The learning disability service within Enfield Community Care NHS Trust has a 10-bed non-secure assessment and treatment unit, serving local and surrounding London health districts. It caters for people with challenging behaviour and mental health problems that cannot be managed in the community. The multi-disciplinary service team comprises psychiatrists, nurses, psychologists, occupational therapists, speech therapists and art and music therapists. The team meet weekly, to coordinate care plans and discuss issues such as violence; however, it is the nursing staff who deal with the overwhelming majority of violent incidents (all of those described in this study). The nursing staff have training in control, restraint and breakaway techniques relating to violence.

There is published guidance on how to manage violence in NHS settings (Department of Health, 2000), within general psychiatric settings (Royal College of Psychiatrists, 1998) and learning disability settings (Reference Harris, Allen and CornickHarris et al, 1996; British Institute of Learning Disability, 2000). Studies show this to be a widespread problem with no easy solution. A survey of attitudes of staff to offending behaviour among people with learning disability in Cambridgeshire (Reference Lyall, Holland and CollinsLyall et al, 1995) showed that tolerance of offending behaviour was extremely high. It also showed an apparent inability of the police to prosecute even when serious crimes (including sexual offences and assault) were reported. Alexander and Singh (Reference Alexander and Singh1999) stated that violent behaviour was the reason for over three-quarters of admissions to a learning disability inpatient service. Kiely & Pankhurst (Reference Kiely and Pankhurst1998) surveyed staff within the learning disability service of an NHS trust, assessing violence experienced over a 12-month period; they showed that 81% of staff within the service had experienced violence over the previous 12 months, and that new and inexperienced staff were particularly vulnerable. They offered suggestions for putting in place human resource strategies to reduce the incidence of violence and to provide appropriate post-incident support. Crichton (Reference Crichton1999) outlined the importance of moral judgement and staff attitude to disturbed behaviour in the understanding of how such behaviours are responded to and conceptualised.

We aimed to analyse the nature of — and the response to — violent and threatening acts and behaviours occurring within our in-patient learning disability service.

Method

Following a violent or threatening act, the staff member involved completes a critical incident form. This report should include circumstances leading up to the event, a description of the incident itself, and all interventions following the incident. The survey compiled the data from all forms completed over a 6-month period, from November 1999 to May 2000. The total number of incidents was recorded, and categorised into type of incident. The categories were: physical violence; aggression; racial abuse; and sexually threatening behaviour. For each incident a retrospective decision was made using knowledge of the patient and recorded inpatient notes of the patient's mental state around the time of the incident, to ascertain whether patients were aware of their actions, whether they knew what they were doing and whether they knew what they were doing was inappropriate or wrong. Although such criteria are based on a legal framework for establishing an insanity defence (the ‘McNaughton rules’; West & Walk, 1977), such additional information was thought to be potentially useful in view of what we perceived as a widespread popular belief that such patients are both unaware and unaccountable for their actions.

Police involvement was noted, looking at whether or not police officers attended the unit, to whom they spoke, whether the patient was interviewed, and any further action taken by the police, such as a caution, removal from the unit or arrest. We recorded whether the victim or a witness to the incident was prepared to make a statement of the event at the time. Short-term interventions at the time of the incident were assigned to one of the following categories: talking to the patient, behavioural measures, restraint, medication (for emergency sedation) or involvement of professional staff other than nurses.

The patients in this study group had a combination of problems that could explain the individual acts of violence or threatening behaviour. These included mental health problems, epilepsy, experiences of abuse or neglect, social disadvantage and personality dysfunction, and all had some degree of learning disability. Commenting on the individual causes of each incident is beyond the scope of this paper.

Results

The results are summarised in Tables 1 and 2. A total of 96 incidents were recorded, of which 55 involved physical violence, 25 involved aggressive intimidation, threatening behaviour or verbal abuse, 8 consisted of racial abuse and 8 of sexually threatening behaviour. In all cases it was felt that the patients at the time of the incident were aware of their behaviour and knew what they were doing was inappropriate or wrong.

Table 1. Analysis of violent incidents

	Physical violence	Aggression (intimidation, threats, verbal abuse)	Racial abuse	Sexually threatening behaviour
Number of incidents (total 96)	55	25	8	8
Patient aware of actions (%)	100	100	100	100
Police contacted	In 5 cases (9%)	No	No	No
Police action following contact	Did not attend (3 cases)
	Spoke to staff (2 cases)	No	No	No
Victim or witness prepared to make statement	No	No	No	No

Table 2. Analysis of interventions following violent incidents

	Physical violence n=55	Aggression n=25	Racial abuse n=8	Sexually threatening behaviour n=8	All-incidents n (%)
Talked to patient	2	8	2	7	19 (20)
Behavioural measures	8	1	Unknown	1	10 (10)
Restraint	18	0	Unknown	NA	18 (19)
Medication	27	11	Unknown	NA	38 (40)
Other professional involvement	6	5	Unknown	NA	11 (11)

The police were contacted in five instances, all of which involved physical abuse. They attended the unit on two occasions. On attending they spoke only to staff and not to the patient. No one involved in the two cases was prepared to make a statement or press charges. For this reason no interview occurred with an alleged offender and no comment can be made about the use of an ‘appropriate adult’ to help the accused in dealing with the police. Responses from individual police officers included comments that they could not pursue any matter involving a patient who has a learning disability and that it is impossible to press charges against a patient held under the Mental Health Act.

