AI and the Gap Between Medical Information and Clinical Decisions (Preprint)

UNSTRUCTURED Much of the discussion about generative artificial intelligence (AI) in medicine has focused on whether AI-generated information is accurate, biased, explainable, or prone to hallucination. These are important concerns, but they may overlook a more subtle problem. Generative AI can take fragmented and uncertain evidence and assemble it into a coherent narrative that feels more complete than the underlying evidence actually is. We call this premature resolution: the point at which a plausible and fluent synthesis creates enough confidence to encourage action even though clinically important uncertainty remains unresolved. Consider a patient who uses AI to integrate molecular findings, published literature, clinical trials, and experimental treatment options. They may arrive in clinic with a coherent narrative and a level of confidence that exceeds the strength of the evidence supporting it. The difficulty is not necessarily that the information is incorrect, but that AI's fluency can obscure what is missing, uncertain, or poorly applicable to the individual patient. The clinical task then becomes one of restoring context: identifying what the AI could not access, could not adequately contextualize, or could not weigh against competing clinical priorities. This requires asking not only whether an AI-generated answer is correct, but what is absent from it, how directly the evidence applies to the patient, and what risks follow from treating a synthesis as a decision. As generative AI, AI-enabled Software as a Medical Device (SaMD), and increasingly agentic clinical systems become part of patient information seeking and medical practice, distinguishing a coherent answer from a clinically justified decision will become an increasingly important skill.

Authors

Publication Details

Published
2026-11-16
DOI
https://doi.org/10.2196/preprints.116724
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

AI and the Gap Between Medical Information and Clinical Decisions (Preprint)

Mustafa Khasraw, Robin Smetana, Anna Macon Corcoran, Nik Zeps
Artificial Intelligence in Healthcare and Education
preprint

AI and the Gap Between Medical Information and Clinical Decisions (Preprint)

Mustafa Khasraw, Robin Smetana, Anna Macon Corcoran, Nik Zeps
preprint en

Abstract

UNSTRUCTURED Much of the discussion about generative artificial intelligence (AI) in medicine has focused on whether AI-generated information is accurate, biased, explainable, or prone to hallucination. These are important concerns, but they may overlook a more subtle problem. Generative AI can take fragmented and uncertain evidence and assemble it into a coherent narrative that feels more complete than the underlying evidence actually is. We call this premature resolution: the point at which a plausible and fluent synthesis creates enough confidence to encourage action even though clinically important uncertainty remains unresolved. Consider a patient who uses AI to integrate molecular findings, published literature, clinical trials, and experimental treatment options. They may arrive in clinic with a coherent narrative and a level of confidence that exceeds the strength of the evidence supporting it. The difficulty is not necessarily that the information is incorrect, but that AI's fluency can obscure what is missing, uncertain, or poorly applicable to the individual patient. The clinical task then becomes one of restoring context: identifying what the AI could not access, could not adequately contextualize, or could not weigh against competing clinical priorities. This requires asking not only whether an AI-generated answer is correct, but what is absent from it, how directly the evidence applies to the patient, and what risks follow from treating a synthesis as a decision. As generative AI, AI-enabled Software as a Medical Device (SaMD), and increasingly agentic clinical systems become part of patient information seeking and medical practice, distinguishing a coherent answer from a clinically justified decision will become an increasingly important skill.

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