Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website
HSRD Conference Logo



2019 HSR&D/QUERI National Conference Abstract

Printable View

1069 — Leveraging VA electronic health record data to assess associations of social determinants of health with suicide risk among Veterans

Lead/Presenter: John Blosnich,  COIN - Pittsburgh/Philadelphia
All Authors: Blosnich JR (Center for Health Equity Research and Promotion, Pittsburgh; University of Pittsburgh), Montgomery AE (National Center on Homelessness Among Veterans, University of Alabama School of Public Health), Dichter ME (Center for Health Equity Research and Promotion, Philadelphia; University of Pennsylvania) Gordon AJ (Informatics, Decision-Enhancement and Analytic Sciences Center, University of Utah) Kavalieratos D (Center for Health Equity Research and Promotion, Pittsburgh; University of Pittsburgh) Taylor L (National Director for VA Social Work, Care Management, and Chaplaincy) Bossarte RM (VISN2 Center of Excellence for Suicide Prevention; West Virginia University)

Objectives:
Social determinants of health (SDH) are strong predictors of suicide risk, but most electronic health records (EHR) do not include SDH data. We determined the extent to which SDH were detectable in the Veterans Health Administration (VHA) EHR and how SDH were associated with suicide ideation and attempt.

Methods:
Data were from patients in the Veterans Integrated Service Network Region 4 with > 1 medical visits during 2016 (n = 293,872). SDH were operationalized using VHA coding for services and ICD-10 codes, encompassing 6 categories: violence, housing instability, financial/employment problems, legal problems, familial/social problems, and lack of access to care/transportation. We defined suicide morbidity by ICD-10 codes and data from the Suicide Prevention Application Network (SPAN). Logistic regression assessed associations of SDH with suicide morbidity, adjusting for socio-demographics and medical comorbidities.

Results:
Most patients were white (79.7%), male (91.7%), and ? 60 years old (70.9%). One in 10 patients had 1 SDH and 5.6% of patients had at least 2 SDH. After adjusting for medical comorbidity and socio-demographic factors, there were significant dose response-like relations between SDH and suicidal ideation and suicide attempt. For instance, compared to patients with no SDH, the odds of suicide attempt for patients with 1 SDH was 4.9 (95% confidence interval[CI]:3.3-7.2), with 2 SDH was 7.4 (95%CI: 4.6-11.8), with 3 SDH was 10.3 (95%CI: 6.1-17.6), with 4 or more SDH was 19.2 (95%CI: 11.8 - 31.1).

Implications:
To our knowledge, this is one of the largest epidemiologic studies of SDH indicators using EHR from the United States' largest integrated healthcare system. SDH were the variables most strongly associated with suicide ideation and suicide attempt; stronger than medical comorbidity, which was remarkable because comorbidities included conditions associated with suicide morbidity (e.g., substance use disorders, depression).

Impacts:
Suicide prevention is one of the VA Secretary's top priorities. Although over 80% of suicide decedents had a medical visit in the year preceding their deaths, healthcare systems struggle to identify risk factors for suicide. Complementing medical factors with standardized integration of SDH data in healthcare systems could improve suicide prevention and intervention.