One of the key challenges in suicidality research is that outcomes may be multi-determined and rare. The causes of suicidal behaviors may consist of a combination of factors, which vary over time, and between individuals and groups. Risk factors may be distal or proximal, and distal exposures may increase risk for suicide given changes in proximal risk factor(s). Despite literature on suicidality among lesbian, gay, bisexual, and transgender (LGBT) persons, little research has examined LGBT Veterans. We hypothesize that LGBT status is a pre-disposing factor that increases risk due to minority stress, the stress of prior and potentially ongoing discrimination, including military sexual trauma, that may be further increased due to proximal factors such as comorbid mental health and medical conditions, a lack of engagement in care, system and providers lack of awareness or insensitivity to their unique needs, and pain among other factors. With the 2011 repeal of the Don't Ask, Don't Tell (DADT) policy, LGB personnel are currently able to openly serve in the US military, while the status of T personnel remains uncertain. While LGBT Veterans, especially those who served prior to repeal of DADT, may suffer similar discrimination, stigma and stress, they may each have dissimilar suicide risks. One study in the VA found that the prevalence of gender identity disorder diagnosis was >5 times higher than in the general US population, and that the rate of suicide-related events was > 20 times higher than the VA population. However, the engagement of LGBT Veterans in VA care is difficult to estimate: sexual orientation and gender identity (SO/GI) data is not collected in a structured manner. The VA's Evidence Review: Social Determinants of Health for Veterans found “only two articles examined sexual orientation for Veterans and non-Veterans and included only women. No articles addressed gender identity for Veterans and non-Veterans”. In addition to LGBT status, there is less study of other suicide risk factors such as pain and opioid prescriptions as well as potential protective factors such as receipt of complementary and integrative health services, which may treat both pain and PTSD. Little is known about how race/ethnicity, gender, morbidity, and treatment intersect with sexual orientation/gender identity to impact suicidality risk. Using VA electronic health record data, will identify LGBT Veterans to ascertain pathways to suicidal behavior, and assess both risk and protective factors related to LGBT individuals in all gender, age, and racial and ethnic groups. In a current project (IIR 16-262), our team is developing natural language processing (NLP) and machine learning (ML) tools to study CIH services' effect on opioid prescribing among Veterans with musculoskeletal disorder diagnoses (MSD) and PTSD. In another project (CRE 12-012), we identified 15,000+ Veterans with suicide attempts near the MSD diagnosis date, and 17,000+ suicide deaths in the cohort (N=5.2 million). That work informs this proposal. Because suicide attempts (SA) might not be captured if a Veteran presents at a non-VA hospital, we will integrate CMS data and develop NLP and ML tools to extract information from VA clinical notes as part of care. The objective and aims of this study are to: Aim 1: Develop informatics tools to identify LGBT Veterans and suicide attempts and events. Aim 2: Characterize suicide risk factors in a national sample of veterans. Aim 3: Examine the differential effect of risk factors on suicide by LGBT status. Achievement of these objectives and Specific Aims promise to expand our understanding of potential areas of unmet need and outreach to all individuals at risk for suicide. This area of VA health services research is in its infancy. Next steps will include: examinations of service(s) use after policy changes and VHA directives such as repeal of DADT (2010) and DOMA (2013), and VHA Directives 1340 and 2011-024 as noted in the letter of support from the VA's LGBT Health Program (10P4Y); using SHEP data to examine differences in satisfaction with care; and sorely needed data on differences in outcomes of care for LGBT Veterans.

This project will fill several critical gaps in the area of suicide research by developing methods to identify sexual minority status, suicide attempts and completed suicide using data from the Veterans Administration (VA), the nation's largest integrated healthcare system. We will examine whether there is variation in suicide risk and protective factors among LGBT and other Veterans in VA care, with emphasis on mental health and substance use disorders, pain, and use of complementary health approaches such as acupuncture, which may help to reduce suicidality. Results have important public health implications, and the methodology may be adapted to expand our understanding and promote outreach to all individuals at risk for suicide both within and outside VA.