Most mental health nurses engage at some point with clients who harm themselves and these nurses often experience strong negative emotional reactions. Prolonged engagement with relapsing clients can lead to antipathy, and 'malignant alienation'. The study reported here has the aim of developing a brief, robust instrument for assessing nurse attitudes in this area. The Self-Harm Antipathy Scale, developed here on a sample of 153 healthcare professionals, has 30 attitudinal items with six factors. It has acceptable face validity, good internal consistency and some evidence of good test-retest reliability. It discriminates effectively between criterion groups. Overall this is evidence for the complexity of nurses' responses to this client group but such complex attitudes can still be assessed using a relatively brief structured instrument.

Nurses' attitudes toward service users who repeatedly self-harm can be negative and may interfere with the user's willingness to engage with services. The effectiveness of an educational intervention aimed at improving nurses' attitudes in this area was tested in this study. The intervention consisted of attendance on an accredited course on self-harm over a period of 15 weeks and the outcome of interest was attitudes as measured by the Self Harm Antipathy Scale. When deployed in a before-and-after design with two non-randomly allocated groups, there was evidence of a 20% reduction in antipathy toward self-harm among course attenders maintained over a period of at least 18 months (compared with a 9% reduction in a comparison group). Three of the six Self Harm Antipathy Scale attitude dimensions showed significant short-term change with some further long-term effects. This is preliminary evidence for the effectiveness of the chosen intervention in reducing overall antipathy toward self-harm clients and enhancing a sense of competence, a valuing of the care process and an awareness of the factors contributing to self-harm.

Just as Chicken Little did in warning that "he sky is falling '' the news media cornrnitfindamentd errors of ath-ibution in their coverage of risk situations, by treating them as rwvelties, failing to analyze the entire system, and using im@ciently analytical language.The sky has n e z w fallen and nezw tctll. Thozigh the sky has nezw fallen, it zcdl inezttably full. The likelihood of the skjj jalling increases etlery day. Our sky is beginning to shou~ eziidence of fulling. Someone sau' apiece oj-the sky full in u remote location. The sky fell this morning.

No abstract

In this study we seek to enhance the assessment of imminent violence risk by providing empirical data on the types of verbal and non-verbal behaviour exhibited by 31 psychiatric inpatients immediately prior to assaulting a staff member, and 31 non-aggressive controls. Verbal abuse, high overall activity level and standing uncomfortably close to the intended victim were the most common behaviours immediately prior to the assault, but most preassault behaviours were also exhibited when patients were not assaulting staff. In the 3 days prior to the assault, aggressive patients differed from non-aggressors in terms of verbal abuse, abnormal activity level (P < 0.05), threatening gestures and threatening stance (P < 0.01). Only one patient was aggressive in the absence of any predictive behaviours. We conclude that most patients exhibit easily identifiable signs of imminent aggression, but that many of these signs occur in the absence of aggression.