Vanderbilt University et al found that an automated approach for the Addiction Behaviors Checklist based on clinical notes performed better than diagnostic codes and can be generalizable to any electronic health record system. It can facilitate earlier identification of problematic opioid use.