Predicting suicide attempts two years in advance

A Florida State University researcher has used advancements in machine learning and artificial intelligence to predict suicide attempts up to two years before they occur.

The study began by analyzing more than two million patient electronic health records (EHRs) to identify 3,200 patients who had attempted suicide. Jessica Ribeiro, the lead researcher on the study, developed machine learning to be able to predict which individuals would attempt suicide two years in advance with up to 80 percent accuracy.

“The machine learns the optimal combination of risk factors,” Ribeiro told Florida State University News, of the study set to appear in Clinical Psychological Science. “What really matters is how this algorithm and these variables interact with one another as a whole. This kind of work lets us apply algorithms that can consider hundreds of data points in someone’s medical record and potentially reduce them to clinically meaningful information.”

Machine learning was able identify certain information within EHRs to “learn” which factors lead to attempted suicide. That accuracy of 80 percent two years before increased to 92 percent a week before a suicide attempt. This machine learning appears able to bring preventive medicine to those most in need.

“This study provides evidence that we can predict suicide attempts accurately,” Ribeiro said. “We can predict them accurately over time, but we’re best at predicting them closer to the event. We also know, based on this study, that risk factors—how they work and how important they are—also change over time.”