Suicide screening protocols implemented in veteran and general populations reduced suicides, but making such screening tools more efficient remains a challenge, according to a pair of studies.
After the Veterans Health Administration implemented the Veterans Affairs Suicide Risk Identification Strategy (Risk ID) in 2018, results were positive in outpatient visits for 3.5% and 0.4% of primary and secondary screenings, respectively, and results were positive in emergency settings for 3.6% and 2.1% of primary and secondary screenings, respectively, reported Nazanin Bahraini, PhD, Rocky Mountain Regional VA Medical Center in Aurora, Colorado, and colleagues.
Compared to patients in ambulatory care, patients screened in the emergency department were more likely to endorse suicidal ideation with intent (odds ratio 4.55, 95% CI 4.37-4.74, P<0.001), have a specific plan (OR 3.16, 95% CI 3.04-3.29, P<0.001), and report recent suicidal behavior (OR 1.95, 95% CI 1.87-2.03, P<0.001) during secondary screening, Bahraini’s group wrote in JAMA Network Open.
Risk ID includes three tiers of screening starting with the nine-item Patient Health Questionnaire, followed by the Columbia Suicide Severity Rating Scale Screener and the VHA’s Comprehensive Suicide Risk Evaluation.
“The greater acuity of suicide risk among patients in the ED or [urgent care clinics] cohort compared with those in the [ambulatory care] cohort supports national implementation of evidence-based suicide prevention programs, such as Safety Planning in the ED,” Bahraini and co-authors wrote.
Suicide screening is recommended for all veterans, who have a 21% higher risk of dying by suicide than the general population.
However, close to two-thirds of veterans who die by suicide do not seek healthcare from the VA, indicating that many suicides will still be missed in VA screening initiatives, noted Roy H. Perlis, MD, MSc, of Massachusetts General Hospital and Harvard Medical School in Boston, and Stephan D. Fihn, MD, MPH, of the University of Washington in Seattle, in an accompanying editorial.
“Although the VA is not representative of general clinical practice, these numbers provide a useful reference for estimating the yield of routine screening in these settings,” Perlis and Fihn wrote.
Another study published in JAMA Network Open, from the Kaiser Permanente system in California and Washington, examined the practicality of a risk-based alert system to identify patients who may attempt suicide.
The tool was developed from electronic health records and included characteristics such as depression symptoms, mental health visits, and past suicide attempts to estimate suicide risk. When patients entered a certain adjustable risk threshold, the tool would deploy an alert to physicians.
Using the 95th percentile threshold, the tool yielded 162 daily alerts and demonstrated a positive predictive value of 6.4%. (That works out to a number-needed-to-screen of 17 to prevent one suicide attempt, Perlis and Fihn calculated.) In contrast, the system sent out four alerts per day at the 99.5th risk percentile.
“While this provides useful estimates for planning, many key effectiveness, clinical, operational, ethical and legal questions remain,” the study’s lead author Andrea H. Kline-Simon, MS, of Kaiser’s research division in Oakland, California, told MedPage Today