Has Speech Recognition Software Advanced Enough for Safe Use with the EHR

​Speech recognition (SR) software needs significant development before it is safe and effective to use for documentation within the electronic health record (EHR), according to a July 27, 2017, article in the Journal of the American Medical Informatics Association. The authors randomly allocated 35 clinicians in an emergency department to perform documentation tasks on the EHR using a keyboard and mouse or SR software. Tasks varied in complexity and in whether interruptions occurred. Mean task completion times were 18.1% slower when using SR software compared with completion times for keyboard and mouse: 17.0% slower for simple tasks and 18.4% slower for complex tasks.