Association of ambient artificial intelligence scribe availability with patient instruction documentation and readability in ambulatory care

Abstract Objective This retrospective cohort study evaluated the effect of an artificial intelligence (AI) scribe on patient instructions documentation in ambulatory internal medicine settings. Materials and Methods We analyzed 391 061 adult encounters at a single medical center from December 2023 to July 2025. A difference-in-differences analysis estimated the relationship of AI scribe availability with patient instructions documentation frequency. Mixed-effects logistic regression estimated the association between encounter-level AI scribe utilization and documentation, adjusted for patient, clinician, and encounter characteristics. Length and readability were compared across instructions created by AI alone, clinicians assisted by AI, and clinicians alone. Character-level edit distances between AI drafts and final documentation were calculated. Results AI scribe availability was associated with a 10.5 percentage-point increase in patient instructions documentation frequency (P = .018), and encounter-level AI use was associated with higher odds of documentation (aOR 4.32; 95% CI, 4.14-4.50). AI instructions were longer than clinician-authored content (median 1994 vs 482 characters) and scored higher on 3 text-complexity indices (all P < .001). Clinician editing of AI content increased length without meaningfully reducing complexity. AI drafts and final documentation differed substantially, with a median character-level edit distance of 39.7%. Discussion AI scribe availability was associated with increased patient instructions documentation. However, AI-generated instructions were longer and scored as more linguistically complex than clinician-authored instructions, and clinician editing did not close this gap. Conclusion Ambient AI scribes can improve availability but may increase complexity of patient instructions. Research into downstream effects including patient comprehension and satisfaction is warranted.

Authors

Institutions

Publication Details

Journal
Journal of the American Medical Informatics Association
Published
2026-10-22
DOI
https://doi.org/10.1093/jamia/ocag177
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Association of ambient artificial intelligence scribe availability with patient instruction documentation and readability in ambulatory care

Maria E. Byron, Kevin X. Shi, Megan Creasman, Jinoos Yazdany et al.
Journal of the American Medical Informatics Association
Artificial Intelligence in Healthcare and Education
article

Association of ambient artificial intelligence scribe availability with patient instruction documentation and readability in ambulatory care

Maria E. Byron, Kevin X. Shi, Megan Creasman, Jinoos Yazdany, Aris Oates, Gabriela Schmajuk, Augusto García-Agúndez
article en

Abstract

Abstract Objective This retrospective cohort study evaluated the effect of an artificial intelligence (AI) scribe on patient instructions documentation in ambulatory internal medicine settings. Materials and Methods We analyzed 391 061 adult encounters at a single medical center from December 2023 to July 2025. A difference-in-differences analysis estimated the relationship of AI scribe availability with patient instructions documentation frequency. Mixed-effects logistic regression estimated the association between encounter-level AI scribe utilization and documentation, adjusted for patient, clinician, and encounter characteristics. Length and readability were compared across instructions created by AI alone, clinicians assisted by AI, and clinicians alone. Character-level edit distances between AI drafts and final documentation were calculated. Results AI scribe availability was associated with a 10.5 percentage-point increase in patient instructions documentation frequency (P = .018), and encounter-level AI use was associated with higher odds of documentation (aOR 4.32; 95% CI, 4.14-4.50). AI instructions were longer than clinician-authored content (median 1994 vs 482 characters) and scored higher on 3 text-complexity indices (all P < .001). Clinician editing of AI content increased length without meaningfully reducing complexity. AI drafts and final documentation differed substantially, with a median character-level edit distance of 39.7%. Discussion AI scribe availability was associated with increased patient instructions documentation. However, AI-generated instructions were longer and scored as more linguistically complex than clinician-authored instructions, and clinician editing did not close this gap. Conclusion Ambient AI scribes can improve availability but may increase complexity of patient instructions. Research into downstream effects including patient comprehension and satisfaction is warranted.

Journal of the American Medical Informatics Association
University of California, San Francisco (US)
University of Oxford, National Institute of Diabetes and Digestive and Kidney Diseases
Good health and well-being
Openalex Percentile: Top 21%
Artificial Intelligence in Healthcare and Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon