• Skip to primary navigation
  • Skip to main content
  • Skip to primary sidebar

information for practice

news, new scholarship & more from around the world


advanced search
  • gary.holden@nyu.edu
  • @ Info4Practice
  • Archive
  • About
  • Help
  • Browse Key Journals
  • RSS Feeds

Ambient AI and Measurement Bias in Psychiatric Notes

To the Editor In JAMA Psychiatry, Castro and colleagues reported that ambient artificial intelligence (AI) scribe use during primary care visits was associated with higher large language model (LLM)–estimated Research Domain Criteria (RDoC) symptom documentation across all 6 domains yet lower odds of a documented psychiatric intervention (depression-related International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10] code, antidepressant prescription, or behavioral health referral). The study advances evaluation beyond clinician time and satisfaction to clinical content. A key implication, however, is that the symptom end point is text native, meaning it is inferred from note narratives that the ambient system helps generate.

Read the full article ›

Posted in: Journal Article Abstracts on 08/06/2026 | Link to this post on IFP |
Share

Primary Sidebar

Categories

Category RSS Feeds

  • Calls & Consultations
  • Clinical Trials
  • Funding
  • Grey Literature
  • Guidelines Plus
  • History
  • Infographics
  • Journal Article Abstracts
  • Meta-analyses - Systematic Reviews
  • Monographs & Edited Collections
  • News
  • Open Access Journal Articles
  • Podcasts
  • Video

© 1993-2026 Dr. Gary Holden. All rights reserved.

gary.holden@nyu.edu
@Info4Practice