iCBT for depression: how does it work?


The elves have blogged about the effectiveness of Computerised Cognitive Behavioural Therapy (cCBT) sometimes known as Internet Cognitive Behavioural Therapy (iCBT) for depression on many previous occasions. Although there does seem some evidence that it reduces symptoms in children and adolescents, evidence in adults is at best weak. Reviewing the literature, one can clearly see how incredibly attractive the concept of using online access to CBT is to commissioners and patients in terms of cost savings and accessibility, such that it sometimes appears to me that we are desperate to prove it works because it is a cheap alternative to face-to-face, costly but effective therapy!

I should declare here that I am sceptical about cCBT/iCBT having been involved in its practical application in primary care settings in England through my role working for IAPT services; we stopped offering it following two years of disappointing outcomes and very poor uptake.

Sceptical or not, I was intrigued when our very own Elrond asked me to blog about a recent study (carried out by a Romanian group from Babeș-Bolyai University) entitled: “Internet-Based Psychotherapy for Adult Depression: What About the Mechanisms of Change?” – especially by the tagline.

The group defined Mechanisms of Change as:

…psychological factors assumed to be targeted by the intervention and to mediate the change in symptom.

The authors’ stress that by:

Knowing the mechanisms of change in iCBT can increase our understanding of the etiological factors underpinning depression and lead to the development of better (computerized) interventions.

and were also keen to emphasise that the mechanisms of change for iCBT are not necessarily the same as conventional (face-to-face) CBT.

The objective of this study was to:

..provide an overview of the state of the art regarding the mechanisms of iCBT for adult depression, in the context of iCBT efficacy.

This then is clearly a very different approach to the standard data crunching comparison of treatment versus control.

This new qualitative systematic review aimed to identify the mechanisms of change for Internet delivered CBT.

This new qualitative systematic review aimed to identify the mechanisms of change for Internet delivered CBT.


This is a systematic qualitative review. The authors’ searched the scientific literature (PsycINFO and MEDLINE databases) for studies on iCBT which met their standard criteria (RCTs, designed specifically to assess the efficacy of iCBT for depression in adults and/or the mechanisms underlying iCBT for depression, in English and published in a peer-reviewed journal). Quality of the included studies was assessed using the “risk of bias” assessment tool developed by the Cochrane Collaboration.

This current study applied the principles of a previous study carried out by David and Montgomery (2011) to assess the scientific status. According to this work, a psychological intervention can show the following types of clinical efficacy:

  • Absolute efficacy (i.e., the therapeutic package is significantly better than no treatment, waiting list control, and placebo condition);
  • Relative efficacy (i.e., the therapeutic package is equivalent to or better than another evidence-based psychological intervention); or
  • Specific efficacy (both the conditions for absolute and relative efficacy are met; the theory of the therapeutic package is also empirically supported)


The review identified 37 studies which fulfilled their inclusion/exclusion criteria (12 from a research group based in Australia; 6 from a group in Sweden, 6 from a Dutch group; with the remaining studies published by research teams from Germany and Switzerland (3 studies), United States (3 studies), United Kingdom (3 studies), Canada (2 studies), Ireland, Norway and Australia (1 study).

iCBT Efficacy

  • Overall, studies showed iCBT was equally efficacious with other active interventions including an iCBT component
  • 20 studies showed iCBT to be superior to waitlist (WL) control
  • 2 out of 6 studies reported iCBT to be better than placebo (regular phone calls to discuss lifestyle and environmental factors related to depression or less structured online activities related to decreasing depression such as online bibliotherapy, online discussion groups)
  • 3 out of 7 reported iCBT was as effective as treatment as usual (TAU), while the other four reported iCBT was somewhat superior to TAU
  • 13 studies compared iCBT to other active interventions, including an iCBT protocol
  • Some evidence was presented that suggested the MoodGym program was better than various combinations of its modules
  • 1 study reported that more severely depressed patients benefited most from the tailored iCBT protocol compared to the untailored one
  • 4 studies compared iCBT to face-to-face group CBT and reported iCBT was equally efficacious
  • Therapeutic gains yielded by iCBT were generally maintained at follow-up.

