Response

In their editorial summarising key points from the recent NICE guideline on management of self-harm,¹ Mughal and colleagues^{Reference Mughal, Burton, Fletcher, Lascelles, O'Connor and Rae2} lead with the need for empathy – a rallying call that seems on the face of it unproblematic. It is however underspecified here (as it so often is in psychiatry) both in terms of empathy's defining features and what exactly should be the interventions to ensure it happens.

Interestingly the word ‘empathy’ appears for the first time only in the supplement to the OED.³ It has gained currency since then but with a rather blurry feel-good meaning. The OED defines empathy as: ‘The power of entering into the experience of […] emotions outside ourselves’. An early use referred to the experience of a work of art, the ability to ‘feel oneself into it’.³ These definitions too are rather hard to grasp but they point to a state that is unrealistic to expect a clinician to achieve – especially in relation to somebody met only briefly and in unusual circumstances, whose emotions and responses to them are likely to arise from experiences well outside the clinician's own. Instead, I suggest the need for three attributes that are clustered around the general idea of sensitive and non-aversive care.

First is courtesy and professionalism. These, unlike empathy, can be taught and supervised by attention to behaviour – how to introduce yourself, how the patient wants to be addressed, attention to privacy and confidentiality and so on. And if that fails, there is a question about lack of professionalism, for which there are other remedies.

Second is being well-informed about causes and consequences. My own experience suggests that many clinicians are not au fait with what is now known about reasons for self-harm – in the sense of what its functions might be. One indication is the frequency with which discussions centre around diagnosis, which is – except in a minority of cases – no help in understanding what is going on. Another is the persistence of stereotypes about self-injury. It is difficult to undertake a sensitive and meaningful psychosocial assessment if you don't know what you're looking for. I wouldn't downplay the importance of person-centred care or the value of service-user involvement in training, but clinically oriented postgraduate education also needs to develop in this space. This sounds too fact-based to have anything to do with ‘empathy’ but then I have always thought that a better word than ‘empathy’ is ‘sympathy’ – the sense of feeling onside with somebody, that comes from a shared understanding of the situation. And how to elicit that shared understanding can be taught.

Third is the question of competence in practice, and here there is a real challenge. Hardly anybody provides comprehensive psychological or psychosocial treatment services in the post-acute period – not in liaison psychiatry where most acute presentations are seen, nor in clinical psychology, nor in CMHTs. Good management involves, for sure, a professional attitude and behaviour and sound knowledge both about therapies and about the specific problem being tackled – but also the generic (transferable) skills, behavioural and emotional repertoire and expertise that come from experience. How can we enhance care in this area if we don't provide the services within which all this can be developed? ED-SAFE 2^{Reference Boudreaux, Larkin, Vallejo Sefair, Ma, Li and Ibrahim4} does not convince as a model answer, notwithstanding the authors’ enthusiasm.

The intentions of my colleagues on the NICE group are sound, but if we are to improve how we treat people we need to go beyond rather general appeals to good practice. We need to develop self-harm services and to specify the curriculum for education and supervised training that will develop those working in such services to deliver sensitive and non-aversive care.