Improving Identification of Suicide Risk Via Routine Screening for Depression and Alcohol Use

Abstract

Background: Annual rates of suicide have been on the rise for the past two decades in the U.S. In 2017, there were 47,173 confirmed suicide deaths and an estimated 1.4 million suicide attempts. Alcohol use and suicidal ideation are both independently associated with increased risk of suicide attempt. Depression assessment tools that include questions about suicidal ideation can help providers identify patients at risk of suicide attempt in the clinical setting where many seek care in the months prior to an attempt. Using routine self-reported assessment tools for risky patterns of alcohol use may also be useful for optimizing identification of suicide attempt risk. Additionally, understanding more about the patient experience answering questions about suicidal ideation could inform screening and assessment practices during routine care delivery. Therefore, this dissertation studied patients from a large regional healthcare system—Kaiser Permanente Washington—to evaluate whether patterns of patient-reported alcohol use can be used to identify patients at risk of suicide attempt (Aim 1) whether and how that association varies by report of suicidal ideation (Aim 2), and to understand how patients experience and answer questions about suicidal ideation (Aim 3). Methods: Aims 1 and 2 used electronic health record data to identify outpatient visits to a mental health provider with documented assessments for unhealthy alcohol use (via Alcohol Use Disorders Identification Test-Consumption or AUDIT-C) and depression symptom severity (via 9-item Patient Heath Questionnaire or PHQ-9) in Kaiser Permanente Washington (1/1/2010-6/30/2015). Logistic regression models were fit using generalized estimating equations were fit to conduct visit-level analyses, accounting for correlation between individuals’ assessments. Separate models evaluated the association of (1) level of alcohol consumption and (2) frequency of heavy episodic drinking (HED) with suicide attempt within 90 days following each visit. Additional stratified analyses were used to evaluate these associations by self-reported suicidal ideation (PHQ-9 ninth question). Primary models adjusted for age, gender, race/ethnicity and visit year. Aim 3 used criterion sampling to identify primary care patients who had recently completed the PHQ-9 ninth question with whom we conducted semi-structured interviews by phone, which were recorded and transcribed. Directive and conventional content analysis were used to apply knowledge from prior research and elucidate new information from interviews; thematic analysis was used to organize key content. Results: Aim 1 primary analyses included 60,273 patient visits among 44,106 patients. At the first study period visit, 22% reported nondrinking, and 39%, 35%, and 5% reported low-, moderate-, and high-level drinking, respectively. Further, 65% reported no HED, 22% reported “less than monthly,” 7% “monthly,” 4% “weekly,” and 2% reported “daily/almost daily” HED. Patients reporting high-level alcohol use were 2.59 times (95% CI, 1.80-3.74) more likely to attempt suicide compared to those reporting low-level use. Patients reporting daily/almost HED were 3.62 times (95% CI, 1.80-3.74) as likely to attempt suicide than those reporting no HED. Aim 2 primary analyses included 59,705 patient visits among 43,706 patients. At the first study period visit 26% reported having suicidal ideation in the past two weeks. Among these patients, risk of suicide attempt was significantly increased for those reporting high-level (OR 9.77, 95% CI, 6.23-15.34), moderate-level (OR: 4.94, 95% CI 3.49-6.98) and non-drinking (OR 5.86, 95% CI 4.00-8.58), relative to low-level alcohol use. Risk was increased for those reporting HED monthly or more (OR 6.80, 95% CI 4.77-9.72) and less than monthly (OR 5.16, 95% CI, 3.67-7.26) relative to those reporting no HED. Among patients reporting no suicidal ideation, no associations between alcohol use (consumption or HED) and suicide attempt risk were observed. In Aim 3, qualitative analyses revealed: 1) Participants believed being asked about suicidality was contextually appropriate and valuable, 2) Some participants described a mismatch between their lived experience and the PHQ-9 ninth question, 3) Suicidality disclosures involved weighing hope for help against fears of negative consequences, and 4) Provider relationships and acts of listening and caring facilitated discussions about suicidality. Conclusions: In these studies, we found that while patients believed being asked questions about suicidal ideation was appropriate, some described experiencing stigma and distanced themselves from suicidality. Further, we found that patterns of alcohol use obtained via routine self-reported assessment may help identify patients at higher risk of suicide attempt, but largely only among patients disclosing suicidal ideation. Findings underscored the value of offering alcohol-related care to patients reporting risky alcohol use patterns, especially to those also reporting suicidal ideation. More work is needed to optimize identification of patients at high-risk of suicide attempt who do not report suicidal ideation.