Asking about self-harm during risk assessment in psychosocial assessments in the emergency department: questions that facilitate and deter disclosure of self-harm

Abstract

Background Emergency departments are key settings for suicide prevention. Most people are deemed to be at no or low risk in final contacts before death. Aim To micro-analyse how clinicians ask about suicidal ideation and/or self-harm in emergency department psychosocial assessments and how patients respond. Method Forty-six psychosocial assessments between mental health clinicians and people with suicidal ideation and/or self-harm were video-recorded. Verbal and non-verbal features of 55 question–answer sequences about self-harm thoughts and/or actions were micro-analysed using conversation analysis. Fisher's exact test was used to test the hypothesis that question type was associated with patient disclosure. Results (a) Eighty-four per cent of initial questions (N = 46/55) were closed yes/no questions about self-harm thoughts and/or feelings, plans to self-harm, potential for future self-harm, predicting risk of future self-harm and being okay or keeping safe. Patients disclosed minimal information in response to closed questions, whereas open questions elicited ambivalent and information rich responses. (b) All closed questions were leading, with 54% inviting no and 46% inviting yes. When patients were asked no-inviting questions, the disclosure rate was 8%, compared to 65% when asked yes-inviting questions (P < 0.05 Fisher's exact test). (c) Patients struggled to respond when asked to predict future self-harm or guarantee safety. (d) Half of closed questions had a narrow timeframe (e.g. at the moment, overnight) or were tied to possible discharge. Conclusion Across assessments, there is a bias towards not uncovering thoughts and plans of self-harm through the cumulative effect of leading questions that invite a no response, their narrow timeframe and tying questions to possible discharge. Open questions, yes-inviting questions and asking how people feel about the future facilitate disclosure.