iCBT mechanisms of change

Evidence regarding the presumed mechanisms of change supporting the iCBT for depression proved to be scarce

  • 10 studies considered specific cognitive factors assumed to act as mechanisms of depression
  • 2 of the studies conducted mediation analyses, suggesting that positive beliefs about rumination, dysfunctional attitudes, worry, negative problem orientation and perceived control mediated the iCBT effect on depressive symptoms

  • 3 studies considered factors that could act as general mechanisms of change (e.g., therapeutic alliance and treatment credibility and expectancies for improvement)
  • 13 studies considered other individual- and treatment-related variables that could have influenced the outcome of the treatment (e.g., symptoms severity at baseline, educational level, employment status, satisfaction with treatment, therapist contact)
  • 6 studies conducted moderation analysis and reported that symptoms severity at baseline, marital status, educational level, and parental psychiatric history moderated the iCBT effect on depressive symptoms.

Risk of Bias

Only one study (of the 37 included) met all of three risk of bias criteria (adequate generation of the random sequence, allocation concealment, and adequate blinding of the assessors). 32 trials reported adequate randomisation, 12 reported allocation concealment, and only 3 reported adequate blinding of the assessors. The researchers summarised this state of affairs by saying: “The quality of iCBT randomized clinical trials proved to be suboptimal”.

This review shows us that iCBT mechanisms of change are clearly under-investigated.

This review shows us that iCBT mechanisms of change are clearly under-investigated.


The reviewers concluded that:

iCBT for depression seems to have only preliminary support (at most) in terms of mechanisms of change, while data supporting the therapeutic package can be described, in our view, as equivocal.

Strengths and limitations

This study set out to try to examine if iCBT is effective in the treatment of adult depression and to try to establish how it might work, in terms of mechanisms of change. From there study they claim that:

iCBT appears to yield significant clinical benefits compared with WL control.

However, the data that they have used in their study has many limitations, mostly identified by the authors’ themselves. Most obvious is both the lack of quantity and the lack quality of the studies included. These are too heterogeneous to distil any real conclusions, other than a need for more structured studies.

It is intriguing that they hypothesise that the mechanisms of change for iCBT could differ from standard face-to-face CBT and I for one remain sceptical that this will prove to be the case. Also from a purely personal view, I feel concerned that trying to treat depression, often characterised by withdrawal and isolation, using internet-based methods might serve to promote isolation and withdrawal.

The authors’ do state that the study is qualitative and not quantitative in design. However, I found it easy to become confused by their inclusion of “quantitative” data and the conclusions they make based on this. I would have preferred them to fully (meta)analyse the data to support their conclusions on efficacy.

Perhaps the most important contribution this study makes lies in the section headed: “Directions for Future Research: How Considering iCBT Mechanisms of Change Could Improve iCBT Protocols”. In this they make a clear case for their final conclusion:

The iCBT theory should be clearly specified and adequately investigated to design and implement highly efficacious therapeutic packages. Without considering the iCBT mechanisms of change along with iCBT efficacy, the extent to which iCBT is an empirically validated treatment remains questionable.

Do we need to understand how iCBT works or is it sufficient to simply know that it does work?

Do we need to understand how iCBT works or is it sufficient to simply know that it does work?


Primary paper

Mogoașe C, Cobeanu O, David O, Giosan C, Szentagotai A. (2016) Internet-Based Psychotherapy for Adult Depression: What About the Mechanisms of Change? J Clin Psychol. 2016 Sep 29. doi: 10.1002/jclp.22326. [PubMed abstract]

Other references

David, D., & Montgomery, G. H. (2011). The scientific status of psychotherapies: A new evaluative framework for evidence-based psychosocial interventions. Clinical Psychology: Science and Practice, 18(2), 89–99. http://doi.org/10.1111/j.1468-2850.2011.01239.x

Photo credits

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