The National Audit of Violence: in-patient care for adults of working age

AIMS AND METHOD 

We audited 184 psychiatric wards against clinical practice guidelines for the management of violence. Staff and service users completed anonymous questionnaires. Environmental inspections were performed by two teams.

RESULTS 

There were 4460 questionnaires returned. Nurses (78%) were significantly more likely to report the experience of violence than service users (37%). Drugs were reported by 72% of nurses and alcohol by 61% as causing problems. Other standards frequently not met included staffing levels, training, provision of activities, ward design and ambience.

CLINICAL IMPLICATIONS 

Specific issues are identified that must be addressed by national and local action. A baseline is set against which the impact of this action can be judged. Priorities must include tackling drug and alcohol use in psychiatric wards.

The National Audit of Violence: in-patient care for adults of working age

AIMS AND METHOD
We audited 184 psychiatric wards against clinical practice guidelines for the management of violence. Staff and service users completed anonymous questionnaires. Environmental inspections were performed by two teams.

RESULTS
There were 4460 questionnaires returned. Nurses (78%) were significantly more likely to report the experience of violence than service users (37%). Drugs were reported by 72% of nurses and alcohol by 61% as causing problems. Other standards frequently not met included staffing levels, training, provision of activities, ward design and ambience.

CLINICAL IMPLICATIONS
Specific issues are identified that must be addressed by national and local action. A baseline is set against which the impact of this action can be judged. Priorities must include tackling drug and alcohol use in psychiatric wards.
The protection of staff and service users against violence in healthcare settings is a public health priority. In England, this is reflected by initiatives such as the Department of Health Zero Tolerance Policy (Department of Health, 2000) and, in mental health specifically, by the prioritisation of ward safety by the National Patient Safety Agency (Lelliott, 2004). A range of factors contribute to the risk of violence in psychiatric wards. These include poor ward design and physical environment, low staff numbers and inadequate training, negative staff attitudes, the increasing prevalence of service users at high risk of violence on wards, boredom and frustration among service users and the use of alcohol and illegal drugs (Royal College of Psychiatrists, 1998). Violence has been shown to occur frequently against staff, including psychiatric trainees (Pieters et al, 2005) and can result in staff experiencing significant psychiatric morbidity (Wildgoose et al, 2003).
The National Audit of Violence in mental health in-patient settings was funded by the Healthcare Commission. It repeats an earlier audit conducted on a smaller scale (McGeorge et al, 2001) that showed high rates of violence against service users, and highlighted problems with staff training, drug and alcohol use and ward environments. The audit standards, drawn from clinical practice guidelines published by the Royal College of Psychiatrists (1998), were revised to incorporate new recommendations by the group developing the National Institute for Clinical Excellence (NICE) guidelines (NICE, 2005). Data collection for the audit began in Spring 2004. This paper presents the main findings of the National Audit of Violence for wards for adults of working age (Healthcare Commission, 2005).

Method
A full description of the method is given in the final report of the audit (Healthcare Commission, 2005). Forty-one National Health Service (NHS) trusts in England enrolled voluntarily in the audit and data were collected for 184 psychiatric wards for adults of working age. This report includes those identified by the ward staff as acute wards, psychiatric intensive care units, forensic wards and rehabilitation or continuing care wards. Each trust set up a project team that attended a regional training event. These took place prior to the data collection and at the end of the audit. The audit had two components.

Questionnaires
A questionnaire for service users and one for staff were used to determine the factors linked to in-patient safety and violence. These were completed between April and August 2004 and returned anonymously to the College Research and Training Unit. Staff were categorised as nurses, or other clinical (doctors, psychologists, pharmacists, etc.) or non-clinical (e.g. administrators, maintenance staff, porters) staff. Each questionnaire contained a mixture of closed ('Yes/No') questions and boxes for free-text comments. Each local project team was encouraged to devise its own strategy for targeting staff and service users in order to maximise the response but preserve confidentiality. It was therefore not possible to calculate the refusal rate.

Inspections
Two teams, one of staff from the ward concerned and the other of people who did not work on that ward (trust managers, service user advocates, etc.), inspected and rated each ward independently against a set of evidencebased standards relating to the safety of the physical environment. This was carried out between September and October 2004 and ended with a meeting of the two teams to agree the final ratings.
In addition, information was collected about staffing, including the use of agency and bank staff in the week leading up to the audit. The data collection was supplemented by information about local ward conditions and obstacles to improvement gathered at regional events at which staff from participating wards met to consider the audit methods and findings.

