Formal reports of violence against health staff are on the increase (National Audit Office, 2003) despite the Zero Tolerance Zone Campaign (Department of Health, 1999) message to the public that aggression, violence and threatening behaviour would no longer be tolerated in the National Health Service.

Variation in reporting makes it impossible to say with certainty how far the increase in reported violence reflects an actual increase in incidents, changes in staff reporting practice, or is a true reflection of how trusts are performing in addressing aggression and violence.

The National Audit Office (2003) estimated the direct cost of work-related violence and aggression as £69 million in 2001-2002. This figure does not include staff replacement costs and compensation claims. The human costs include demoralisation, high staff turnover and sickness rates, and deterioration in the service delivered to patients.

Managing aggression has been a challenge for mental health services for many years, but the challenge has been compounded in recent years by increases in substance misuse, the use of weapons and violence in society generally. A Royal College of Psychiatrists’ working party on safety for psychiatrists has recently reviewed the safety literature in this field (Royal College of Psychiatrists, 2006).

Studies from the USA (Reference Black, Compton and WetzelBlack et al, 1994; Reference Schwartz and ParkSchwartz & Park, 1999), Canada (Reference Chaimowitz and MoscovitchChaimowitz & Moscovitch, 1991) and Belgium (Reference Pieters, Speybrouck and De GuchtPieters et al, 2005) showed prevalence rates of physical assaults to psychiatric trainees of between 26 and 56%. These studies varied greatly in terms of the prevalence period surveyed and the definition of ‘violence’ used. In addition, all these studies concentrated on psychiatric trainees. In the UK (South Wales), there has been just one study in the past 10 years which systematically studied violence to psychiatrists of all grades (Reference DaviesDavies, 2001). In this study 17% of respondents reported one or more assaults over 1 year, with senior house officers (SHOs) significantly more likely to have experienced an assault or threat.

The object of our survey was to establish the number of assaults in 1 year on all grades of psychiatrists in a London mental health trust. We also related the number of such assaults in 1 year to the grade, specialty, previous training in the prevention and management of aggression and the circumstances of the aggression.

Method

We undertook a retrospective questionnaire study of violent assaults in the past 12 months on psychiatrists of all grades in a large London mental health trust, where all sub-specialties of psychiatry are represented. Assaults were defined as in the South Wales study (Reference DaviesDavies, 2001) as actual physical attacks, for example ‘hit, struck with an object, shake, pushed, throttled, etc.’ and not threats or other verbal aggression alone.

A questionnaire was developed which was informed by previous research, especially Davies (Reference Davies2001), and by discussions held at the North Thames Regional Psychiatric Committee. The first part of the questionnaire addressed demographic details, experience of working in psychiatry, specialty and psychological attitude to assaults from patients. The second part of the questionnaire was completed for any assault suffered in the past 12 months.

Statistical analysis was performed using χ²-tests with a level of significance of P<0.05. Power calculations were undertaken to establish the risk of type II error.

Results

We approached all 199 psychiatrists identified as being employed within the trust at the time of the survey. We had 129 returned responses (64.8%).

Assaults

Of the 129 psychiatrists who completed the questionnaire, 16 (12.4%) had been assaulted by a patient in the previous 12 months. Of these, 3 (18.7%) had been assaulted twice within this period (all SHOs). No doctor had been assaulted more than twice within the survey period. Eleven (68.7%) of the 16 victims were males. Males constituted 59.7% of the responding psychiatrists.

Of the 19 assaults, 11 were on SHOs (57.9%) and 6 (31.6%) were on consultants. Overall 11 of the 34 SHOs (32.4%) were assaulted and 6 out of 59 consultants (10%). The difference between SHOs and consultants in terms of likelihood of being assaulted was statistically significant (χ²=5.698, d.f.=1, P<0.02, two-tailed). Only 1 assault (5.2%) was on SpRs and 1 (5.2%) on staff grade doctors. The sub-specialty and grade of those assaulted is shown in Table 1.

Table 1. Sub-specialty and grade of those assaulted

	Senior house officer n (%)	Specialist registrar n (%)	Consultant n (%)	Staff grade n (%)	Associate specialist n (%)
General adult	10	0	2	1	0
Forensic	0	1	3	0	0
Old age	0	0	1	0	0
Learning disability	1	0	0	0	0
Total	11 (57.9)	1 (5.2)	6 (31.6)	1 (5.2)	0 (0)

The majority of the assaults were on those working in general adult psychiatry (13 out of 19; 68.4%), followed by forensic psychiatry (4 out of 19; 21.1%). Out of the 49 psychiatrists working in general psychiatry, 13 were assaulted compared with 4 of the 38 in forensic psychiatry. However, the difference was not statistically significant (χ²=2.54, d.f.=1, P=0.11, power 0.813). One assailant was being treated within each of old age, general medical and prison services.

