A groundbreaking study from Vanderbilt University Medical Center reveals that AI-driven clinical alerts may help doctors detect patients at risk of suicide more effectively. This development could significantly improve prevention efforts in routine medical settings.
The research centers around the Vanderbilt Suicide Attempt and Ideation Likelihood model (VSAIL), which uses AI to prompt doctors to screen for suicide risk during regular clinic visits. Led by Colin Walsh, MD, MA, the study tested VSAIL in three neurology clinics at Vanderbilt University Medical Center (VUMC).
STUDY DESIGN AND COMPARISON
The team compared two alert systems. These systems included interruptive pop-up alerts and a more passive method. The passive method displays risk information in the patient’s electronic health record. The results demonstrated a significant difference in effectiveness between the two systems.
INTERRUPTIVE ALERTS SHOW GREATER EFFECTIVENESS
The interruptive alerts briefly interrupted the doctor’s workflow. They prompted doctors to conduct suicide risk assessments in 42% of the cases. In contrast, the passive system triggered screenings in only 4% of visits.
ADDRESSING THE URGENT NEED FOR SUICIDE RISK SCREENING
Suicide rates have surged in the U.S., now being the 11th leading cause of death. Studies reveal that 77% of individuals who die by suicide have seen a healthcare provider within the past year. They had at least one visit before their death. Despite this, routine suicide screenings stay impractical in many medical settings.
VSAIL’S ROLE IN IDENTIFYING HIGH-RISK PATIENTS
The VSAIL system, developed at Vanderbilt, analyzes routine electronic health data to calculate a patient’s 30-day risk for suicide attempts. Previous testing showed its effectiveness, flagging one in 23 patients who later expressed suicidal thoughts.
WHY NEUROLOGY CLINICS WERE CHOSEN
The study specifically focused on neurology clinics due to the higher suicide risk linked to certain neurological conditions. These patients are more vulnerable, and the team sought to test the AI model in this high-risk group.
EXPLORING FEASIBILITY IN BUSY CLINICS
The AI system flagged just 8% of all patient visits for suicide risk screening. This makes it a more practical solution for busy medical settings. This selective approach allows healthcare providers to prioritize high-risk patients without overwhelming their workflow.
STUDY DETAILS AND OUTCOMES
The study reviewed 7,732 patient visits over six months, generating 596 screening alerts. Despite the prompt screenings, no patients in the study were found to have attempted suicide or reported suicidal ideation in the following 30 days.
BALANCING EFFECTIVENESS AND ALERT FATIGUE
Interruptive alerts were more effective. Nonetheless, the researchers cautioned against potential “alert fatigue.” This occurs when doctors become overwhelmed by frequent notifications. Future studies should focus on balancing effectiveness with minimizing burnout for healthcare professionals.
IMPLICATIONS FOR FUTURE SUICIDE PREVENTION
Walsh emphasized the need for a balance between effective alerts and their potential drawbacks. He believes that automated systems like VSAIL can assist doctors. When combined with well-designed notifications, these systems could help identify more patients in need of suicide prevention services.
