AI scribes & ambient AI

Sin Hung Timothy Yu, Julie Li +4

What shapes adoption of digital scribes, and what do they change for patients, clinicians, and organisations? This PRISMA-ScR scoping review searched MEDLINE, CINAHL, Web of Science, SCOPUS, and EMBASE for original studies or case reports evaluating digital scribe implementation in real-world care, mapping themes to the updated Consolidated Framework for Implementation Research and its Outcomes Addendum. Of 4772 studies screened, 29 were included. Scribes were generally acceptable (n=11) and usable (n=8), though nine reported accuracy concerns; reported impacts included reduced documentation burden (n=18), improved clinician wellbeing (n=12), and better patient-clinician interaction (n=10). Only three examined cost or productivity. The abstract reports no pooled effect sizes.

Zhaoqiang Zhou, John Geracitano +3

How do rural patients view ambient AI scribes, and which characteristics predict acceptance? A cross-sectional analysis of 1,050 rural respondents in the 2024 Canadian Digital Health Survey dichotomized three outcomes — trust in documentation accuracy, perceived interaction benefit, and future-use preference — and fit XGBoost classifiers using sex, age, race/ethnicity, education, employment, income, chronic disease, self-reported health, and high-speed internet access, summarizing subgroup differences as marginally standardized predicted probabilities. Endorsement declined across the three domains. Future-use preference was 0.388 for males versus 0.313 for females, 0.466 for graduate degrees versus 0.308 for less than high school, and 0.408 with chronic disease versus 0.298 without. Internet access showed similar probabilities throughout.

Jose María Alcázar-Peral, Juan Antonio Álvaro-de la Parra +7

Can an ambient AI scribe scale across emergency departments without degrading documentation or patient experience? This 12-month multicenter retrospective observational study covered five emergency specialties at 48 Spanish hospitals (February 2025–January 2026), including all level 4 and 5 consultations among roughly 2.27 million ED visits. The scribe was used in 1,032,558 consultations (45.3%), with monthly adoption rising from 7.7% to 57.8% and 2,097 physicians using it at least once. Scribe-assisted consultations were shorter (mean relative time savings 21.8%, p<0.001), transcription accuracy averaged 93.9%, and audited report quality and patient Net Promoter Scores were higher; the abstract reports no effect sizes for the quality and experience comparisons.

Alexandra Rabotin, Efren Aguilar +6

Do ambient AI scribes carry interpreter errors into the clinical note? Using simulated English- and Spanish-language clinical encounters mediated by interpreters, the authors evaluated whether documentation generated by ambient AI scribes reproduced interpretation errors introduced during the visit. Scribes propagated interpreter errors into the resulting notes, with propagation patterns differing by speaker role and by error type. The published abstract reports no effect sizes, error counts, or comparative rates. The authors frame the results as a case for further evaluation of AI-scribe performance in multilingual and interpreter-mediated care.