BACKGROUND Seclusion is an intervention widely used in Dutch mental health care. The intervention can be effective in acute situations to avert (further) aggression or self-harm. However, seclusion is also a controversial intervention that may not have any positive effect with regard to symptom improvement. In general patients report negative effects after being secluded e.g. anxiety and having had a traumatic experience.
The main reason for seclusion is not manageable aggressive behaviour of a patient. Earlier studies reported several risk factors that may contribute to seclusion, regarding patients’ characteristics, but also with regard to staff characteristics, working protocols and unit characteristics. Because of unequivocally results there is the need for a longitudinal prospective study to examine staff- and unit determinants in association with seclusion.

AIMS The objective of this study is to determine which nursing staff and unit characteristics are associated with seclusion following aggression in hospitalized adult psychiatric patients. We hope to create a predictive model to estimate the risk of seclusion on an acute psychiatric ward.

METHODS We will conduct a prospective observational study on a closed psychiatric ward of an academic hospital. Patients are aged 18 – 65 years and are admitted when their psychiatric condition leads to an immediate threat to the patient themselves or their surroundings.
All nurses on the ward are all qualified nurses and registered in the Dutch registration of healthcare professionals. They are trained every six months in techniques of verbal de-escalation and safe physical restraint. For both nurses and the patients baseline characteristics are monitored. Every shift (day, evening, night) data are gathered on the patients, nurses and unit. Data are retrieved from the electronic patient chart, including information of the Brøset Violence Checklist. Furthermore, the exchange of information among nurses is measured using the Grid instrument. Data will be analysed using multilevel regression analysis. Data will be collected for a period of 2 years, which started January 2013.

RESULTS The primary endpoint in our study is the incidence of seclusion. As a secondary endpoint, the duration of the seclusion is measured. These endpoints are measured using the Argus registration system and will be linked to predictors of seclusion, with special focus on the nursing staff- and unit determinants.