study of Computer-assisted Cognitive Behaviour Therapy (cCBT) for depression by University of York Researchers and published in the BMJ has recently hit the headlines as it has suggests that cCBT may offer little or no benefit for depression.

MindTech have had the following response to this study published - please see the BMJ website for the full letter and other rapid responses.

Re: Computerised cognitive behaviour therapy (cCBT) as treatment for depression in primary care (REEACT trial): large scale pragmatic randomised controlled trial

We commend the authors of the REEACT trial for conducting a large-scale pragmatic randomised controlled evaluation of the effectiveness of offering either computerised cognitive behaviour therapy (cCBT) (Beating the Blues or MoodGYM) in addition to usual care for depression in UK primary care settings.

Computerised CBT packages available on the internet, especially ones that are free to the consumer and the NHS, widen access to treatment in a variety of important ways, such as choice, immediate access, contextualisation to personal need, not having to ask other people for help and lack of stigma from having a consultation for depression recorded in case notes. These considerations regarding scale of delivery, efficiency and accessibility are particularly important for people with depression, for whom it is estimated that only 25 per cent receive any health service treatment at all [1].

The authors correctly state that as a pragmatic trial they have not set out to test the efficacy of cCBT for depression already established in meta-analyses [2]. However, in our view, their headline conclusion that ‘supported (our italics) computerised cognitive behaviour therapy confers modest or no benefit over usual GP care’ overstates what the REEACT study can tell us regarding the effectiveness of cCBT more generally, as the cCBT interventions in this trial were unguided and unsupported by clinicians. Research has consistently shown that unguided cCBT delivered without clinician support has smaller effects and poorer adherence than clinician guided cCBT [3, 4].

Hence, the results of the REEACT trial are hardly surprising given what we already know. However, our main concern is that that the authors’ conclusions regarding the lack of effectiveness of ‘supported’ cCBT for depression in primary care could easily be misinterpreted by policy makers and commissioners of services as undermining the benefits of cCBT and digital interventions more generally, in particular where clinician support is provided and effectiveness has been clearly demonstrated and equivalence shown to face to face CBT [5].

In our view, the most important result of this trial, and key reason for lack of effectiveness of the interventions, is the very low acceptability and uptake of cCBT when delivered without clinical or therapist support for patients with depression in primary care. The self-guided stand-alone computerised interventions, delivered without clinician support/ guidance, were unpopular and infrequently used by patients. Less than 20% of patients allocated to either Beating the Blues or MoodGym completed the full package and the median number of sessions utilised was just one in both groups. It is well known that clinician supported cCBT produces larger effects than unsupported self-guided cCBT, and the author’s acknowledged that the trial was probably underpowered to detect the small effect sizes (d=0.2 to 0.25) reported in trials of unsupported cCBT [6]. The larger effect sizes for ’supported’ cCBT interventions all include some level of remote therapist feedback, motivational support and interaction. In the REEACT trial, only ‘technical’ support was offered which amounted to an average of only 6 minutes in both intervention arms. Therefore, we think it is potentially misleading to characterise the REEACT trial as demonstrating the effectiveness of ’supported’ cCBT. The risk of labelling these interventions as ‘supported’ cCBT is that clinician supported cCBT may be viewed by policy makers and commissioners as equally ineffective. In addition, given the poor adherence to cCBT in this trial, it would have been helpful if the authors had presented a ‘dose-response’ analysis to test whether the adherence to the cCBT intervention predicted a better clinical response.

As well as to the likelihood that the REEACT trial was underpowered to detect the smaller effect size expected with self-guided cCBT delivered without clinician support, other design factors such as cCBT contamination effects between usual care and treatment arms, high levels of antidepressant use and the choice of a categorical recovery threshold may have masked or minimised a genuine, albeit small, intervention effect. Given that the PHQ-9 score at baseline was in the moderate to severe range, it is possible that self-guided cCBT could lead to small but worthwhile improvement in depressive symptoms although patients may still remain above the PHQ-9 threshold of 10 for ‘caseness’.

