Deliberative multi-agent large language models improve clinical reasoning in ophthalmology

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Misaghi, Ehsan, Berkowitz, Sean T, Chen, Bing Yu, Chen, Qingyu, Duval, Renaud, Keane, Pearse A, Mammo, Danny A, Ong, Ariel Yuhan, Sevgi, Mertcan, Sharma, Sumit, Srivastava, Sunil K, Tham, Yih Chung, Antaki, Fares
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914413734264832
author Misaghi, Ehsan
Berkowitz, Sean T
Chen, Bing Yu
Chen, Qingyu
Duval, Renaud
Keane, Pearse A
Mammo, Danny A
Ong, Ariel Yuhan
Sevgi, Mertcan
Sharma, Sumit
Srivastava, Sunil K
Tham, Yih Chung
Antaki, Fares
author_facet Misaghi, Ehsan
Berkowitz, Sean T
Chen, Bing Yu
Chen, Qingyu
Duval, Renaud
Keane, Pearse A
Mammo, Danny A
Ong, Ariel Yuhan
Sevgi, Mertcan
Sharma, Sumit
Srivastava, Sunil K
Tham, Yih Chung
Antaki, Fares
contents Large language models (LLMs) show potential for ophthalmic clinical reasoning, yet individual models risk introducing harm. We evaluated whether multi-agent LLM deliberative councils improve diagnostic performance and mitigate harm compared to individual LLMs. In a comparative cross-sectional study, we assessed 12 individual LLMs and three multi-agent councils on 100 ophthalmology clinical vignettes. Each council comprised four models assembled by type: proprietary flagship, proprietary fast, and open-source. Models independently answered a vignette, anonymously ranked one another's responses, and a designated chair synthesized all responses and peer reviews into a final answer. Councils consistently outperformed pooled individual models across all three tiers. Accuracy improved for proprietary flagship (95.0% vs 90.8%; risk difference [RD]: 4.25 [95% CI: 0.45, 8.05]), proprietary fast (96.0% vs 86.5%; RD: 9.50 [5.31, 13.59]), and open-source councils (91.0% vs 83.2%; RD: 7.75 [4.17, 11.33]). Harm rates declined for proprietary flagship (10.0% vs 22.5%; RD: -12.50 [-16.86, -8.14]), proprietary fast (16.0% vs 31.8%; RD: -15.75 [-21.49, -10.01]), and open-source councils (22.0% vs 38.5%; RD: -16.50 [-22.27, -10.73]). Coverage analysis revealed net positive gains for accuracy (ΔCoverage: 4.4-9.8 percentage points) and safety (ΔCoverage: 13.6-20.6), indicating councils recovered correct diagnoses and averted harm. Councils elevated correct diagnoses to higher rank positions; and produced more complete differentials and management plans (all P<.05). Harmful council responses showed reduced combined commission-and-omission errors and tended to be less severe. Structured deliberation via multi-agent LLM councils may enhance the reliability of LLM-assisted ophthalmic clinical reasoning.
format Preprint
id arxiv_https___arxiv_org_abs_2603_21447
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Deliberative multi-agent large language models improve clinical reasoning in ophthalmology
Misaghi, Ehsan
Berkowitz, Sean T
Chen, Bing Yu
Chen, Qingyu
Duval, Renaud
Keane, Pearse A
Mammo, Danny A
Ong, Ariel Yuhan
Sevgi, Mertcan
Sharma, Sumit
Srivastava, Sunil K
Tham, Yih Chung
Antaki, Fares
Computers and Society
Large language models (LLMs) show potential for ophthalmic clinical reasoning, yet individual models risk introducing harm. We evaluated whether multi-agent LLM deliberative councils improve diagnostic performance and mitigate harm compared to individual LLMs. In a comparative cross-sectional study, we assessed 12 individual LLMs and three multi-agent councils on 100 ophthalmology clinical vignettes. Each council comprised four models assembled by type: proprietary flagship, proprietary fast, and open-source. Models independently answered a vignette, anonymously ranked one another's responses, and a designated chair synthesized all responses and peer reviews into a final answer. Councils consistently outperformed pooled individual models across all three tiers. Accuracy improved for proprietary flagship (95.0% vs 90.8%; risk difference [RD]: 4.25 [95% CI: 0.45, 8.05]), proprietary fast (96.0% vs 86.5%; RD: 9.50 [5.31, 13.59]), and open-source councils (91.0% vs 83.2%; RD: 7.75 [4.17, 11.33]). Harm rates declined for proprietary flagship (10.0% vs 22.5%; RD: -12.50 [-16.86, -8.14]), proprietary fast (16.0% vs 31.8%; RD: -15.75 [-21.49, -10.01]), and open-source councils (22.0% vs 38.5%; RD: -16.50 [-22.27, -10.73]). Coverage analysis revealed net positive gains for accuracy (ΔCoverage: 4.4-9.8 percentage points) and safety (ΔCoverage: 13.6-20.6), indicating councils recovered correct diagnoses and averted harm. Councils elevated correct diagnoses to higher rank positions; and produced more complete differentials and management plans (all P<.05). Harmful council responses showed reduced combined commission-and-omission errors and tended to be less severe. Structured deliberation via multi-agent LLM councils may enhance the reliability of LLM-assisted ophthalmic clinical reasoning.
title Deliberative multi-agent large language models improve clinical reasoning in ophthalmology
topic Computers and Society
url https://arxiv.org/abs/2603.21447