Louise Harriss1, Keith Hawton. 1. Centre for Suicide Research, Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford OX3 7JX, UK.
Abstract
BACKGROUND: Suicidal intent at the time of deliberate self-harm (DSH) is known to be associated with risk of future suicide in DSH patients. The predictive power of suicidal intent to identify particular individuals who are at high risk of suicide is unclear. METHOD: Clinical and demographic data on 4156 DSH patients presenting to hospital 1993 and 1997 were analysed. Follow-up information on eventual suicide was available for 2489 patients (mean follow-up 5.2 years). The predictive value of the Beck Suicidal Intent Scale (SIS) was examined using Receiver Operating Characteristic (ROC) plots. RESULTS: Thirty males (2.9%) and 24 females (1.7%) died by suicide. Despite a robust association between suicidal intent and eventual suicide, the positive predictive value (PPV) of the SIS was low (4%). Predicted probabilities of suicide assigned to individual patients were also low, even for those who eventually died by suicide. LIMITATIONS: Assessment and follow-up information was not available for all patients. CONCLUSIONS: The SIS cannot predict which individual patients will ultimately die by suicide. Nonetheless, information regarding suicidal intent is valuable in the clinical risk assessment and management of DSH patients.
BACKGROUND: Suicidal intent at the time of deliberate self-harm (DSH) is known to be associated with risk of future suicide in DSH patients. The predictive power of suicidal intent to identify particular individuals who are at high risk of suicide is unclear. METHOD: Clinical and demographic data on 4156 DSH patients presenting to hospital 1993 and 1997 were analysed. Follow-up information on eventual suicide was available for 2489 patients (mean follow-up 5.2 years). The predictive value of the Beck Suicidal Intent Scale (SIS) was examined using Receiver Operating Characteristic (ROC) plots. RESULTS: Thirty males (2.9%) and 24 females (1.7%) died by suicide. Despite a robust association between suicidal intent and eventual suicide, the positive predictive value (PPV) of the SIS was low (4%). Predicted probabilities of suicide assigned to individual patients were also low, even for those who eventually died by suicide. LIMITATIONS: Assessment and follow-up information was not available for all patients. CONCLUSIONS: The SIS cannot predict which individual patients will ultimately die by suicide. Nonetheless, information regarding suicidal intent is valuable in the clinical risk assessment and management of DSH patients.
Authors: Vincent C H Chen; Jen-Yu Chou; Ti-Chen Hsieh; Hung-Jung Chang; Charles T C Lee; Michael Dewey; Robert Stewart; Happy K L Tan Journal: Soc Psychiatry Psychiatr Epidemiol Date: 2013-04-06 Impact factor: 4.328
Authors: Shirley Yen; M Tracie Shea; Zach Walsh; Maria O Edelen; Christopher J Hopwood; John C Markowitz; Emily B Ansell; Leslie C Morey; Carlos M Grilo; Charles A Sanislow; Andrew E Skodol; John G Gunderson; Mary C Zanarini; Thomas H McGlashan Journal: J Clin Psychiatry Date: 2011-01-25 Impact factor: 4.384
Authors: E Antretter; D Dunkel; C Haring; P Corcoran; D De Leo; S Fekete; K Hawton; A J F M Kerkhof; J Lönnqvist; E Salander Renberg; A Schmidtke; K Van Heeringen; D Wasserman Journal: Int J Methods Psychiatr Res Date: 2008 Impact factor: 4.035
Authors: Matthew Miller; Katherine Hempstead; Tuan Nguyen; Catherine Barber; Sarah Rosenberg-Wohl; Deborah Azrael Journal: Am J Public Health Date: 2013-04-18 Impact factor: 9.308
Authors: Zimri Yaseen; Curren Katz; Matthew S Johnson; Daniel Eisenberg; Lisa J Cohen; Igor I Galynker Journal: BMC Psychiatry Date: 2010-12-14 Impact factor: 3.630