In hindsight, an internal pilot with the first 20 to 30 participants might have been carried out to find out the minimum therapeutic support people would be willing to receive before embarking on a RCT testing interventions that were unacceptable and unused in 691 people. The authors claim that they offered more support than for cCBT provided by IAPT but don’t provide any evidence to back up this statement. Self-guided treatments for depression without clinician support are likely to appeal only to a minority who are very motivated for this kind of treatment and are capable of finishing the intervention without stimulation from an external therapist or coach.

Finally, for policy makers and commissioners this study offers a useful lesson concerning what we should not do when implementing this type of intervention. It also illustrates the importance of considering the type of human support available and setting in which these interventions are delivered. For patients with depression in primary care, cCBT when delivered as part of stepped-care is most likely to be accepted and take-up when there is at least a minimal level of clinical guidance and support. Accurately specifying this level of support is particularly important when comparing the effectiveness of different interventions. Increasingly, cCBT and other digital interventions are being offered in IAPT and other mental healthcare settings as part of ‘blended’ care delivery which integrates clinician contact and guidance with automated computerised self-help. Because self-guided (unsupported) cCBT can be delivered a near-zero marginal cost, despite small effects and adherence, it may be an effective public health intervention where delivered at scale via sites such as NHS Choices ‘Moodzone’ [7] to those who do not access traditional health services such as primary care. Currently, research supported by NIHR is underway to guide commissioning decisions regarding the clinical- and cost-effectiveness of offering internet-based peer support and self-guided (unsupported) cCBT at a population level outside primary care [8].

Chris Hollis, Professor of Child and Adolescent Psychiatry, University of Nottignham and Clinical Director NIHR MindTech Healthcare Technology Co-operative
Richard Morriss, Professor of Psychiatry, University of Nottignham and Mood Disorder Theme Lead NIHR MindTech Healthcare Technology Co-operative
Gerhard Anderssen, Professor of Clinical Psychology, Linköping University and Karolinska Institute
Elizabeth Murray, Professor of eHealth and Primary Care, University College London

References:
[1] Davies, S.C. “Annual Report of the Chief Medical Officer 2013, Public Mental Health Priorities:
Investing in the Evidence” London: Department of Health (2014)
[2] Richards D, Richardson T. Computer-based psychological treatments for depression: a systematic review and meta-analysis. Clin Psychol Rev. 2012;32:329-42.
[3] Christensen H, Griffiths K, Groves C, Korten A. Free range users and one hit wonders: community users of an Internet-based cognitive behaviour therapy program. Aust N Zealand J Psychiatry. 2006;40:59-62.
[4] Johansson R, Andersson G. Internet-based psychological treatments for depression. Expert Rev Neurother. 2012;12:861-70.
[5] Andersson G, Topooco N, Havik OE, Nordgreen T. Internet-supported versus face-to-face cognitive behavior therapy for depression. Expert Review of Neurotherapeutics. In press.
[6] Cuijpers P, Donker T, Johansson R, Mohr DC, van Straten A, et al.) Self-Guided Psychological Treatment for Depressive Symptoms: A Meta-Analysis. PLoS ONE 2011; 6(6): e21274. doi:10.1371/journal.pone.0021274
[7] http://www.nhs.uk/conditions/stress-anxiety-depression/pages/low-mood-st...
[8] Morriss R, Hollis C, Coulson N, Noran P, Avery A, Tata L et al. Randomised controlled trial of an established direct to public peer support and e-therapy programme (Big White Wall) versus information to aid self-management of depression and anxiety. Protocol. National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care East Midlands.29th September 2015.

 

Join MindTech (click)

We would love to keep in touch with you regarding our work and how you might be able to help. Please fill in this form so we can contact you.

Have you had your say yet? Tell us your top 3 research questions about digital for mental health… https://t.co/5IVZ929SuE

What if tech tried to be healing instead of just addictive? Interesting article about designing wellbeing into tech https://t.co/yOlYVF6f2v