Could suicide risk be predicted from a patient’s records?

A study led by Boston Children’s Hospital and Massachusetts General Hospital demonstrates that a predictive computer model can identify patients at risk for attempting suicide from patterns in their electronic health records — an average of two years ahead of time.