Advance-CTR

Advance RI-CTR Mentored Research Award (2022)

"Developing a Computational Approach for Identifying Suicidal Ideation and Behavior using the Electronic Health Record"

Suicide is a leading, and growing, cause of death in the United States. Dr. Arias' study leverages electronic health record (EHR) data to develop classification algorithms for identifying suicidal ideation and suicide attempt cases. These data have the potential to inform interventions for monitoring suicidal patients that could serve as an early warning sign to help identify high-risk  patients.

With the goal of reducing suicide rates by 20% in the next few years, there is an immediate need for research that can inform suicide intervention development. Considering that over 75% of individuals receive some form of healthcare prior to suicide, data in the electronic health record (EHR) can be very informative for suicide detection and prevention efforts. However, leveraging these data for clinical research can be challenging. In addition, the current practice for studying suicide risk involves the use of ICD-9/10-CM codes, which have been found to significantly underestimate suicidal ideation (SI) and suicide attempt (SA) cases in the EHR.

Although there have been numerous analyses of suicide outcomes within EHRs, there is limited information on the critical initial step of developing effective methods for reliable identification of SI/SA cases documented at the point of care to advance prediction and monitoring of future suicide risk. Dr. Arias' study will develop procedures for creating a suicide database using data from the Care New England (CNE) Epic EHR system that will then be used to create and validate classification algorithms for identifying SI and SA cases within the EHR.