Vanderbilt researchers think this AI solution will help hospitals flag suicide risk

Data scientists from Nashville, Tenn.-based Vanderbilt University Medical Center created a machine-learning algorithm to identify which patients may be at risk of suicide, according to Quartz.

The algorithm uses hospital-admissions data like diagnostic history, medication, age, gender and zip code to predict the likelihood a patient could attempt suicide. Using a dataset representing more than 5,000 patients admitted to the hospital for either self-harm or suicide attempts, researchers found in a trial that the algorithm accurately predicted if someone would attempt suicide in the next week about 84 percent of the time. The algorithm was 80 percent accurate at predicting whether a patient would attempt suicide within the next two years.

Colin Walsh, MD, an assistant professor at Vanderbilt's Department of Biomedical Informatics and a project lead on the machine-learning algorithm, told Quartz he wants to see widespread adoption of the solution within hospitals and medical centers. Most patients are only assessed for suicide risk when they seek psychiatric help or present outward symptoms like self-harm. Catching suicide risks early will give physicians more time to give proactive treatment for those not exhibiting obvious symptoms.

Dr. Walsh is working with physicians to create an intervention program based on the algorithm. In high-risk cases, a patient may be asked to spend several supervised days at the hospital, whether the patient volunteers or not. Follow-up care like appointments with psychiatrists after discharge and post-visit phone calls would be part of an intervention program for patients with lower suicide risks.

For the full report, click here.

More articles on artificial intelligence:
AI and the revenue cycle: Why automation should be your next core competency
Dr. Eric Topol turns critical eye on landmark FDA-approved AI device
Indiana U School of Medicine partners with Fujifilm for AI research

© Copyright ASC COMMUNICATIONS 2020. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.


Featured Webinars

Featured Whitepapers