Date of Submission

2015

Degree Type

Dissertation

Degree Name

Doctor of Psychology (PsyD)

Department

Psychology

Department Chair

Robert A DiTomasso, PhD, ABPP, Chair, Department of Psychology

First Advisor

Bradley Rosenfeld, PsyD, Chairperson

Second Advisor

Bruce S Zahn, EdD, ABPP

Third Advisor

Gregory K Brown, PhD

Abstract

The ability to comprehensively and effectively identify those individuals who are at greatest risk to engage in self-directed violence (SDV) forms the cornerstone for all professional suicide prevention activities. To that end, mental health professionals have come to rely on the use of risk stratification to identify at-risk individuals as a way to inform and guide risk management and treatment, without having the benefit of empirical evidence to support such practices. The current program evaluation examined archival data comprised of suicide risk assessments conducted by mental health professionals on suicidal veterans (N = 1,560) in the emergency department of a large, urban Veterans Administration medical center (VAMC) located in the northeastern United States over 2 years, as well as data for subsequent suicide attempts among this sample (n = 110). Results indicate that the current practice of stratifying suicide risk into specific categories (high, moderate, and low) effectively distinguishes the majority of those veterans who subsequently engaged in SDV from those who did not after being assessed by emergency department evaluators utilizing routine, naturalistic clinical judgment. Additional findings revealed that the categorical risk ratings of high, moderate, and low were distinguishable from one another, whereby a rating of high risk was more likely than both moderate and low risk to distinguish those suicidal veterans who subsequently engaged in SDV. Moderate risk identified at-risk veterans more accurately than those at low risk.