Beyond Model Choice: A Harness-Centered Framework for Designing Clinical Large Language Model Agents
Journal of Medical Internet Research · 09.11.2026 · 28 · e101269
DOI: 10.2196/101269 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Medical Internet Research, online erschienen am 09.11.2026. Verfasst von Jialin Liu und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medical Informatics, West China Hospital of Sichuan University, Chengdu, Sichuan, China. Hauptthemen laut MEDLINE: Large Language Models, Medical Informatics. Die Arbeit nennt 36 Literaturangaben.
Abstract (Originalsprache)
Clinical large language model agents depend on both model capability and the surrounding system that selects information, constrains actions, and verifies execution. In this Viewpoint, we argue that this harness should be an explicit unit of clinical agent design, evaluation, and reporting. Our aim is to help clinical informatics and development teams specify this layer and to help evaluators and institutional governance teams assess its contributions and risks. Selected benchmarks show substantial performance differences across harness configurations, but architecture comparisons, bundled improvements, and within-model component comparisons support different levels of causal attribution. We propose a revisable 3-tier framework: tier 1 specifies general orchestration functions, including prompts, tools, output schemas, control flow, memory, and generic verification; tier 2 specifies their clinical adaptations for knowledge integration, differential reasoning, missingness and temporality, safety verification, and workflow alignment; and tier 3 specifies a failure-driven evaluation cycle applied to both. Our contribution is a framework that makes these connections explicit. We provide classification rules, an illustrative clinical example, a proposed evaluation protocol, and a minimum reporting checklist. Observable execution traces are distinguished from private model reasoning, with provenance and privacy safeguards supporting retrospective reconstruction. Evaluation should separate task success, clinical correctness, process safety, and operational reliability. Lightweight fine-tuning can complement harness design; their interaction requires controlled study. The framework is a proposal for empirical testing, not a validated taxonomy or evidence of deployment safety.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Medical Informatics, West China Hospital of Sichuan University, Chengdu, Sichuan, China.
- Department of Otolaryngology-Head and Neck Surgery, West China Hospital of Sichuan University, Chengdu, Sichuan, China.
- Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, TN, United States.
- Department of Computer Science, Vanderbilt University, Nashville, TN, United States.
ORCID-Profile: Jialin Liu, Siru Liu
Zitieren
APA
Liu, J., & Liu, S. (2026). Beyond Model Choice: A Harness-Centered Framework for Designing Clinical Large Language Model Agents. Journal of Medical Internet Research, 28, e101269. https://doi.org/10.2196/101269
Vancouver
Liu J, Liu S. Beyond Model Choice: A Harness-Centered Framework for Designing Clinical Large Language Model Agents. Journal of Medical Internet Research. 2026;28:e101269. doi:10.2196/101269
BibTeX
@article{Liu2026,
author = {Liu, Jialin and Liu, Siru},
title = {Beyond Model Choice: A Harness-Centered Framework for Designing Clinical Large Language Model Agents},
journal = {Journal of Medical Internet Research},
year = {2026},
volume = {28},
pages = {e101269},
doi = {10.2196/101269}
}
EUVIMED