Can Suicide Be Prevented with Predictive Analytics Model Data?

The statistics can be staggering. Every 11 minutes, a person dies by suicide. That means more than 48,000 deaths per year. As high as that number is, 12.3 million adults contemplate suicide with 1.7 million attempting it. Prevention is the best way to reduce deaths, but how do we identify who is most likely to contemplate suicide?
 

Preventing Suicide Attempts

According to the Centers for Disease Control (CDC), suicide can be prevented using a comprehensive, data-driven approach. Healthcare claims can be analyzed for potential risk factors, including substance use, certain behavioral health diagnoses, inpatient mental health stays, and prior suicide attempts. Using the resulting data in a predictive analytics model can help identify people who are at higher risk for contemplating suicide.

Having specially trained mental health professionals proactively contact people at risk of contemplating suicide and offering them support can help reduce the likelihood they will attempt suicide. Feeling a sense of connection to others is one of the protective factors that can reduce suicide attempts. Sometimes just feeling heard can stop people from taking their lives.
 

Putting the Data to Work

Dr. Jessica Chaudhary, medical director for Elevance Health’s Carelon Behavioral Health and a psychiatrist by training, agrees suicide can be prevented. “Determining who is at risk — in time to make a difference — is the key,” Chaudhary said.

Chaudhary and her team developed a predictive model that identifies people who have at least a 10% risk of attempting suicide over the next 12 months. People found to be at risk are then contacted by behavioral health case managers, peer recovery specialists, or people who have lived experience. “Knowing who is at risk means you can reach out, connect with them, and give them access to care and assistance,” said Chaudhary.
 

Reducing Risk and Increasing Safety

Chaudhary and her team’s efforts started in 2018. They found they could potentially identify someone at risk and intervene five months before a suicide attempt. “That’s what our data told us, that we had this window of time to offer an intervention and try to change the trajectory,” Chaudhary said.

Members have the choice of accepting the additional support the trained specialists provide. When they do, they can build safety plans based on their experiences and resilience to reduce the risk of attempting suicide. Safety plans provide detailed methods and techniques to identify and respond to triggers when someone is having thoughts of suicide.

These plans can include distraction methods to get people to think of things other than suicide, having an accessible list of self-selected people to call (often family and friends), listings of professional resources to contact such as a therapist or 988, as well as ways to cultivate a safe home environment that reduces access to the means to act on attempting suicide.

Results from the proactive interventions demonstrate a more than 20% reduction in adolescent and young adult suicidal events for commercial risk-based members engaged in the program relative to control groups. Currently 4,200 members are participating in the program which is available to the following:

• Members age 10 and up in employer-sponsored plans
• Members in individual Medicare Advantage plans
• Members in Medicaid plans in Georgia, West Virginia, Wisconsin, Kentucky, South Carolina, Iowa, and Minnesota

Continually Improving

The team will continually adjust the predictive models as more data are gathered and analyzed to mitigate any bias as well as keep up with emerging and changing trends. “If we can identify people at high risk for these devastating types of attempts, intervene, and help someone onto a safer path, I don’t know what could be better than that,” Chaudhary said.