Suicide is the tenth leading cause of death in the United States. It has been rising throughout the country for a long time. Suicide claims the lives of 14 in 100,000 Americans each year. Nationally, around 8.5 percent of suicide attempts end in death.
A machine-learning algorithm capable of predicting suicide attempts recently underwent a prospective trial at the same institution where it was originally developed, Vanderbilt University Medical Center.
How the Algorithm was Developed
The study was carried out over 11 consecutive months, which concluded in April 2020. Multiple adult patients were seen at VUMC, where predictions ran silently for the entire period. The algorithm is named Vanderbilt Suicide Attempt and Ideation Likelihood (VSAIL) model and uses routine information from electronic health records. This information is used to calculate the 30-day risk of return visits for suicide attempts and suicidal ideation.
In the institute, adult patients were divided into eight groups according to their risk scores per the algorithm. The topmost group accounted for more than one-third of the total suicidal attempts documented in the study, and approximately half of the cases of suicidal ideation. As documented in the electronic health reports, one in 23 in the high-risk group reported having suicidal thoughts, and one in 271 attempted suicide.
How AI (artificial intelligence) can Save the Day
Over the 11 months, around 78,000 adult patients were seen in the hospital, emergency room, and surgical clinic at VUMC. As per the HER documentation, 395 individuals in this group reported having suicidal thoughts. 85 people lived through at least one suicidal attempt, and 23 surviving repeated attempts. For every 271 people identified in the high-risk group, one returned for treatment for a suicide attempt.
It is not possible asking all the patients, who are visiting a medical facility about suicidal thinking, nor is it ok to ask all patients. In such cases, the algorithm might be efficient to detect the risk of suicide and help the patients accordingly.