Sayon Dutta, MD, MPH

Emergency physicians experience high patient volume and substantial documentation burden, which can contribute to burnout. Human scribes can reduce this burden, but can be cost prohibitive.

Ambient artificial intelligence (AI) offers a promising alternative by automating note generation during clinician-patient conversations, but its impact on documentation workloads has been unclear.

A retrospective study led by Sayon Dutta, MD, MPH, a physician in the Mass General Brigham Department of Emergency Medicine, and Jacqueline Guan-Ting You, MD, MBI, of the Division of General Internal Medicine, compared how ambient AI scribes and human scribes affected physicians’ documentation time and productivity.

The team reviewed more than 198,000 emergency department encounters across four different sites, during which the attending physician used either ambient AI scribes (4.3%), human scribes (8%) or no scribe (87.7%).

Human scribes lowered median physician documentation times by about 3.3 minutes per note when compared to encounters without scribes. Ambient AI scribes produced longer notes and a smaller decrease of 1.6 minutes per note. Both effects were more pronounced in academic centers than community hospitals.

Neither approach increased hourly clinical productivity, suggesting that both ambient AI scribes and human scribes reduce clerical burden but do not boost throughput or revenue.

Published in Annals of Emergency Medicine on June 11, 2026 | Read the paper: “AI and Human Scribes Reduce Emergency Physician Documentation Burden by Varying Degrees”

Summary reviewed by Sayon Dutta, MD, MPH