CHR Researchers Engage with Health Systems to Prevent Suicide
By Katie Essick, CHR Writer and Editor
Frances L. Lynch, PhD
Gregory N. Clarke, PhD
Bobbi Jo Yarborough, PsyD
From algorithms that can predict suicide risk to apps that promote mental health, Center for Health Research investigators are immersed in efforts to prevent suicide.
It’s a daunting challenge. Despite half a century of research, suicide rates have budged little. In 2021, nearly 50,000 people died by suicide in the U.S., and three times that many attempted it, while more than 12 million people seriously thought about taking their own lives, according to the Centers for Disease Control and Prevention.
Recent advances in technology show promise, including in determining who might be at risk of suicide so that steps can be taken to prevent it.
A decade ago, Senior Investigator Frances Lynch, PhD, began working with the nationwide Mental Health Research Network (MHRN) to develop one of the first algorithms to identify — through electronic health records — individuals at risk of dying by suicide.
Four algorithms were developed through the MHRN and found to be successful in identifying at-risk adults. One of these algorithms —referred to here as the Suicide Risk-Prediction model — predicts suicide attempts and deaths in the 90 days following a mental health visit.
This model has also been tested among adolescents. Suicide is the second leading cause of death among people ages 10-14 and third among those 15-24 years old, according to the National Institute of Mental Health (NIMH). Lynch and Distinguished Investigator Greg Clarke, PhD, conducted one study that found the risk model useful in identifying adolescents who needed further evaluation.
But until now, the algorithm hasn’t been employed in health care systems except for two very small pilot programs. CHR Senior Investigator Bobbi Jo Yarborough, PsyD, is directing a new study, funded by the NIMH, that will observe how the model is embedded in healthcare delivery in three health systems—Kaiser Permanente Northwest, Henry Ford Health System in Michigan, and HealthPartners in Minnesota.
“These models take electronic health record data — past depression and suicide risk screening scores, current and past mental health and substance use diagnoses, mental health medications a patient has received, past emergency department or inpatient care received, past suicide attempts or other injuries or poisonings, current and past medical conditions, gender, age, race and ethnicity, and other demographic and health factors — and pull them together to create a risk score,” Yarborough said.
“The model is efficient at pulling this information together, saving the provider the time required to wade through a patient’s chart to identify these risk factors. When you visit your provider, the model might flag you as someone who merits additional attention. It doesn’t tell the provider what to do, it’s not meant to replace their clinical judgment. It just tells the provider, ‘When you see this particular patient, you should be thinking about assessing suicide risk.’”
Yarborough’s previous studies have found patient perspectives on and experiences with the Suicide Risk-Prediction Model to be positive, and physicians to be supportive.
Yarborough and her team expect that the Suicide Risk-Prediction model will augment current methods used to identify those at risk of attempting suicide, boosting the use of outpatient behavioral health services, reducing the need for emergency and inpatient crisis services, and reducing suicide attempts.
The new study will focus on whether use of the model results in patients getting more preventive services and better preventive care, Yarborough said. “Personally, and professionally,” she shared, “I’ve known a lot of people who have struggled with suicide ideation or attempted suicide, and we, as health care systems, don’t handle that very well. I want to be involved in improving how we identify people at risk and care for them.”
Interventions and apps
Risk-prediction models are among several suicide-prevention studies in which CHR researchers are engaged.
Currently, Clarke, several other CHR researchers, and Joan Asarnow, PhD, a colleague at the University of California, Los Angeles, are leading an NIMH-funded suicide-prevention study, Step2Health, among 300 KPNW members ages 14-24 years.
The study compared “usual care” services for preventing self-harm, which includes the health system’s Zero Suicide program to increase suicide screening and appropriate treatment referrals, to a “stepped care” intervention with services matched to the level of risk, including dialectical behavior therapy (DBT) groups for those at greatest risk.
Lynch is conducting research on medical providers’ views of a new app called BRITEPath, which aims to provide youth who are at risk of suicidal behavior with a safety plan they can access easily and the ability to interact with their providers. “It’s well known that many youth and young adults are experiencing unprecedented levels of depression, including suicide ideation, and unfortunately a proportion of these will attempt suicide or be at high risk of suicide,” Lynch said. “They need close contact with their providers.”
Other suicide-related research
Clarke has conducted mental health research for more than 30 years, centered on both adults and youth. His innovative approaches have included the use of videos, websites, and apps. Recent suicide-prevention studies he has led or collaborated on include:
Lynch, a health economist, also has extensive experience in mental health research and has served as site principal investigator for the MHRN for more than 10 years. Her suicide-prevention research includes two NIMH-funded studies that:
- Used patient electronic health records and claims data from eight MHRN health systems to determine risk of suicide among a general population with substance use disorders. The research team found a significant risk, especially among women.
- Focused on gender identity disorder (GID, also called gender dysphoria) and suicide. The research team’s finding was that “patients with GID were at higher risk for suicide than patients with any other psychiatric diagnoses previously linked to suicide mortality.”
CHR researchers remain committed to preventing suicide. Lynch offered a message of hope. “I have worked with a variety of vulnerable populations, and I have seen firsthand that people can come from difficult situations and go on to live healthy lives. There’s a sense that it’s hopeless, but it is possible to survive and live well. I believe in that scientifically and personally, and it’s important to me to carry that message.”
If you or someone you know is struggling or in crisis, the 988 Suicide & Crisis Lifeline is available 24/7. Call or text 988 for help.