The editorial by Wilson et al ^{Reference Wilson, Murray, Harris and Brown1} highlights important dilemmas faced by mental health professionals in relation to reporting violence perpetrated by mentally disordered patients. We welcome the proposals made by the authors, but unless there is a significant change in staff attitudes to reporting non-trivial violence perpetrated by psychiatric patients, progress in this area is unlikely to occur.

Our observation is underpinned by the results of two surveys which we carried out in a medium secure unit in Middlesbrough in 2006 and 2008. There were 80 incidents of assaults on staff by in-patients, the majority of incidents having been perpetrated by a minority (2006: 43 assaults, n = 10/100; 2008: 37 assaults, n = 14/100). Despite being a medium secure unit, the majority of assaults were perpetrated by patients detained under Part 2 of the Mental Health Act and by female patients. Only 10 incidents (12.5%) were reported to the police, despite 70% of nursing staff being aware of the memorandum of understanding (www.cp3.gov.uk/publications/agencies/mounhs.html). We explored the attitudes of nursing staff using self-report attitude questionnaires (each of the 13 attitude statements measured on a 5-point Likert scale) to identify enablers or barriers to reporting incidents.

In both surveys, approximately a third of respondents feared that reporting incidents would result in a breakdown of therapeutic relationships with patients and a half feared reprisal from patients following reporting. In 2006, half of respondents considered being assaulted as an ‘occupational hazard’, but encouragingly this attitude was reported only by a quarter of respondents in 2008. Although 84% of nursing staff understood that they had a ‘right to report’, a fifth believed that reporting incidents was a bureaucratic exercise without any benefits and for 60% the required reporting forms and procedures were difficult to complete. Staff were more likely to report incidents perpetrated by patients with personality disorder than those with other mental illness. About 20% of staff stated that they would only report incidents which resulted in physical injury. Only 40% believed that reporting incidents would strongly deter patients from re-assaulting. Some of these free-text comments capture the ambivalence in this area: ‘I came to the nursing profession to help patients, not to be a punch bag’; ‘I would report only if the assaults were due to “badness” not “madness’”; ‘Disillusioned towards the police dealing with incidents’; ‘Waste of time’; ‘Zero tolerance should mean zero tolerance’.

In summary, whereas we acknowledge the value in developing robust policies, procedures and systems to address this important issue, significant progress in this area is unlikely to occur unless considerable efforts are made to shift attitudes of mental health professionals. Campaigns and systems to report and reduce violence are akin to taking a horse to water. Making a change will require a change of attitudes in relation to reporting violent incidents to the police. We propose that this can be achieved by discussing patient assaults in staff induction training, appraisal, supervision sessions and trust audits.