We thank Dr Hubbeling for her interest in our article (Kapur et al. Reference Kapur, Hunt, Windfuhr, Rodway, Webb, Rahman, Shaw and Appleby2012). She questions the relevance of treatment setting in suicide risk and prevention. We would disagree that the investigation of treatment setting is uninformative. The aetiology of suicide is complex and requires multifaceted prevention programmes. Treatment setting is crucial in that it helps services to focus on specific preventative efforts. Indeed, as Hubbeling herself notes, preventative actions vary according to where care is being provided. For example, preventing access to ligature points is more relevant to in-patient suicide, whilst limiting the availability of certain drugs might have a greater effect on community suicides.

Whether risk is transferred to different settings is an important point raised by Hubbeling. We explored this in our study and found an increase in the rate of post-discharge suicides, and in the number of suicides under crisis resolution home treatment (CRHT) teams. As the respondent suggests, with the substantial fall in in-patient suicides, there could feasibly have been a compensatory increase in suicides among community mental health patients. We were unable to provide rates of suicide among community patients due to a lack of appropriate denominators. However, findings from our recent report showed a sustained fall in the rate of suicide among all mental health patients in England between 2004 and 2009 (Appleby et al. Reference Appleby, Kapur, Shaw, Hunt, Flynn, While, Windfuhr, Williams, Rodway, Roscoe and Rahman2012) – a fall which is unlikely to have been accounted for solely by the fall in in-patient deaths. Therefore we maintain our results probably do not indicate a transfer of risk to the community setting per se. However, vigilance is needed for patients under CRHT and this is one treatment setting that would certainly benefit from further research.