Related Experiment Videos
Ambient AI Scribes to Create Educational Feedback Notes for Medical Students: Randomized Trial
Jaideep S Talwalkar1, David Chartash2, Lisa Zhang2
1Departments of Medicine and Pediatrics, Yale School of Medicine, 367 Cedar Street, Bldg D, New Haven, CT, 06510, United States, 1 203-737-4190, 1 203-737-4199.
JMIR Medical Education
|May 28, 2026
Summary
Ambient artificial intelligence (AI) scribes enhance medical education feedback quality without increasing instructor workload. This AI scribe-assisted workflow shows potential for transforming feedback documentation, despite minor inaccuracies requiring review.
Area of Science:
- Medical Education
- Artificial Intelligence in Healthcare
- Clinical Competence Development
Background:
- High-quality feedback is crucial for medical trainees' clinical competence and professional growth.
- Faculty face challenges in providing written feedback due to documentation burden.
- Ambient AI scribes offer a potential solution by capturing and structuring clinical interactions.
Purpose of the Study:
- To evaluate ambient AI scribes in generating educational feedback for first-year medical students.
- To assess the impact of AI scribe-assisted workflows on feedback quality and instructor effort.
Main Methods:
- Randomized controlled trial comparing human-only and AI scribe-assisted feedback workflows.
- AI scribe generated transcripts, summarized into notes using large language models, and edited by instructors.
- Feedback quality measured by the Evaluation of Feedback Captured Tool (EFeCT); task load and usability assessed via NASA-TLX and SUS.
Main Results:
- AI scribe-assisted feedback (human-edited and unedited) demonstrated significantly higher EFeCT scores than human-only feedback (P<.001).
- AI-assisted outputs were longer than human-only narratives (P<.001).
- AI-generated feedback had low rates of mischaracterization (6.8%) and hallucination (1.7%), with most errors corrected during editing; task load and usability were comparable between groups.
Conclusions:
- Ambient AI scribe-assisted workflows improve written narrative feedback quality without increasing instructor effort.
- This technology has the potential to revolutionize feedback documentation in medical education.
- Careful review is necessary to address occasional inaccuracies in AI-generated feedback.
Related Concept Videos
02:07Blind Procedures
Ideally, the people who observe and record the children’s behavior are unaware of who was assigned to the experimental or control group, in order to control for experimenter bias. Experimenter bias refers to the possibility that a researcher’s expectations might skew the results of the study. Remember, conducting an experiment requires a lot of planning, and the people involved in the research project have a vested interest in supporting their hypotheses. If the observers knew which child was...
00:46Feedback Inhibition
Biochemical reactions are occurring constantly in cells, converting starting substances to different products, usually with the help of enzymes that speed the reactions. Without enzymes, it would take far too long for most reactions to occur to be useful to the cell!