Meta-review of the effectiveness of computerised CBT in treating depression

Depression and anxiety are amongst the most prevalent psychiatric conditions, affecting up to 8.8% of the population in the UK [1]. It is estimated that depression in adults could cost the nation over £9 billion a year and could lead to 109.7 million working days lost per year [2]. To address the problem, there is a strong evidence base for psychological therapy featuring cognitive behaviour therapy (CBT) [3]. However, conventional CBT is delivered face-to-face, meaning that the resources which would be needed to treat everyone who could potentially benefit for depression are vast.

The past decade has seen growing use of computers and the internet as an adjunct to face-to-face CBT, and in some cases as an alternative to individual contact with a clinician. A computer can be used in different ways; to provide information in written or audio/video formats; as a tool for screening or diagnosis; as a self-help tool, which entails following a structured programme on-line; to simulate situations and permit graded exposure in highly controlled conditions; and as a means of communication between patients and health care professionals for administrative or therapeutic purposes. The attractions of computer-mediated CBT (cCBT) include: its capacity to deliver structured input consistently with precision, low cost, easy accessibility and potential to offer a flexible approach in a non-stigmatising environment. Yet if best use of this resource is to be made, clear knowledge about the effectiveness and costs of alternative cCBT programmes is needed to guide planning and provision of care.

The main focus of this paper is on the delivery of computerised CBT (cCBT) for treating depression with or without anxiety. From a pragmatic perspective, users, clinicians and commissioners of health care are most likely to ask 'what works for people with depression?'. Since we were engaged in a study of MoodGYM, which is a publicly-available cCBT programme developed by the Centre for Mental Health Research at the Australian National University [4, 5], we wanted to place this study in the context of relevant literature concerning both MoodGYM and other similar programmes which are also primarily designed to treat depression. There have been different reviews on the topic, but their quality and comprehensiveness of their scope has not been studied yet. We therefore set out to review existing reviews, appraise these and draw together what they indicate regarding cCBT for depression, particularly if their results could help us to compare alternative cCBT packages for the same disorder.

The search strategy is given in the Additional file 1. The authors collaborated in the review and searched the following electronic data-bases: Medline, Embase, CINAHL (EBSCO), the British Nursing Index, the Cochrane Library, and PsycINFO. All systematic reviews examining efficacy of cCBT for depression with or without anxiety in adults were included at the outset, limited to the last ten years (initially searching 1999 to 2009, then updating to Feb 2011). The evolutionary development of cCBT and its reliance on ever-increasing speeds of internet communications make earlier reviews unlikely to be relevant to our aim. We also considered papers that were suggested to us via personal communications. We excluded books, chapters, commentaries and reviews whose methodology was not systematic or reviews whose methodology could not be ascertained, as well as systematic reviews that did not directly address effectiveness, e.g. those looking at patient acceptability or drop-out alone. To verify the methodology of articles, authors were consulted via email when necessary. The search was restricted to publications in English, since the cCBT packages of interest use the English language. The PRISMA statement guided the approach [6] and the Critical Appraisal Skills Programme (CASP) tool [7] was used to extract data from the papers identified.

The initial search identified 7031 studies, from which 6978 were excluded by screening the title and abstract. The remaining 52 papers were obtained and read; as a result 41 were excluded. Additional file 2 lists these and the reasons why. One paper identified through personal communication was included in the review. Twelve papers covering ten reviews of the effectiveness of cCBT for depression with or without anxiety in adults were included (Table 1 and additional file 3), and are summarised below (Figure 1).

Having critically appraised the quality of several reviews, we had to exclude 22 reviews that did not describe their methodology. Thirty-one other papers were excluded for reasons such as having a focus other than efficacy of cCBT on depression. Even the included reviews provided minimal information about individual packages, and their conclusions were based on only a small number of RCTs on each individual package (table 1). In effect we had to unpick even the better reviews to arrive at a conclusion about cCBT for depression.

By singling out the evidence pertaining to depression we have revealed reasonably consistent inferences drawn by from ten systematic reviews (twelve papers) of varying quality and relevance. Bearing in mind the (declared) interests of certain authors and the largely uncritical approach taken by some reviewers, we cautiously infer that certain cCBT packages, specifically MoodGYM, Beating the Blues and Colour Your Life, can have a positive effect on symptoms of depression. Yet there is limited information available to compare alternative packages, which is an important challenge given the differences in costs. The finding that cCBT reduces therapist time could have important implications for service organisation and costs, but the effect size of each intervention as well as the costs need to be compared before superior efficacy for cCBT can be claimed, which highlights the need for high-quality comparative studies as new packages are developed.

At present there is limited evidence which indicates the effectiveness of MoodGYM, Beating the Blues and Colour Your Life, but not enough of such evidence to prefer one package for depression over another. Two studies currently in the field will have an important bearing on this position. The Randomised Evaluation of the Effectiveness and Acceptability of Computerised Therapy (REEACT) Trial is due to end in 2013 (ISRCTN91947481) and Computerised CBT for Common Mental Disorders: RCT of a Workplace Intervention (ISRCTN 24529487) concludes in August, 2011. They will partly meet the need highlighted here for more RCTs on specific cCBT packages, both in relation to outcomes and in relation to costs. As new packages become available, their effectiveness will also need to be established, particularly if they charge for use.