Continuing care old age psychiatry units often have high levels of behavioral disturbance. Shah (Reference Shah1999) reported that levels of aggressive behavior can be influenced by several factors including patient variables, patterns of hospital use and the nature of the ward environment. The ward environment includes variables such as staffing levels, staff attitudes and perceptions, staff morale and the character of staff-patient interaction. The relationship between aggressive behavior and the nature of the ward environment has been less well studied. We carried out a study to examine the relationship between aggressive behavior by patients, nursing staff attitudes and perception and stress amongst nurses. We report findings from our study on the relationship between aggressive behavior by patients on old age psychiatry wards, nursing staff attitudes and perceptions.

Ethical approval for this study was granted by the local Ealing Research Ethics Committee. The study was carried out over a three-month period, on two 16-bed old age psychiatry wards on the John Conolly Wing, St Bernard's Hospital, serving a catchment area of 40,000 older people. All patients were included in the study with both wards often at full occupancy. The patients studied were all over the age of 65 and had a functional or organic mental disorder.

One senior nurse on each ward completed the RAGE scale (Patel and Hope, Reference Patel and Hope1992) to measure aggressive behavior on all patients at monthly intervals. The RAGE scale is a 21-item scale that provides a composite measure of the quantity and severity of aggressive behavior. Although the original RAGE scale was designed to measure aggressive behavior over the preceding three days, it has since been adapted for use over the preceding week because the original reliability study reported little change over a week. This has been confirmed by a previous study on continuing care psychogeriatric wards (Shah and De, Reference Shah and De1998).

The attitude and perception of nursing staff toward patients was measured using the MAS (Malignant Alienation Scale). The original alienation scale (Morgan and Stanton, Reference Morgan and Stanton1997) had been developed for use among younger psychiatric patients to measure the phenomenon of malignant alienation. The MAS is an 8-item scale measuring the following items: (1) feeling distant from the patient, (2) deliberate use of symptoms by the patient, (3) alienation of the patient, (4) vulnerability of the patient, (5) withdrawn/uncommunicativeness, (6) variable mood, (7) depression, and (8) aggression. The MAS has previously been used with old age psychiatry patients (Baheerathan et al., Reference Bahareethan and Shah2000). In our study, the MAS was completed at monthly intervals by a senior nurse on each ward. Information used for completion of the MAS included direct nursing observations, nurse/patient interactions, nursing handovers and nursing notes and informal nursing discussions over the preceding month. Training was given to nurses at the start of the study by one of the researchers (IN) and also throughout the duration of the study.

Spearman's rank correlation coefficient was used to examine the correlation between the RAGE scale and the MAS. The correlations between the RAGE and MAS scores were similar for each month so we combined the data for all three months.

Ratings from 97 patients were collected and analyzed. Data were missing from seven patients in the final month and so these were excluded. We correlated the RAGE scores with the MAS scores and the MAS subscales for period 1 and then all 3 periods.

Our study showed a correlation between the RAGE total score and the MAS total score (ρ = 0.4, p < 0.0001). The total RAGE score showed a positive correlation with five out of the eight MAS sub-items. These sub-items were: feeling distant from the patient (ρ = 0.48), alienation of the patient (ρ = 0.32), withdrawn/uncommunicativeness (ρ =0.216), variable mood (ρ = 0.24) and aggression (ρ = 0.62).

Although the MAS was developed for use on adult psychiatry wards for patients with suicidal behavior, it has since been used on a continuing care psychogeriatric ward showing modest internal consistency and inter-rater reliability (Baheerathan and Shah, Reference Bahareethan and Shah2000). It is possible that some behavior may have been missed because our MAS ratings were done on a monthly basis. We believe that the risk of this occurring was reduced by including information from nursing handovers and by using an appropriately trained and experienced senior nurse to complete the ratings.

Our study showed a positive correlation between the total RAGE score and the following sub-items of the MAS: aggression, feeling distant from the patient, variable mood, alienation of the patient and withdrawn/uncommunicativeness. These findings are consistent with the findings of a similar study in London (Baheerathan and Shah, Reference Bahareethan and Shah2000). We carried out a cross-sectional study that demonstrated evidence of an association between aggression by patients and alienation by nurses. We were unable, however, to establish whether high levels of aggression by patients resulted in alienation by nurses or whether alienation by nurses in turn led to high levels of aggression by patients. It could be argued, however, that irrespective of the direction of causality, nurses should be given training in therapeutic interventions to reduce aggressive behavior. The expectation is that this would lead to an improvement in nursing attitudes and perceptions, thus reducing feelings of alienation. Alternatively, training could focus on improving nursing staff attitudes and their perceptions of patients in order to reduce levels of aggressive behavior by patients. The most effective approach to reducing aggressive behavior is likely to be one that combines interventions aimed at reducing aggressive behavior and improving nursing attitudes and perceptions. Further research will be required to examine the most effective approach to delivering such training.