To the Editor: We read with interest the article in the June issue entitled "Crowding and Aggression on Inpatient Psychiatric Wards," in which Nijman and Rector (1) report a modest correlation between the number of patients on two acute wards and the number of aggressive incidents. Crowding on acute psychiatric wards is a topical but underinvestigated issue. It is of concern to most mental health service providers given the global trend to reduce the number of inpatient beds on psychiatric wards.
As the authors acknowledge, most studies of violence on acute inpatient units have concentrated on individual patient variables. Only a few well-designed studies have specifically examined the relationship between violence and crowding on acute units. Although an association between crowding and aggressive incidents sounds plausible, it is not supported by many studies. Lanza and associates (2) found no relationship between the percentage of occupied beds on a unit and the number of assaults, and in a recent study Hardie (3) concluded that an increase in patient density was not associated with an increase in violent incidents. Therefore Nijman and Rector's study is useful.
However, because the authors did not report on the extent of patient use of the enlarged space created by opening a courtyard in one of the wards, it is difficult to draw any conclusions about the significance of the finding that the number of aggressive incidents did not decline after the courtyard was opened. If we accept that an association exists between crowding and violent behavior, then additional space must be utilized in order to reduce the number of aggressive incidents.
In our view, aggressive incidents on a crowded ward may occur because of stress. Stress has been suggested as an explanation for the observed association between violence and crowding in a number of animal and human studies (4,5). Psychiatric hospitalization in itself is stressful for most patients, and this stress can be exacerbated by crowding. Crowding creates intrusions into personal space, disrupts patient and staff activities, and leads to patient frustration about being forced to socialize in a confined space. These factors may act independently or together to aggravate the stress and lower frustration tolerance, leading to aggressive incidents.
The modest association Nijman and Rector found between aggressive incidents and crowding was identified by correlation analysis. The use of correlation is offered as an explanation of why a causal relationship between crowding and aggressive incidents could not be established. A weekly occupancy rate was used, and we assume a weekly incident rate was calculated. One problem with the use of correlation is the daily population changes. Regression analysis might have been more a more useful statistical approach because the outcome variables were dichotomous; that is, they indicated whether there was an incident or not. A regression model with crowding, ward, and time as explanatory variables, defined as before or after the ward was enlarged, might have been a more powerful analysis.
Dr. Kumar and Dr. Ng are affiliated with Rotorua Hospital in Rotorua, New Zealand. Ms. Robinson is with the department of community medicine in the Auckland School of Medicine in Auckland, New Zealand.