Good guidelines but should the emphasis be on implementation?

The recent launch of the College guidelines on the management of patients who present with deliberate self-harm (Carter et al., 2016) is an opportunity to welcome useful advice on how to help this population, but it also prompts the broader question, ‘Should the College be producing guidelines such as these and those on mood disorders (Malhi et al., 2015) and schizophrenia (Galletly et al., 2016) at all?’ Self-harm is an important problem because it is common, it is associated with significant distress, it is linked to suicide and there is a high premature mortality from all causes after self-harm. The population of people who present to hospital with deliberate self-harm is one of the highest risk groups for subsequent suicide that any mental health service will encounter.

What is in the guidelines?

One of the difficulties in this area, particularly for non-clinicians, is defining exactly what is meant by self-harm, and the guidelines clearly describe the issues. The authors of the guideline have decided to retain the term ‘deliberate’ in the description of self-harm, a term that has been dropped outside Australia and New Zealand as service users dislike the negative connotations that are associated with the word. However, it is welcome that the authors have excluded from the guideline trials which focused specifically on the management of non-suicidal self-harm – a diagnosis with poor inter-rater reliability and which assumes that people’s motivation for self-harm can be dichotomised into suicidal or non-suicidal. In practice, motivation is hard to assess, there may be more than one motive and it often changes over time.

The guideline contains 46 recommendations, many of which relate to service design rather than the details of which particular treatment to offer. As such, these are based on expert consensus rather than the results of randomised controlled trials. Probably the most important is that everyone who presents to hospital with deliberate self-harm should be seen by a mental health professional and that risk assessment scales should not be used to decide who gets treatment. These recommendations are also made in the National Institute for Health and Care Excellence (NICE) guidelines on the management of self-harm (www.nice.org.uk/guidance/cg133). Alongside these recommendations are important statements about the need to collect routine high-quality data about the extent of the problem especially among Maori, Aboriginal and Torres Strait populations.

Are the guidelines any good?

There are guidelines about guidelines (sic) which help to answer the question ‘are these guidelines any good?’ The best known of these is the Appraisal of Guidelines for Research and Evaluation better known as AGREE II (www.agreetrust.org/). This lists 23 questions in six different domains which include scope and purpose, stakeholder involvement, rigour of development, clarity of presentation, applicability and editorial independence. Applying the AGREE II to the College self-harm guidelines, the scope and purpose are clearly specified. The guidelines are for psychiatrists and other health professionals who assess and treat people after deliberate self-harm in Australia and New Zealand and include guidance on assessment, clinical treatment, aftercare and organisation of services. Stakeholder involvement appeared to include a range of professional, public and other groups although the contribution of service users and carers was unclear. The guidelines were rigorously developed with systematic methods used for searching the evidence. Of the 46 recommendations, however, only 15 are based on evidence from randomised controlled trials which reflects the fact that assessing the best way to organise services is difficult. Unfortunately, there appears to be no procedure in place for updating the guidelines. The guideline is clearly presented although it is weak in describing how the guideline should be applied, the resource implications of applying the recommendations and how implementation of the recommendations should be monitored.

Is there any content missing?

Self-harm is often a sign of having had a difficult life, and many of the risk factors for self-harm also increase the risk of developing non-psychiatric disorders. It is not a surprise that while half of the premature mortality following self-harm is due to suicide, the other half is due to premature mortality from other causes including cardiovascular disease and cancer (Ostamo and Lonnqvist, 2001). The prevention of premature mortality and morbidity from non-psychiatric causes after self-harm is rarely researched and as a consequence does not appear in these or other guidelines on the problem. However, logically preventing premature mortality from non-psychiatric causes should be given equal weight to preventing suicides. This should be a topic for further research and may need to rely on a change of thinking from self-harm being seen as a mental health problem to it being considered as a health problem that is an indicator of ill health in multiple domains.

There is also a need for consideration of how the guidelines relate to existing guidance in this area. For example, the document ‘Working with the suicidal person’ produced by the Victorian Government Department of Health recommends only that emergency physicians consider referral to mental health professionals and that risk assessments guide treatment planning (www2.health.vic.gov.au/about/publications/policiesandguidelines/suicide-guidelines-working-with-suicidal-person). This is very different to what is recommended in the new College guidelines. So what should clinicians in Victoria do? Some statement from the College would be helpful here.

Should the College be producing guidelines at all?

Given that these are generally ‘good’ guidelines, the question arises should the College be in the business of producing guidelines at all? There are several arguments against medical Colleges producing guidelines. The first is that they are expensive in terms of time and money to produce and disseminate. To produce robust guidelines means having the resources to collate and appraise the literature on a topic. In the United Kingdom, the National Collaborating Centre for Mental Health, which had as its core task the writing of guidelines on the treatment of mental health disorders, had a 2014 budget of about AU$2.1 million but only produced about two guidelines a year. However, the direct costs to the Royal Australian and New Zealand College of Psychiatrists (RANZCP) of producing these self-harm guidelines were low as most of the work was voluntary, although relying on individuals’ generosity with their time is unlikely to be sustainable over time.

The second argument for not doing guidelines is that everyone else is doing them – often other organisations that are better funded and have as their sole purpose the production of guidelines. An obvious example is NICE in the United Kingdom whose mission is to produce guidelines on the management of different disorders. They already produce guidelines on the management of self-harm which have similar recommendations to the new College guidelines. The Cochrane Collaboration produces systematic reviews of the effect of treatments, and their methodology is seen as the gold standard. While Cochrane reviews are not guidelines, they clearly play a role in informing treatment and are often quoted in guidelines. It is unfortunate that three major Cochrane reviews on the treatment of self-harm were published in the last 12 months. This represents at the very least an expensive duplication of effort with the College guidelines. Also other psychiatric colleges around the world are not in the business of producing guidelines. For example, the American Psychiatric Association has only produced one guideline in the last 5 years and the UK Royal College of Psychiatrists only produced guidelines through separate contracts with NICE. An alternative argument for medical Colleges being involved in the guideline business is that they should focus on dissemination and implementation in a local context rather than the production of new guidelines.

Third, there needs to be a clear process for deciding what to write guidelines about and how to update them so that they stay relevant. One of the concerns about the College self-harm guidelines is that there does not seem to be any process in place for reviewing and updating them. Again this is expensive, and there needs to be some justification for how different disorders are prioritised and the frequency of updates.

Last, the College lacks the mandate to oversee implementation and monitoring as it does not fund or provide health services. Where medical Colleges develop guidelines on their own, they will inevitably be weak on how to put them into practice. This may explain why the guidelines offer little in describing how they should be applied which is a concern as most of the recommendations are about service delivery rather than technical aspects of which treatment to offer. Part of dissemination of the guidelines should involve a partnership between the College and health providers to assess how the guidelines are implemented and monitored.

So, to conclude, the guidelines are well written and contain sound advice. However, they are weak in how they should be applied which reflects the problem of whether medical Colleges should be in the business of producing guidelines at all.