Citation

Murphy E, Kapur N, Webb R, Cooper J. J. Adv. Nurs. 2011; 67(1): 127-139.

Affiliation

Elizabeth Murphy BSc Research Associate Centre for Suicide Prevention, Manchester Academic Health Science Centre, The University of Manchester, UK Navneet Kapur MD FRCPsych Professor of Psychiatry and Population Health Centre for Suicide Prevention, Manchester Academic Health Science Centre, The University of Manchester, UK Roger Webb MSc PhD Senior Research Fellow Centre for Suicide Prevention, Manchester Academic Health Science Centre, The University of Manchester, UK Jayne Cooper PhD RMN Senior Research Fellow Centre for Suicide Prevention, Manchester Academic Health Science Centre, The University of Manchester, UK.

Copyright

(Copyright © 2011, John Wiley and Sons)

DOI

10.1111/j.1365-2648.2010.05484.x

PMID

20969616

Abstract

Aim.  This paper is a report of a study conducted to compare risk assessments by psychiatrists and mental health nurses following an episode of self-harm. Background.  Self-harm assessments by nurses and psychiatrists are similar in terms of overall content, but risk assessment may vary by professional discipline. To our knowledge previous researchers have not compared the positive predictive value of risk assessments by nurses and psychiatrists, the factors that inform those assessments in clinical practice or the management of people assessed as being at high risk. Methods.  We conducted a prospective cohort study (2002-2006) of 3491 individuals presenting with self-harm to three hospitals in the North West of England. A standard assessment form including detailed demographic and clinical data was completed by the assessing psychiatrist or nurse. Results.  The positive predictive value of risk assessments for self-harm repetition was 25% (95% CI: 20-31) among nurses and 23% (95% CI: 13-37) among psychiatrists. There was strong agreement on factors associated with high risk assessment by both professions. Following assessment of high risk, psychiatrists were much more likely than nurses to admit people for inpatient treatment (RR = 5·6, 95% CI: 3·2-9·7). This difference remained highly statistically significant after controlling for case-mix differences (RR = 4·3, 95% CI: 2·4-7·7). Conclusion.  Our finding that risk assessments were comparable by profession supports the provision of nurse-led assessment services. However, inpatient admission was influenced largely by assessor type rather than patient characteristics. This has important implications for equity of care and may reflect professional differences in referral practices.

Language: en