The most common intervention recorded was emergency sedation through medication (40% of cases). Talking to the patient was specifically recorded in 20% of cases. The involvement of other professionals was recorded in 11% and behavioural measures were recorded in 10% of cases. The majority of the recorded interventions were for incidents involving physical violence, performed by a small minority (three) of the patients. The small proportion of sexually threatening acts recorded all involved the same patient.

Discussion

A limitation of our study was the use of retrospective data obtained from critical incident forms. Staff acknowledge that they are more likely to fill them in following a serious incident, rather than following the more frequent incidents that are of a lower intensity. The incidents reported included damage to property and assaults on staff and fellow in-patients, some involving injury and subsequent staff sickness. The results show that a small group of individuals were responsible for the vast majority of the incidents.

The most popular intervention recorded was the use of medication, with ‘talking to the patient’ recorded as an intervention much less frequently. This probably reflects the tendency of staff to record interventions that they feel need to be justified at a later date. Similarly, the recorded involvement of other professional staff is low. The incidents themselves were all managed and recorded by nursing staff. Staff from other disciplines are involved in a more general way, as described previously. Future analysis of such incidents should look at these issues more closely.

In learning disability services the issue of violence is largely conceptualised under the broad category of ‘challenging behaviour’ (Reference Emerson, Barnett and BellEmerson et al, 1987). This concept was introduced in an attempt to replace the judgemental implications of previous terminology such as ‘behaviour problems’. It was intended to encourage service providers to take some ownership of the problems, try to understand them, and come up with solutions. This concept is widely accepted. It may militate against encouraging people with learning disability to be held accountable for their own actions. Our study has shown that in all incidents recorded, the patients involved were able to understand that their actions were wrong. Although a formal evaluation of legal insanity is a matter for the courts and not merely medical opinion, it does not necessarily obviate the need for a police inquiry and court disposal, especially in the cases of serious offences when a restriction order may be considered.

Although forensic learning disability services exist, they are available only for patients with criminal charges against them. By the end of this survey period no charge had been made against any patient within the learning disability service in Enfield within the memory of staff working there. It was apparent that there were misunderstandings among the local police force about issues regarding mental illness, learning disability and the Mental Health Act. Some members of the local police even stated the erroneous belief that in-patients in a mental health or learning disability setting (or both) are unable to be prosecuted. The authors believe that the police training on these issues varies across the forces, and is limited. When the police believe that allegations are unlikely to result in prosecution of an alleged perpetrator or when the perpetrator is already detained under the Mental Health Act, they sometimes express the view that there is little to be gained from carrying out an investigation (Reference Claire, Carson, Churchill, Brown and CraftClaire & Carson, 1997). All staff expressed reluctance to go through the process of pressing charges against a patient after violent incidents. This philosophy is shared with carers, relatives and the police themselves, in the authors' experience. If local police are unwilling to become involved, then it is the opinion of the authors and others (Reference Claire, Carson, Churchill, Brown and CraftClaire & Carson, 1997) that the service should ensure that a report of the incident is recorded by the police, with an added statement to the effect that important evidence may be lost if the police do not act promptly.

It seems both unfair and dangerous for people with learning disabilities to be placed on open wards with fellow patients who are both violent and threatening, when there are unsuitable resources to cope. Caring for a marginalised and disadvantaged group of people such as those with learning disability is stressful, but the additional issue of violence exacerbates the burden of care. The authors agree with those who advocate police involvement when a person with a learning disability commits a crime. We believe that just as people with learning disability should be valued and treated as normal citizens, they should also be held accountable when they commit crimes. Claire and Carson (Reference Claire, Carson, Churchill, Brown and Craft1997) have detailed how the reasoning behind a decision to prosecute is a fine balance of a number of competing pressures. They conclude that often the decision is not clear-cut and that the end result may be little different from the original situation. However, they state that a clear advantage would be that a criminal act would have been identified and acted upon within the law, and that such results provide considerable leverage with purchasers and commissioners to provide appropriate resources for such individuals. Norko et al (Reference Norko, Zonana and Philips1992), writing about patients in a general psychiatric setting, outlines arguments for and against police involvement. These arguments apply equally to people with learning disability. Arguments in favour include encouraging patient responsibility, improving staff morale and willingness to treat violent patients, deterring violence, allowing public scrutiny of violence in institutions, and possibly also representing a type of reality therapy by limit-setting interventions. Arguments against prosecuting patients include its impracticality; scapegoating patients for inadequacies in the treatment environment; the alienating effect on patients; subversion of the therapeutic alliance; violation of confidentiality; and the possibility of countersuit by the patient. It might also reflect an acting-out of counter-transference on the part of staff. Norko et al (Reference Norko, Zonana and Philips1992) also recommended guidelines for determining appropriateness of patient prosecution.

Since this survey our service has engaged in a closer working relationship with the local police liaison service. Staff are being made aware of the various legal criteria of what constitutes a crime, how and when to involve the police, and what the expected police action should be. There is an awareness that the police are often less able to deal with violent incidents in a learning disability setting than the unit staff themselves. We recommend use of the ‘appropriate adult’ system in criminal justice proceedings to ensure that fair and due process is adhered to.

Conclusion

The level of violent and threatening acts occurring within in-patient learning disability units is high. We feel that there is a need to acknowledge the scale of the problem and make appropriate resources available to deal with it.

We are currently participating in the College's audit of violence within learning disability settings. We hope to repeat the survey reported here with more positive findings when we act on lessons learned from these exercises.

Declaration of interest

None.