The City-128 study of observation and outcomes

The City-128 study was designed to assess the contribution of patient, service environment, physical environment, containment usage, staff demography and staff attitudes to rates of self-harm and suicide by patients on acute psychiatric wards. Using a multivariate cross-sectional design, data was collected from more than the planned sample of 128 wards in three regions of England. Over 50,000 end of shift reports and more than 10,000 questionnaires were collected during the study. Multilevel modelling was used to analyze the data. Constant special observation was not related to rates of self-harm, but intermittent observation was associated with reduced self-harm. Ward security policies were not related to self-harm, but locking of the ward door was associated with more self-harm. Some other coercive measures were associated with more self-harm. Admissions during the shift and the types of patient admitted were also influential factors. Greater numbers of qualified nursing staff on duty were associated with reduced self-harm. Many staff factors thought to be important were not associated with self-harm rates. There were indications that a more liberal psychiatry with more emphasis on patient freedom and responsibility, and less use of containment (with one interesting exception) reduced rates of self-harm. Some significant changes to policy and practice may be required.