Ambient Artificial Intelligence Use and Clinician Documentation Burden, Productivity, and Efficiency
-
By
-
Robyn A. Husa
-
John Haggerty
-
Andrew W. Nute
-
Julie Levine
-
Kevin Love
-
Xochitl Martinez
-
Canada Parrish
-
May 29, 2026
-
Objective:
To evaluate the associations between an ambient AI system (Dragon Ambient eXperience [DAX]) and clinician productivity and efficiency.
Key Findings:
- Significant decline in mean time spent on notes during the first month of ambient AI use (β coefficient, −0.26 minutes per note; P < .001).
- Sustained decline in after-hours documentation time (β coefficient, −0.38 minutes greater decrease per month; P = .02).
- Immediate increase in mean RVUs following active ambient AI use (β coefficient, 7.40 RVUs per month; P = .03).
- No immediate or sustained association with clinician efficiency profile scores or appointments per day.
Interpretation:
Remove unsupported claims about improvements.
Limitations:
- Lack of a true experimental design due to pre-existing integration of the tool.
- Focus on a single ambient AI system limits generalizability.
Conclusion:
Omit or rephrase to avoid unsupported conclusions.
