Trainees'forum The effect environmental change has on the frequency of violent incidents

Violence is an important consideration in the management of psychiatric patients. Probably one in ten psychiatric patients assault staff (DHSS, 1988), although this may be an underestimate (Lion et al, 1981). Fellow patients are assaulted less frequently (Nobel & Rodger, 1989). Several factors have been associated with violent behaviour in patients: suffer ing from schizophrenia, delusions, hallucinations (Noble & Rodger, 1989);being young, compulsorily detained and an increased use of temporary nursing staff (JAimese/al, 1990). This survey compares the frequency of violent incidents before and after transfer from a rehabili tation ward to a health authority house or hostel.

Violence is an important consideration in the management of psychiatric patients. Probably one in ten psychiatric patients assault staff (DHSS, 1988), although this may be an underestimate (Lion et al, 1981). Fellow patients are assaulted less frequently (Nobel & Rodger, 1989). Several factors have been associated with violent behaviour in patients: suffer ing from schizophrenia, delusions, hallucinations (Noble & Rodger, 1989); being young, compulsorily detained and an increased use of temporary nursing staff (JÃ¡mese/al, 1990).
This survey compares the frequency of violent incidents before and after transfer from a rehabili tation ward to a health authority house or hostel.

The study
In order to improve the psychiatric rehabilitation service in Doncaster, staff accommodation within the hospital grounds was converted into two hostels and eight houses for patients' occupancy.
As accommodation became available patients were transferred from the rehabilitation wards. All patients transferred since 1987 were included in the survey, provided they were in-patients on the ward for 12 months prior to transfer and have remained resident in the hostel/house to date.
The hostels and houses are occupied by five and three patients respectively. All patients have a single bedroom; the remaining accommodation is for communal use. Psychiatric nurses staff the hostels 24 hours a day, the houses 14 hours a day. The staff/ patient ratio varies: in the hostels a total of 15 staff provide 24 hour cover for ten patients; in the houses a total of five staff provide 14hour cover for 21 patients. The wards from which patients were transferred were two 28-bedded rehabilitation wards, each with 20 single and two 4-bedded bedrooms off a main corridor. In addition they had two sitting rooms.
Records are kept of all violent incidents on a 'violent incident' form, regarding the patient's name, the date, time, nature of the incident and the action taken by staff. These forms were examined, in order to quantify the incidence of violent incidents in the 12 months before and after transfer. The significance of the change was measured using the Wilcoxon signed rank test.

Findings
Twenty-four patients fulfilled the above criteria and were therefore included in the survey. The total number of violent incidents initiated by these patients reduced from 47 in the 12 months before transfer to eight in the 12 months after transfer (P<0.0\). There was no documented evidence of any violent behav iour, before or after transfer, for eight patients. The reduction in violent incidents was therefore accounted for by the remaining 16 patients. Further analysis of the data in the 12months before transfer gives a mean number of violent incidents per patient of 1.96 (range 0 to 14, SD 3.46 and a variance of 12.0). Analysis of the data in the 12 months after transfer gives a mean number of violent incidents per patient of 0.33 (range 0 to 2, SD 0.70 and a variance of 0.49).

Comment
It is suggested that the reduction in 'violent incidents', following transfer to a staffed house/hostel was pri marily due to the change of environment, as other variables were kept to a minimum. Staff changes were negligible as ward based staff were transferred with the patients, allowing continuity of care and minimising differences in staffs perception of criteria for complet ing 'violent incident' forms. The influence of patients' episodes of violence diminishing as their illness improved, was reduced by only including long stay patients, none of whom have yet progressed to a less supportive environment. The frequency of 'violent incidents' reported for individual patients, during the 12 months prior to transfer, implies that episodes of violence had not waned before transfer.
A number of influencing factors are important in explaining the findings. Ward based patients have to abide by restrictions which are imposed for their own welfare and safety. A small number of patients living in a house allows a more 'normalised' environment 490 to develop. The environmental change, attained by the transfer to house/hostel accommodation, is conducive with encouraging patients to establish a greater degree of autonomy. Patient/patient and staff/patient relationships improve as individuals become more aware of each other's needs. Close working relationships and regular feedback from house meetings support this development. The staff/ patient ratio varies depending on individual needs, allowing time to be spent with patients when they need it, rather than when the ward situation permits. The number of patients living in close proximity is considerably less, reducing the potential for inter personal conflicts. On occasions, patients on re habilitation wards exhibit disturbed behaviour, which may precipitate a volatile atmosphere. In response, particularly vulnerable patients exhibit violent behaviour. It is hypothesised that the reduced number of patients residing in the houses/hostels has substantially decreased the frequency of this occurrence, benefiting susceptible patients.
In this survey, the measure used for improvement (i.e. frequency of violent incidents) was invalid for non-violent patients, although it demonstrates that these patients did not exhibit violent behaviour as a result of transfer. Patients who displayed a high incidence of minor violence, while resident on a rehabilitation ward, demonstrated a significant Rusius reduction in such behaviour, following transfer to a health authority house or hostel. It is therefore suggested that such behaviour should not be an argument against, but should reinforce the argument "for" transfer from a ward to the more "nor malised" environment of an adequately staffed house or hostel.

A cknowledgemen ts
I would like to thank Gillian Staley, Dr Maureen O'Leary, Consultant Psychiatrist, and Jan Hughes, Clinical Psychologist, for their help in the preparation of this paper.
The Audit Commission has drawn attention to local champions of change in mental health services. Good Practices in Mental Health (GPMH) ( 1985)has high lighted a district which has overcome some of the myths about the impossibility of transforming the service, and recently the Institute of Health Services Management (IHSM) Working Group (1991) has entered the debate with "good psychiatric services can be developed in areas where managers are deter mined to introduce improved services". The Audit Commission singled out Torbay Health Authority, GPMH highlighted Exeter Health Authority, and the IHSM Working Group have listed 12exemplary health authorities (including Torbay and Exeter) where good local services have been developed.
The common theme running through each report is that much more can be done to redeploy existing resources within mental health services rather than simply asking for more money.
My view is that this is too simplistic an approach to apply globally, given the considerable variance in funding levels and site asset values available to each unit. The picture is further complicated by the relative priority given to mental health services by regional and district authorities which may (or may not) make capital or bridging finance available for service development.
The IHSM report makes the fundamental point that "considerable resources are currently available within the Health Service and more can be achieved