Attitudes and vulnerability to assaults

We explored the relationship between four different attitudes of psychiatrists to violence by patients and vulnerability to attacks (Table 2). There was no statistically significant difference between those assaulted and those not assaulted in terms of these attitudes.

Table 2. Distribution of attitudes of surveyed psychiatrists towards assaults by patients

	Those assaulted		Those not assaulted
Attitude	Agree n (%)	Disagree n (%)	Agree n (%)	Disagree n (%)	P	Power
Occasional assaults an acceptable hazard of work	5 (31.3)	11 (68.8)	18 (16.0)	95 (83.2)	0.25	0.67
Assaults can be predicted and hence avoided	11 (68.8)	5 (31.3)	74 (65.5)	37 (32.7)	0.91	0.91
My manner with patients makes me vulnerable to assaults	0	16 (100)	10 (9.0)	101 (91.4)	0.45	1
I take needless risks with potentially violent patients	0	16 (100)	7 (6.2)	103 (91.2)	0.65	1

Discussion

More than 1 in 10 (12.4%) of the 129 responding psychiatrists were assaulted in the 1-year study period. Senior house officers were the most vulnerable to assaults (11 out of 34, 32.4%). Davies (Reference Davies2001) found a higher rate (17%) of assaults in 1 year among all responding psychiatrists but a lower rate (28.6%) among SHOs. The next most commonly assaulted group in our study was the consultants, of whom 6 out of 59 (10%) were assaulted. However, they were significantly less likely to be assaulted than SHOs during the year of the study. The fact that SpRs and staff grade doctors were the least likely to be assaulted does not suggest a straightforward direct inverse relationship between seniority and vulnerability to assaults. Differences between grades of psychiatrists and their vulnerability to assaults could, of course, also be due to differences in exposure time to patients.

We anticipated that doctors were more likely to be assaulted during an urgent rather than a routine assessment. This was the finding in the study by Davies (Reference Davies2001), where half of assaults occurred during urgent assessments. We were, however, surprised that most of the assaults in our study occurred during a routine in-patient assessment or during casual contact on the ward. Only 1 attack (5.3%) occurred during an urgent out-of-hours assessment. Most of the assaults occurred in the psychiatric ward. Also, of note was that 3 out of 16 (15.8%) of the assaults occurred on the general medical wards and none in the accident and emergency department. One assault had occurred during an assessment in prison despite the presence of a prison officer.

Overall, these findings may reflect the improvement in recent years in policies, procedures and security related to urgent assessments, including those in accident and emergency departments, and do not necessarily imply an increase in assaults in psychiatric wards. However, our study suggests that greater attention might now need to be directed to practice on psychiatric wards.

As many as 16 out of the 19 assailants had a history of previous assaults. Despite this information being available at the time of assessment in the majority (62.5%), a further assault was not prevented.

Only two of the victims required medical treatment after the assault, first aid in one case and accident and emergency treatment in the other. Only 3 out of 16 victims had any kind of psychological support. This was informal support from colleagues. This may reflect the true need or reluctance to seek formal support after an assault. It may also reflect deficiencies in the organisation's response to such incidents, including ensuring formal support is offered even if this is not taken up.

Although attendance at courses on the prevention and management of violence would seem prudent, and annual attendance is in theory mandatory for psychiatrists in most trusts, it did not significantly reduce the risk of being assaulted in our survey. This might reflect the sporadic and often unpredictable nature of such assaults. In our survey psychiatrists of senior grades were less likely to attend courses but we did not explore differences in efficacy between such courses.

In our survey most psychiatrists did not consider that they took unnecessary risks or that occasional assaults are an acceptable hazard of work. Two-thirds of psychiatrists believed assaults could be predicted and thus avoided. Even if, as a generalisation, this proposition were true, clinical experience suggests assaults by psychiatric patients, especially if psychotic, can be impulsive and/or driven by internal positive psychotic symptoms, and thus outwardly appear unpredictable, and cannot be necessarily prevented by good communication skills, including verbal talk-down, or policies, procedures and security measures.

Although statistically there was no difference in such attitudes between those assaulted and those not, we noted that none of the psychiatrists assaulted believed that they took unnecessary risks with patients or felt that there was something in their approach that made them vulnerable to attacks. This may be a retrospective psychological defence as, of those not assaulted, 6.8% considered that they did take needless risks with potentially violent patients.

Declaration of interest

None.