{"paper":{"title":"BLUEmed: Retrieval-Augmented Multi-Agent Debate for Clinical Error Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A multi-agent debate system augmented with hybrid retrieval detects terminology substitution errors in clinical notes more accurately than single-agent RAG or debate-only approaches.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hanshu Rao, Nguyen Anh Khoa Tran, Qiunan Zhang, Saukun Thika You, Wesley K. Marizane, Xiaolei Huang","submitted_at":"2026-04-12T00:30:31Z","abstract_excerpt":"Terminology substitution errors in clinical notes, where one medical term is replaced by a linguistically valid but clinically different term, pose a persistent challenge for automated error detection in healthcare. We introduce BLUEmed, a multi-agent debate framework augmented with hybrid Retrieval-Augmented Generation (RAG) that combines evidence-grounded reasoning with multi-perspective verification for clinical error detection. BLUEmed decomposes each clinical note into focused sub-queries, retrieves source-partitioned evidence through dense, sparse, and online retrieval, and assigns two d"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"BLUEmed achieves the best accuracy (69.13%), ROC-AUC (74.45%), and PR-AUC (72.44%) under few-shot prompting, outperforming both single-agent RAG and debate-only baselines.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The clinical terminology substitution detection benchmark used for evaluation is representative of real-world clinical notes and error distributions, and the two domain-expert agents possess sufficient clinical knowledge to produce reliable independent analyses without introducing new hallucinations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"BLUEmed combines hybrid RAG with structured multi-agent debate and a safety filter to detect terminology substitution errors in clinical notes, reaching 69.13% accuracy under few-shot prompting and outperforming single-agent and debate-only baselines.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A multi-agent debate system augmented with hybrid retrieval detects terminology substitution errors in clinical notes more accurately than single-agent RAG or debate-only approaches.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"9ff56a033121866695cfcfd2a53d0f67ff84fa77aa24c09d96cac172f4dc8db2"},"source":{"id":"2604.10389","kind":"arxiv","version":2},"verdict":{"id":"b2517862-decb-4aa0-9ecc-3be15226f03b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T16:40:22.999045Z","strongest_claim":"BLUEmed achieves the best accuracy (69.13%), ROC-AUC (74.45%), and PR-AUC (72.44%) under few-shot prompting, outperforming both single-agent RAG and debate-only baselines.","one_line_summary":"BLUEmed combines hybrid RAG with structured multi-agent debate and a safety filter to detect terminology substitution errors in clinical notes, reaching 69.13% accuracy under few-shot prompting and outperforming single-agent and debate-only baselines.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The clinical terminology substitution detection benchmark used for evaluation is representative of real-world clinical notes and error distributions, and the two domain-expert agents possess sufficient clinical knowledge to produce reliable independent analyses without introducing new hallucinations.","pith_extraction_headline":"A multi-agent debate system augmented with hybrid retrieval detects terminology substitution errors in clinical notes more accurately than single-agent RAG or debate-only approaches."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.10389/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":1,"snapshot_sha256":"30c4e653028e18e63ca8eb561435ae8b602bfdef229d0f3336392243d998afd0"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}