A new study from the UC Davis Violence Prevention Research Program (VPRP) suggests that machine learning, a type of artificial intelligence, can help identify gun buyers who are at high risk of suicide. It also identified individual and community characteristics that predict firearm suicide. The study was published in Open JAMA Network.
Previous research has shown that suicide risk is particularly high immediately after purchase, suggesting that the purchase itself is an indicator of elevated suicide risk.
Risk factors identified by the algorithm to predict suicide by firearm included:
- first purchaser of firearms
- White race
- live very close to the arms dealer
- buy a revolver
“While limiting access to firearms among those most at risk for suicide presents a critical opportunity to save lives, accurately identifying those at risk remains a key challenge. Our results suggest the potential utility of firearms registries in identifying high-risk individuals to aid in suicide prevention,” said Hannah S. Laqueur, assistant professor in the Department of Emergency Medicine and lead author of the study. .
In 2020, nearly 48,000 Americans committed suicide, of which more than 24,000 were firearm suicides. Firearms are by far the deadliest method of suicide. Access to firearms has been identified as a major risk factor for suicide and is a potential approach to suicide prevention.
To see if an algorithm could identify gun buyers at risk of firearm suicide, researchers analyzed data on nearly five million firearm transactions from the California Dealer Sale Registry database ( DROS). The records, which spanned from 1996 to 2015, represented almost two million people. They also analyzed firearm suicide data from California death records between 1996 and 2016.
The team generated 41 predictor variables from the transaction data. Among other data points, the researchers looked at firearm categories (such as a revolver or semiautomatic pistol), caliber size, price, where the gun was purchased, the purchaser’s previous gun purchases, weapons, gender, race and ethnicity, and age.
The researchers ran a random forest classification algorithm, which can generate predictions over a wide range of data. They used transaction-level data to predict firearm suicide within a year of purchase.
Among the top 5% of transactions identified as riskiest, about 40%, or 379 of 983, were associated with a buyer who died by gunshot suicide within a year.
Among the very small number of transactions with a random forest score or predicted probability of 0.95 and above, 69%, or 24 of 35, were affiliated with a buyer who died by gunshot suicide within one year.
“Research has established a clear and strong association between firearm acquisition and ownership and firearm suicide risk, but this study adds to the growing evidence that computational methods can aid in the identification of groups of high risk and the development of specific interventions,” Laqueur said. .
The researchers caution that the first-of-its-kind study was largely a “proof of concept.” Still, the results suggest the potential for using firearms registries to identify high-risk individuals to aid suicide prevention. They also noted that many firearm suicides occurred among individuals classified as “low risk,” so other forms of intervention would be needed to prevent firearm suicide in this group.
Other study authors include Colette Smirniotis, Christopher McCort and Garen J. Wintemute of VPRO and the California Gun Violence Research Center.
Materials provided by University of California-Davis Health. Original written by Lisa Howard. Note: content can be edited for style and length.