Article Text

Download PDFPDF

Manual assisted cognitive behaviour therapy is as effective as treatment as usual for deliberate self harm, but is more cost effective
  1. Robert Goldney, MB, BS, MD, FRANZCP, FRC Psych
  1. Professor of Psychiatry, University of Adelaide, Australia

Statistics from

OpenUrlCrossRefPubMedWeb of Science; and OpenUrlCrossRefPubMedWeb of Science

Q Is manual assisted cognitive behaviour therapy more effective and more cost effective than treatment as usual in people with recurrent episodes of deliberate self harm?


Embedded Image  Design:

Randomised controlled trial.

Embedded Image  Allocation:


Embedded Image  Blinding:


Embedded Image  Follow up period:

12 months.

Embedded Image  Setting:

Five centres in Glasgow, Edinburgh, Nottingham, West London, and South London; recruitment May 1998 to April 2000.

Embedded Image  Patients:

480 people, aged between 16 and 65 years, presenting with recurrent (at least 2 episodes) deliberate self harm. Exclusions: diagnosis of a psychotic or bipolar disorder; needing in-patient psychiatric treatment; primary diagnosis of substance dependence.

Embedded Image  Intervention:

Participants were randomised to manual assisted cognitive behaviour therapy (MACT; 239 people) or treatment as usual (TAU; 241 people). The MACT group received a booklet based on cognitive behavioural therapy (CBT) principles and were offered up to 7 sessions with a MACT trained therapist. The TAU group was offered the standard treatment available in each location, including: problem solving approaches (Nottingham); dynamic psychotherapy (South London); GP or voluntary group referral (West London and Edinburgh); or short term counselling (Glasgow). Members of the TAU group already receiving standard care were offered continuation of this care. Assessment interviews were carried out at 6 and 12 months. Incidences of self harm were self reported during a modified Linehan Parasuicide History Interview and verified using GP notes and A&E records. Economic data were collected at baseline, 6 and 12 months, using a modified Client Service Receipt Inventory. Data collected covered all service providing sectors, productivity losses, accommodation and living costs.

Embedded Image  Outcomes:

Efficacy: a repeat self harm episode within the following year. Cost: cost per patient over 12 months. Cost effectiveness: incremental cost effectiveness ratio (ICER) over 12 months.

Embedded Image  Patient follow up:

Efficacy analysis 84%; cost analysis 83%.


Efficacy: at 12 months, the proportion of participants repeating deliberate self harm was not significantly different between groups (AR 39% with MACT v 46% with TAU; OR 0.78, 95% CI 0.53 to 1.14; analysis not by intention to treat). Cost: MACT was significantly cheaper than TAU over 6 months, but did not maintain significance over 12 months (6 month mean difference in cost per patient: −£897, 95% CI −£1747 to −£48; 12 month mean difference in cost per patient: −£838, 95% CI −£2142 to £466). Cost effectiveness: the ICER was −£120 per 1% reduction in the proportion of participants with a repeat self harm episode. The likelihood that MACT is more cost effective than TAU, based on cost effectiveness acceptability curves, is over 90%.


MACT is as effective as usual treatment in reducing recurrences of deliberate self harm, however, it is more cost effective.


Authors note that there were sociodemographic differences between participants included and those excluded from the economic analysis due to missing economic data. There was a centre bias among excluded participants, and they were younger and had cost less in the 6 months prior to baseline. This may affect generaliseability of results.


These companion papers by Tyrer et al and Byford et al are excellent models of psychiatric research which have produced sobering and disappointing results in terms of effecting a reduction in deliberate self harm. However, it would be reassuring to health administrators that there may be a reduction in health care expenditure with the manual assisted cognitive behaviour therapy.

If such a well designed study cannot demonstrate an improved clinical outcome, a finding which is consistent with the majority of previous research, questions arise not only about the wisdom of pursuing further such studies, but also whether the correct issues are being addressed.

Consider the subjects of the study: they have had a previous episode, and therefore they would be a more homogeneous group; they did not require in-patient care, so they may be less ill than some seen in practice; and they did not have a psychotic disorder or bipolar disorder, or a primary diagnosis of substance dependence, therefore precluding a further substantial group seen in practice. And yet they are a patient group united by deliberate self harm, a behaviour related to many disparate stressors. Is such a behaviour sufficient to form a cohesive group worthy of such intensive study?

In 1828 George Burrows1 wrote that “The medical treatment of the propensity to suicide, whether prophylactic or therapeutic, differs not from that which is applicable in cases of ordinary insanity.” This advice to address the underlying condition appears to have been overlooked in our attempts to demonstrate our ability to influence a behaviour.

It is doubtful whether a better study design focusing on deliberate self harm could be implemented. Indeed, perhaps the most important implication from this study is that future psychiatric research should focus on the antecedents of deliberate self harm, that is those conditions which are associated with it rather than the behaviour itself.


Supplementary materials

  • Web-only Table

    The table is available as a downloadable PDF (printer friendly file).

    If you do not have Adobe Reader installed on your computer,
    you can download this free-of-charge, please Click here


    Files in this Data Supplement:

    • [View PDF] - Aripiprazole significantly improved overall PANSS score and CGI-S score compared with placebo


  • For correspondence: Professor Peter Tyrer, Imperial College, St Dunstan’s Road, London, UK

  • Source of funding: The Medical Research Council, UK.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.