The Clinical Utility of PHQ-9 Item 9 for Suicide Prediction: In Reply
IN REPLY: We appreciate Dr. McCusker’s helpful letter. We agree that, as a function of the low base rate for suicide, the PPV for a given patient is limited. However, patient responses to this question offer clinical utility because they provide a distinct indicator of risk of suicidal behaviors.
Our study demonstrated that item 9 of the PHQ-9 was associated with an increased risk of suicide in the study population, even after the analysis controlled for overall depression. As we noted in the discussion section, suicidal ideation as measured by this tool is limited as a predictor of individual patient-level suicide. The Joint Commission’s recent Sentinel Event Alert emphasized the importance of patient assessments to inform care and to enhance suicide prevention for at-risk patient subpopulations (1). The Joint Commission recommends screening all patients for suicidal ideation by using a brief, standardized, evidence-based screening tool, such as the PHQ-9, in nonacute or acute care settings.
Our study documented positive associations between item 9 of the PHQ-9 and suicide risk among patients in the Department of Veterans Affairs (VA) health system. Although we acknowledge the limited PPV for individual patients, we believe assessments provide valuable information to inform clinical care and enhance suicide prevention. We note that VA is working to enhance and apply more comprehensive predictive models for suicide (2); these will be based not only on the information available in the electronic health record but also on real-time clinical assessments. Application of predictive modeling tools offers new strategies for health systems to prevent suicide.
1 Detecting and treating suicide ideation in all settings. Sentinel Event Alert, Feb 24, 2016. https://www.jointcommission.org/assets/1/18/SEA_56_Suicide.pdf.Google Scholar
2 : Predictive modeling and concentration of the risk of suicide: Implications for preventive interventions in the US Department of Veterans Affairs. American Journal of Public Health 105:1935–1942, 2015Crossref, Medline, Google Scholar