therapies in terms of both efficacy and acceptability would be welcome  – especially as behavioural therapies may be more cost-effective and simpler to deliver than other therapies. (Kanter, 2010 – as cited in Shinohara et al, 2013).

Luckily for us, Shinohara et al (2013) have put together a Cochrane review which sets about to do just that very thing.

Behavioural therapies concentrate more on what people are doing, including helping them to strengthen their social skills and assertiveness

Behavioural therapies concentrate more on what people are doing, including helping them to strengthen their social skills and assertiveness

Methods

To produce their review, Shinohara et al searched two clinical trial registers, several electronic databases and reference lists for relevant studies, as well as making personal communication with experts in the field.  They selected only RCTs with particular designs, which included men and women aged between 18 and 75 who were being treated for an acute phase of depression (i.e. not chronic, treatment-resistant or remitted depression).  Inpatient studies were excluded.  Therapy could be delivered individually or in group settings, but self-help, guided self-help and writing therapies were excluded.  25 studies were used in the final analyses.

BT approaches were defined as: behavioural therapy, behavioural activation, social skills training and relaxation training.  Psychological therapy approaches included: CBT, third-wave CBT, psychodynamic therapy, humanistic therapy and integrative therapy.

The following outcomes were assessed:

  • Primary outcomes:
    • Treatment efficacy: the number of participants who demonstrated a reduction of 50% or more on a validated depression scale.
    • Treatment acceptability: the number of participants who dropped out of therapy.
  • Secondary outcomes:
    • Number of remissions during treatment
    • Improvement in depression symptoms (using validated scales)
    • Improvement in overall symptoms using the Clinical Global Impressions scale (CGI)
    • Improvement in anxiety symptoms (using validated scales)
    • Adverse effects (suicide, attempted suicide, worsening of symptoms)
    • Social adjustment (using the Global Assessment of Function measure)
    • Quality of life (using validated scales)
    • Economic outcomes – such as service use, days off work etc

Results

Differences between BT and psychological therapies in the primary and secondary outcomes were analysed first by comparing all BT therapies against all psychological therapies, then by comparing different BT therapies one by one against all psychological therapies, and then, finally, all BT therapies against different psychological therapies one by one.

In all these analyses there was only one statistically significant result: a low-quality evidence finding that CBT was slightly better for treatment efficacy (reduction of 50% or more on a validated scale) than BT.

No relevant data was available to analyse overall improvement in symptoms, adverse effects, quality of life, economic outcomes, third wave CBT or anxiety and social adjustment in humanistic therapy or remission and social adjustment in integrative therapy.

All other analyses found no statistically significant difference between BT and psychological therapies.

Shinohara et al found a small number of studies which included follow up data (within 6 months).  Analysis of these studies found a statistically significant difference between CBT and BT for depression remission, treatment efficacy (reduction by at least 50% on a scale) and depression severity, in favour of CBT.

There was only one statistically significant result: a low-quality evidence finding that CBT was slightly better for treatment efficacy than BT.  All other analyses found no statistically significant difference between BT and psychological therapies.

Most of the analyses found no differences between CBT and behavioural therapies in the treatment of depression.

Conclusions

The overall impression this elf is left with is that Shinohara et al’s review was seriously hampered by lack of data, small sample sizes, and the low quality of the data available.

All of the studies used in the review were from Western countries and published between 1975 and 1996.  This makes them slightly out of date, and difficult to generalise to non-Western settings.  Shinohara et al also highlight a potential for bias: treatment fidelity.  With only six studies giving data about how closely trial therapists kept to the BT model, it is unclear whether the intervention was delivered as intended in each study.  Similarly, no studies reported on which therapy manuals were used in the study – something which Shinohara et al would like to see detailed in future studies to enable replication.

Another problem was that studies often did not clearly state what outcomes they would define as “remission” or “response” to treatment, meaning that Shinohara et al had to impute outcomes themselves.  Nor did any study report details on how allocation was kept from the assessors, and many did not describe how allocation was randomised.

These caveats aside, what can be concluded from Shinohara et al’s review is that there is no strong evidence that BT is any different in effectiveness or acceptability in the treatment of depression than any other psychological therapy.  This is an interesting conclusion because it paves the way for future research to look at BT further – specifically, which aspects of it are the most helpful to patients, and whether it might be a more cost-effective approach in a squeezed NHS because it can be delivered more simply than other psychological therapies.

This review found no strong evidence that behavioural therapies are any different in effectiveness or acceptability in the treatment of depression than any other psychological therapy

This review found no strong evidence that behavioural therapies are any different in effectiveness or acceptability in the treatment of depression than any other psychological therapy

Link

Shinohara K, Honyashiki M, Imai H, Hunot V, Caldwell DM, Davies P, Moore THM, Furukawa TA, Churchill R. Behavioural therapies versus other psychological therapies for depression. Cochrane Database of Systematic Reviews 2013, Issue 10. Art. No.: CD008696. DOI: 10.1002/14651858.CD008696.pub2.

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