A team of researchers at the Perelman School of Medicine at the University of Pennsylvania has developed a data-driven tool that predicts who is at highest risk of stimulant overdoses, using demographic and clinical information to guide earlier intervention.