Research Summary

Ambient AI Scribes Linked to Lower Work Exhaustion in Multistate Pragmatic Trial

Edited by:
Kate Young

Key Highlights

  • Ambient AI reduced work exhaustion and interpersonal disengagement among clinicians.
  • Documentation time decreased by 0.36 hours per day without loss of note quality.
  • Professional fulfillment increased modestly but not significantly.
  • Billing diagnostic code accuracy improved with AI-generated notes.

A stepped-wedge randomized trial published in The New England Journal of Medicine examined whether generative artificial intelligence (AI) scribes could mitigate documentation-related strain among ambulatory health care practitioners. The investigation provides real-world evidence on the impact of ambient AI on clinician well-being and documentation performance.

Researchers conducted a 24-week pragmatic trial across ambulatory clinics in two U.S. states. Sixty-six practitioners were individually randomized to one of three 6-week sequences of ambient AI use. Coprimary outcomes included professional fulfillment and work exhaustion/interpersonal disengagement, measured via the Stanford Professional Fulfillment Index. Secondary outcomes assessed time spent on notes, work outside work (WoW), documentation quality using the Provider Documentation Summarization Quality Instrument 9 (PDSQI-9), and diagnostic coding accuracy reviewed by professional coders. Linear mixed models supported intention-to-treat analyses. Over the study period, 71,487 clinical notes were authored, with 38% generated using the ambient AI tool.

Study Findings

Ambient AI use significantly reduced work exhaustion and interpersonal disengagement (−0.44 points; 95% CI, −0.62 to −0.25; P < .001) on the 5-point scale. Professional fulfillment increased slightly (+0.14 points; 95% CI, 0.004 to 0.28), though this change did not meet the threshold for meaningful improvement.

Secondary outcomes showed reductions in documentation burden. Clinicians spent 0.36 fewer hours per day on notes (95% CI, −0.55 to −0.17). Additionally, WoW decreased by 0.50 hours per day (95% CI, −0.90 to −0.09), though this result was sensitive to outliers and became nonsignificant when the top 3% of daily observations were removed.

Diagnostic billing codes improved significantly with ambient AI (P < .001). Documentation quality, assessed via PDSQI-9, remained high across all domains, with mean values ranging from 3.97 to 4.99 on a five-point scale. No drift in AI software performance was observed.

Clinical Implications

The findings suggest that ambient AI scribes may meaningfully reduce clinician exhaustion without compromising documentation accuracy or quality. While improvements in professional fulfillment were modest, the reduction in time spent on notes may offer workflow benefits in busy ambulatory settings.

Expert Commentary

“In a real-world randomized implementation, ambient AI reduced health care practitioners’ work exhaustion/interpersonal disengagement but did not significantly increase professional fulfillment,” the researchers concluded.

Reference
Afshar M, Baumann MR, Resnik F, et al. A pragmatic randomized controlled trial of ambient artificial intelligence to improve health practitioner well-being. N Engl J Med AI. 2025;2(12):10.1056/AIoa2500945. doi:10.1056/AIoa2500945