Potential triggers of violence
The audit questionnaires enquired about a range of factors that are linked to violence. Of these, staff appeared to be particularly concerned about drug and alcohol use. Drugs were reported to cause trouble by 72% of nurses and alcohol by 61% (Table 1). Service users were less likely than nurses to report drugs (29%; w 2 =629, d.f.=1, P50.001) or alcohol (25%; w 2 =434, d.f.=1, P50.001) to be a problem. However, many service users did report being bored. Although 63% were satisfied with daytime activities and therapy, only 47% expressed satisfaction with evening activities and 41% with activities during the weekend.
Free-text comments about triggers to violence were made by 185 staff and 170 service users. For both groups, illegal drugs and alcohol were mentioned most frequently (by 54 staff and 27 service users). Other issues for staff included inadequate staff numbers (n=34) or training or experience (n=25) and overcrowding (n=10). Some service users (n=19) reported that staff inadvertently provoke violence by their negative attitudes or by restricting patients' freedom (n=15). Consistent with this, 36% of service users answered yes to the question 'do staff ever wind you up?'

Staffing and the management of violence
Most service users had a high opinion of staff and reported that they were available to speak to (83%) and treated them with respect (85%). Most (86%) also agreed that staff dealt effectively with violence between service users, an opinion that was shared by 94% of nurses. Nurses rated highly their support from other staff (overall rate 86%) and satisfaction with communication with colleagues (79%). However, only 57% of nurses were satisfied with the number, skills experience and qualifications of the staff team. Free-text responses reported problems such as inadequate staffing, inexperienced leadership, difficulties with recruiting nurses and an overreliance on bank and agency staff. Ward managers estimated that agency and bank staff had worked an average of 100 h on the ward in the week before the audit. This is equivalent to a mean of 2.7 full-time members of staff per ward. Although 90% of nurses had received some training in the prevention or management of violence in the past 5 years, 39% had had no training before they started working on the ward. Of those who had received training, 20% reported that it was inadequate to equip them to manage violence.

The physical environment
Nurses were generally more critical of the physical environment and ward ambience than service users (Table 1). Several safety issues were highlighted by the environmental audit. Only 36% of wards were judged to have adequate sight lines, 48% had exits that could be seen by staff and 46% had adequate private space. Although 80% of wards had access to outside activity areas, in only 40% was there covered external space, and a separate, low-stimulus quiet area was provided in only 59% of wards. Adequate temperature and ventilation control was judged to be present in only 33% of wards.

Discussion
The strength of the audit is that it involved many wards managed by a large number of NHS trusts. It also gathered the opinions of both staff and service users and used two separate data collection methods -questionnaires and inspections. However, the method was designed to support a national audit and not for health services research. The trusts and wards were volunteers and not selected to be representative of the national picture. Also, we could not calculate the response rate to the questionnaires and so cannot comment on the possible extent of response bias.
The audit highlights that, although nurses are most at risk, the experience of violence is not limited to those who work exclusively on the wards or whose job it is to intervene in violent incidents. Medical and ancillary staff are also affected. The causes of violence are complex and service users emphasised different factors from nurses.
The key messages arising from this audit are: . in-patient wards are frequently noisy, have unsafe physical environments and poor ambience . service users commonly lack a structured day, which results in boredom . staff are concerned about the impact of drug and alcohol use . many wards have inadequate staffing levels and rely upon temporary staff . many nurses report deficiencies with their training, particularly in applying it to real-life situations . despite these problems, service users generally have positive opinions about nurses and nurses feel well supported by colleagues.
Some of the problems identified by the audit require organisational or system changes. Difficulties in staff recruitment and low morale may arise as a result of the experience of violence or lead to increased violence. However, this audit does not support the latter, as both service users and nurses gave consistently high ratings for the way staff managed violence. The status of in-patient nursing must be raised to reduce the exodus of nurses to community posts and so reduce reliance on bank and agency nurses. Only if staff duties are reorganised can nurses spend more time in face-to-face contact with service users. This would both increase therapeutic and occupational activities and reduce boredom among service users, and probably improve staff morale. The open nature of acute psychiatric wards and rapid patient turnover make it difficult to prevent drugs and alcohol getting onto the ward (Quirk et al, 2006). Creative solutions are needed to limit access to drugs and alcohol without compromising patient freedom and choice.
Although the audit standards were chosen because of their link to ward safety, many are also measures of ward quality. The audit therefore sets a baseline against which the impact of national and local action to improve English psychiatric wards can be gauged. This action is backed by »30 million of additional funding (Department of Health, 2004) and a raft of guidance about many aspects of ward design, ward safety and the management of violence (Department of Health, 2002;Marshall et al, 2004;NICE, 2005).

Declaration of interest
None.