The paper introduces LogiMed‑RoB, a benchmark that tests large language models (LLMs) on hierarchical logical consistency in medical risk‑of‑bias assessments, using 860 randomized controlled trials and 14,820 queries based on Cochrane RoB 2.0 expert logic. It evaluates models across four dimensions—Atomic Consistency, Domain Consistency, Aggregation Consistency, and Evidential Faithfulness—revealing a severe Error Compounding Effect where high atomic accuracy does not translate to end‑to‑end consistency. Experiments on ten state‑of‑the‑art LLMs show that even top performers can collapse to 45.13% overall consistency, with some models nearly failing entirely, and that many models struggle to deduce correct outcomes from retrieved evidence.
The paper introduces three interpretable metrics—Explanation Stability Index (ESI), Causal Faithfulness Score (CFS), and Perturbation Stability Score (PSS)—to assess the faithfulness of large language models (LLMs) in clinical decision support. Evaluating six LLMs on 150 MedQA-USMLE questions, the study finds that only 23.3% of cited clinical concepts are causally necessary, and that correct answers often have lower CFS than incorrect ones, with answer consistency negatively predicting faithfulness. These findings suggest that accuracy, consistency, and consensus are incomplete safety signals for clinical decision-making support.
By Bharath Kumar Bolla, Chaitali Deb Purkayastha, Vishnu Surya Reddy Nandi
arXiv:2608.10725v2 Announce Type: replace
Abstract: Large language models (LLMs) often rely on shortcuts rather than systematic reasoning, raising safety concerns in medical applications. Allowing mo...
By Uma Ranjan, Kunal Tilaganji, Aditya Koul, Anurag Mahipal, Dashpreet Singh, Hriday Rana, Manan Jain, Sidharth Gupta, Ajo Babu George, Vineeth Balasubramanian, Nagarajan Natarajan, Amit Sharma
The paper investigates why retrieval‑based open‑ended evaluation fails in medical fact verification. By creating two detailed taxonomies—one for retrieval‑stage errors across five quality dimensions and another for verifier‑reasoning errors across six steps—the authors automatically label evidence quality and reasoning errors using an LLM‑as‑Judge pipeline. Their large‑scale stress tests across multiple retrieval methods and verifier models show that increasing model size, reasoning effort, source breadth, or medical fine‑tuning does not eliminate these failure modes, indicating fundamental limits of the retrieve‑then‑verify paradigm in open‑ended medical contexts.
By Heyuan Huang, Jirui Dai, Alexandra DeLucia, Sonal Joshi, Mahsa Yarmohammadi, Jie Gao, Bernal Jim\'enez Guti\'errez, Mark Dredze
arXiv:2607. 10275v1 Announce Type: new Abstract: Large language models achieve high scores on medical knowledge assessments, yet clinical reasoning requires actively deciding what to investigate under uncertainty.
By Krischan Braitsch, Laura K. Schmalbrock, Theresa Weltermann, Andrew F. Berdel, Isabella Miller, Kai Tran, Michael Heider, Sabrina Kraus, Florian Bassermann, Jacqueline Lammert, Sebastian Ziegelmayer, Marcus Makowski, Lisa C. Adams, Keno K. Bressem
arXiv:2604. 14892v3 Announce Type: replace-cross Abstract: Evaluating medical AI systems using expert clinician panels is costly and slow, motivating the use of large language models (LLMs) as alternative adjudicators.
By Amy Rouillard, Sitwala Mundia, Linda Camara, Ziyaad Dangor, Michael Cameron Gramanie, Ismail Kalla, Shabir A. Madhi, Kajal Morar, Marlvin T. Ncube, Haroon Saloojee, Bruce A. Bassett
The paper introduces BAR, a Budget‑Aware LLM Reasoning framework that enhances post‑discharge risk prediction by integrating external medical knowledge graphs (KGs). BAR refines KGs into disease‑specific evidence graphs with support scores and provenance, then uses an LLM to plan, navigate, and verify evidence within a patient‑specific budget. Experiments on MIMIC‑III and MIMIC‑IV across eight diseases show BAR improves AUPRC by 3.4 points, raises citation precision from 59.8% to 77.9%, and uses only 62‑65% of the allotted budget.
By Chen Chen, Dongjie Wang, Mei Liu, Zijun Yao
arXiv:2608.21948v1 Announce Type: new
Abstract: Complex clinical reasoning requires models to update diagnostic hypotheses as new evidence emerges and to coordinate different medical specialities und...
By Sike Xiang, Shuang Chen, Qian sun, Jia Cheng, Yusi Wei, Amir Atapour-Abarghouei
arXiv:2607. 08038v1 Announce Type: new Abstract: Diagnostic error is a major threat to patient safety, yet current large language model (LLM) systems often treat diagnosis as a one-shot prediction task, lacking safeguards against missed high-risk alternatives or rigorous verification of their reasoning.
By Fan Ma, Mauro Giuffr\`e, Donald Wright, Kent McCann, Mark Iscoe, Lingfei Qian, Mingyang Jiang, Chi Wing Ng, Na Hong, Huan He, Cathy Shyr, Qingyu Chen, Lee Schwamm, Lucila Ohno-Machado, Hua Xu
arXiv:2607. 28677v1 Announce Type: new Abstract: LLM now pass medical licensing examinations and, in curated cases, can rival physicians at diagnostic reasoning.
By Shayndhan Sivanathan, Shravan Nageswaran, Mehdi Zadem, Ryaan Sultan, Nicolas von Mallinckrodt, Max Solovyev, Alexey Matyushkin, Sumon Sadhu, Gabriele C DeLuca, Sanjeeva Jeyaretna, James Hillis, Manoj Ramachandran, Prakash Jayakumar
arXiv:2606. 05463v1 Announce Type: new Abstract: Patient safety event triage, determining whether a clinical event is reportable under jurisdiction-specific policy, is a high-stakes task typically performed manually by patient safety experts.
By Keqi Han, Ryan Young, Annabel Strauss, Lindsey Hughes, Katharine M. Nesbitt, Nicole Schueler, Che Ngufor, Carl Yang, Yuan Xue, Zhijun Yin
The article reviews existing tools for assessing clinical reasoning in large language models, identifying six key dimensions—problem representation, temporal synthesis, differential and management reasoning, counterfactual reasoning, calibrated uncertainty, and reasoning faithfulness. It finds that no single instrument covers all dimensions, with some aspects reasonably addressed while others, especially faithfulness and counterfactual reasoning, remain weakly evaluated. The authors propose combining current tools through binary rubric items, separate completeness and correctness scores, importance weighting, safety caps, temporal consistency checks, and chance‑corrected reliability reporting, and call for further design work on uncertainty, counterfactual reasoning, and faithfulness in longitudinal free‑text contexts.
By Zhangshu Joshua Jiang, Zina Ibrahim, James T. Teo