New Delhi: New research reveals that artificial intelligence (AI) can significantly assist healthcare providers in identifying patients who may be at risk for suicide, thereby potentially enhancing preventive efforts in typical medical environments.

The study, published in the JAMA Network Open, explored and compared two distinct methods for flagging suicide risk: automatic pop-up alerts that disrupt the doctor’s workflow and a non-intrusive approach that simply presents risk information within the patient’s electronic health record.

The findings clearly stated that the interruptive alerts proved far more effective. These alerts prompted doctors to conduct comprehensive suicide risk assessments in response to 42 percent of screening alerts, whereas the passive notifications only led to such assessments in 4 percent of cases.

Colin Walsh, an associate professor of biomedical informatics, medicine, and psychiatry at Vanderbilt University Medical Center, noted that most individuals who die by suicide have engaged with healthcare services within the year leading up to their death, often for matters unrelated to mental health.

In assessing whether their AI model, known as the Vanderbilt Suicide Attempt and Ideation Likelihood model (VSAIL), could effectively prompt healthcare providers in three neurology clinics to screen patients for suicide risk during routine visits, the researchers aimed to streamline the process of identifying vulnerable individuals.

“Universal screening isn’t practical in every context. Therefore, we developed VSAIL to assist in recognizing high-risk patients and initiating focused screening conversations,” stated Walsh.

The VSAIL model processes routine data from electronic health records to evaluate a patient’s risk of suicide attempts within a 30-day timeframe.

The researchers believe similar AI systems could be tested in various healthcare environments to further enhance suicide risk detection.

Walsh emphasized the need for healthcare providers to weigh the benefits of interruptive alerts against their possible drawbacks.

The overall results from this study suggest that a synergy of automated risk detection with well-crafted alerts could enable better identification of patients requiring suicide prevention services.

It is further highlighted that a staggering 77 percent of individuals who die by suicide have interacted with a primary care provider within the year preceding their death.

With inputs from IANS