Copilot or Autopilot? Evaluating the Impact of Generative AI Clinical Decision Support on Independent Clinical Reasoning in Pre-Clinical Osteopathic Medical Students

Summarise with VIRAT
Authors: Michael B. Moore
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Generative AI clinical decision support (GenAI-CDS) tools are entering medical training without evidence on whether AI access during formative learning supports or undermines independent clinical reasoning. Cognitive Load Theory (CLT) predicts that supports bypassing germane processing may impair schema construction in novice learners, producing a "never-skilling" effect distinct from the deskilling documented in experienced clinicians. This mixed-methods study randomizes teams of OMS-II students to GenAI-CDS-access, CDS-delayed, or CDS-absent conditions during an acute stroke simulation. The primary outcome is performance on a key-feature examination administered one week post-simulation without AI access. Secondary outcomes include task performance during simulation, perceived cognitive load (Paas mental effort scale, NASA-TLX), and behavioral coding of AI interaction using the CHART+M framework. The study evaluates the educational impact on learner reasoning, not the clinical validity of the AI platform. The key-feature examination functions as a transfer test: it operationalizes whether schemas formed during simulation persist when AI is removed. This design mirrors the structural logic of licensure board examinations (COMLEX-USA, COMAT), which are administered without AI access and measure independent reasoning capacity.

Fields of Study & Subjects

Artificial Intelligence in Healthcare and Education Clinical Reasoning and Diagnostic Skills Innovations in Medical Education

Publication details

Journal / source
OSF Preprints (OSF Preprints)
Publication date
20 March 2026
Language
EN
Document type
Preprint