2023 HSR&D/QUERI National Conference

1160 — Non- or partial disclosure of suicidal thoughts by Veterans during population-based screening in Veterans Health Administration: National survey study

Lead/Presenter: Steven Dobscha,  COIN - Portland
All Authors: Dobscha SK (HSRD Center to Improve Veteran Involvement in Care (CIVIC)), Elliott VJ (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Niederhausen M (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Salvi A (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Rynerson AL (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Gade PR (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Carlson KF (HSRD Center to Improve Veteran Involvement in Care (CIVIC)) Bahraini N (VA Rocky Mountain Mental Illness Research, Education, and Clinical Center) Post EP (Department of Medicine, University of Michigan Medical School, Ann Arbor) Denneson LM (HSRD Center to Improve Veteran Involvement in Care (CIVIC))

Prior research suggests that disclosure of suicidal thoughts, including in healthcare settings, may be low. In late 2018, the Veterans Health Administration (VHA) implemented a new population-based suicide risk screening program (Risk ID) in ambulatory settings. The main objectives of the current study were to: 1) describe the frequency of non-disclosure or partial-disclosure of suicidal ideation among a nationally-representative sample of Veterans screened in VHA primary care; and 2) examine potential barriers to disclosure in this patient population.

Between February and October 2021, a national sample of 2,001 Veterans screened in primary care were mailed surveys; sampling was stratified by initial screen (Columbia Suicide Severity Rating Scale Screener; C-SSRS) result and sex (as reported in VHA’s Corporate Data Warehouse). Initial mailings were sent within two weeks of screening, and up to three reminder mailings were sent to non-respondents. Survey respondents were compensated $35 for participating. In analyses, data were weighted to account for stratified sampling and survey non-response. T-tests and chi-squared tests were used to compare survey responses of C-SSRS-positive to C-SSRS-negative screen individuals. Analyses are underway to describe demographic, attitudinal and clinical predictors of non-disclosure, which we hope to report on at the national meeting.

Eight-hundred sixty-eight veterans (43.4%) returned surveys, and the average age of respondents was 61 (SD = 16), the majority were male (58%), white (68%) and non-Hispanic (91%). Ninety-seven percent and 85% of individuals with positive and negative C-SSRS screens, respectively, recalled being screened. Sixty percent (weighted) of Veterans with positive screens reported having expressed suicidal thoughts to their providers “very accurately” (as opposed to “somewhat” or “not at all accurately) on the date of screening. Similarly, 52% of Veterans with positive screens reported “very accurately” expressing suicidal thoughts to the nurse or medical assistant. A small proportion of Veterans with negative screens reported “not at all accurately” expressing their thoughts to nurses/medical assistants (6.3%) or providers (5.7%). Compared to individuals with negative screens, individuals with positive screens were more likely to endorse “everyone has the right to commit suicide” (39% vs. 14%; p < .001), less likely to endorse “no matter how sad, isolated and lonely people are, they should not consider suicide as a way out” (42% vs. 87%; p < .001), and more likely to report barriers to care across five subscales of the Barriers to Health Seeking Scale [Ouimette 2011] (all p < .015).

While population-based risk screening has potential to identify individuals at elevated risk for suicide, individuals at risk often experience potential barriers to care and help-seeking, and may under-disclose their levels of risk. Some at-risk individuals may go undetected.

A comprehensive approach to assessing suicide risk that is not overly reliant on screening is recommended to optimize identification of at-risk individuals and to stratify risk among those with positive screens. Further efforts to identify characteristics of individuals at risk for non- or under-disclosure, and to develop clinical approaches to engage these individuals, are indicated.