ABSTRACT
BACKGROUND: With suicide rising in the United States, identifying and preventing suicides is increasingly important. To provide a valuable step toward achieving effective suicide risk assessment, this study examines the agreement between self-report measures and psychiatrist documentation of suicidal ideation and behaviors (SI) at a Veterans Administration (VA) psychiatric emergency clinic. METHODS: A total of 377 veterans presenting at a VA psychiatric emergency clinic completed a self-report survey on SI and other acute risk factors for suicidal behavior. We examined agreement between veterans' self-reported SI and psychiatrists' clinical notes regarding SI. RESULTS: A total of 199 veterans (53%) self-reported SI; 80 psychiatrist notes (21%) indicated SI. Psychiatrists and veterans differed in 44% (164/377) of cases. Among the discordant cases, the veterans' self-report was more severe than the psychiatrists' in 97% of cases. Of the 120 veterans with SI and documented as having no SI by psychiatrists, 31 (26%) reported having a suicide plan and 18 (15%) plan preparations. Findings were similar when controlling for presenting problem, current depression, presence of a standardized suicide risk assessment, psychiatrist training level, past suicide attempt, homelessness, diagnosis of personality, or substance use disorder. CONCLUSIONS: Agreement between veterans' self-reports and psychiatrists' documentation of SI was generally low, with veterans self-reporting SI significantly more often than psychiatrists documented SI in their clinical notes. This suggests that inclusion of a self-report questionnaire provides an additional source of data to complement information gleaned from the clinical interview for a more comprehensive risk assessment, but only if actually examined by the clinician.