Authors' objectives

To assess the effectiveness of computer-based cognitive behaviour therapy (CCBT) for treating anxiety, depression and phobias, and to compare the cost-effectiveness of CCBT with cognitive behaviour therapy (CBT) by conventional methods and treatment as usual (TAU).

Searching

Seventeen bibliographic databases, including grey literature and current research, were searched from 1966 to September 2001 for English and non-English language studies. The authors reported the individual databases and search terms, and also listed the health services research-related resources searched via the Internet. The Science Citation Index and Social Sciences Citation Index were used for citation searches on key papers and authors. The bibliographies of relevant articles and sponsor submissions were examined.

Study selection

Assessment of study quality

The RCTs were scored for quality according to criteria developed by Jadad et al. (see Other Publications of Related Interest no.1). Non-randomised studies were assessed according to criteria modified from the Users' Guides to Evidence-Based Medicine (see Other Publications of Related Interest no.2). The quality assessment was carried out unblinded to author, institution and journal. The authors do not state how many of the reviewers performed the quality assessment.

Data extraction

The data were extracted by one reviewer using a customised data extraction form, and checked by another. Any disagreements were resolved by discussion. Depending on the data available in the individual studies, the authors extracted the mean (and standard deviation) pre- and post-treatment scores on the outcome measures, between-group difference scores and effect size.

Methods of synthesis

Results of the review

Nineteen studies were included. Sixteen of the included studies (n=953) involved CCBT without any other psychological therapy: 11 RCTs (n=793), 2 cohort studies (n=64), 2 pilot studies (n=53) and one comparative study (n=43). There were 3 studies (n=101) where CCBT was used as an adjunct to CBT with a therapist: one cohort study (n=96) and 2 very small studies (n=1 and n=4).

Psychological symptoms and interpersonal and social functioning (some studies had more than one comparator, usually TCBT and TAU).

CCBT versus TCBT (7 studies): there was no difference between CCBT and TCBT in six of the studies, with CCBT being as good as TCBT. One study found TCBT to be more effective than CCBT.

CCBT versus TAU (7 studies): CCBT was more effective than TAU in four studies. Three studies found no significant differences between CCBT and TAU, though one of these found a significant improvement in the CCBT group (but not the TAU group) at one month' post-treatment; there was no difference at later follow-up.

CCBT versus bibliotherapy (2 studies): one study found no difference between CCBT, TCBT and bibliotherapy, with all three groups showing equal improvement. The second study found bibliotherapy to be more effective on two of the six outcomes.

Patient preference, satisfaction and acceptability (12 studies).

In one study no patients expressed regret at receiving CCBT. One study reported no differences in satisfaction between the CCBT and TCBT groups. Three studies reported drop-out or low participation in patients receiving CCBT, with counselling being more popular in one of the studies. In a further study two patients said the computer made things worse, while in another study the therapist was viewed more positively than CCBT or bibliotherapy. In the remaining five studies the views regarding CCBT were positive.

Therapist time (6 studies).

Of the six studies which reported therapist time in an appropriate format, four reported a marked reduction in therapist time for the CCBT group, one reported more therapist time for the CCBT group than the TCBT group, and one only reported time for the CCBT group (45 minutes) and not the TAU group.

Studies of CCBT as a treatment adjunct.

A significant reduction in the patient's panic and anxiety was reported in the case study. In the second very small study, one patient dropped out and three showed improvement. The final study also showed improvement in patients receiving CCBT in combination with TCBT.

Cost information

The economic modelling was based on the evidence provided in four sponsor submissions in relation to the following programmes: Stresspac, Cope, FearFighter and Beating the Blues. In the first year of implementing Beating the Blues, the costs with an assistant psychologist were £21,691 and with a practice nurse £25,192. Stresspac and FearFighter cost £19,902 and £22,574, respectively. The overall costs for Beating the Blues were estimated to be £275 million in England and 13 million in Wales when employing an assistant psychologist, and £237 million and £11 million, respectively, with a practice nurse. Stresspac was estimated to cost £206 million in England and 10 million in Wales. The authors state that the model estimates should be viewed with caution due to the data deficiencies and the large number of assumptions made.

Authors' conclusions

The authors state that the included studies were heterogeneous in terms of the setting, patient populations, comparators and outcomes measures, and that the evidence for the effectiveness of CCBT is uncertain. They also state that there is limited evidence that CCBT may be effective in the treatment of depression, anxiety and phobias, and this evidence is of poor to moderate quality.

CRD commentary

The review question was well-defined in terms of the intervention, participants, the outcomes of interest and study design. A number of relevant electronic databases were searched and the subject headings used in the search strategy were given. The authors also sought unpublished data and studies were not excluded on the basis of language. There was limited information on the study selection and quality assessment components of the review process. The data extraction was carried out by one reviewer and checked by a second, which helps to reduce errors and any bias. Sufficient information on each of the included studies is provided in both the text and tables, and the findings are discussed in the context of study quality and heterogeneity. The authors' conclusions appear to follow from the evidence presented.

Implications of the review for practice and research

Practice: The authors did not state any implications for practice.

Research: The authors make extensive recommendations regarding future research; further details are provided in the paper. They state that the extent of therapist involvement required to produce optimal outcomes for patients using CCBT needs to be investigated.

There is a need for studies carried out within the general practice setting and, in particular, which include patients with co-morbidities routinely treated within that setting. In addition, patients from a range of socioeconomic and ethnic backgrounds, and of different age groups and gender, need to be included. Research should also include patients who cannot access current services because they are housebound.

The role of CCBT within a stepped-care programme needs to be clarified, in addition to how it relates to other methods used to increase access to psychological therapies.

Approaches to treatment aimed at reducing therapist time, such as bibliotherapy and other self-help approaches, should be used as a comparator in studies.

Co-morbidity and medication need to be taken into account when designing studies. Other design issues include research carried out by independent researchers, good-quality RCTs of adequate power, appropriate comparison groups, and well-validated outcome measures.

Further research into the components of CCBT packages, such as the most appropriate length of sessions and software and hardware interface, is also required.

Funding

NHS R&D Health Technology Assessment (HTA) Programme, project number 01/23/01.

Bibliographic details

Kaltenthaler E, Shackley P, Stevens K, Beverley C, Parry G, Chilcott J. A systematic review and economic evaluation of computerised cognitive behaviour therapy for depression and anxiety. Health Technology Assessment 2002; 6(22): 1-100. [PubMed: 12433315]

Original Paper URL

http://www.hta.ac.uk/1254

Other publications of related interest

1. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1-12. 2. Levine M, Walter S, Lee H, Haines T, Holbrook A, Moyer V. How to use an article about harm. JAMA 1994;271:1615-9. 3. Kaltenthaler E, Parry G, Beverley C. Computerized cognitive behaviour therapy: a systematic review. Behav Cogn Psychoth 2004;32:31-55.

Indexing Status

Subject indexing assigned by NLM

MeSH

Adolescent; Adult; Aged; Anxiety /diagnosis /therapy; Behavior Therapy /economics /methods; Cognitive Therapy /economics /methods; Cost-Benefit Analysis; Depressive Disorder /diagnosis /therapy; Female; Humans; Male; Manifest Anxiety Scale; Middle Aged; Prognosis; Psychiatric Status Rating Scales; Randomized Controlled Trials as Topic; Risk Assessment; Sensitivity and Specificity; Severity of Illness Index; Therapy, Computer-Assisted /economics /methods; Treatment Outcome

AccessionNumber

12003008033

Database entry date

30/09/2003

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.