Evaluating Retrieval-Augmented Large Language Models on Anesthesiology Board-Style Questions: Benchmark Study

A Self-Controlled Benchmark of Retrieval-Augmented Generation for Large Language Models on Clinical Guideline Questions

Background/Objectives: Large language models (LLMs) show promise for clinical decision support, yet their accuracy in interpreting specialized medical guidelines remains uncertain. Retrieval-augmented generation (RAG) may enhance performance by grounding responses in authoritative knowledge bases. This study aimed to compare the accuracy, comprehensiveness, and safety of RAG-enhanced versus standard LLMs for answering clinical questions derived from the German S3 guideline for oral cavity carcinoma. Methods: We conducted a prospective, single-blind benchmark study evaluating six LLMs: one RAG-enhanced model (Custom GPT with guideline access), one consensus-based model (ConsensusGPT), and four standard models (DeepSeek-V3.2, Mistral Small 3.2, Qwen3-Next-80B, GPT-OSS-120B). Fifty clinical questions covering 17 guideline domains were presented to each model three times, yielding 900 evaluations. Three expert reviewers assessed responses using 5-point Likert scales for accuracy, comprehensiveness, and clarity, under a single-blind procedure, the effectiveness of which was tested by a pre-specified manipulation check. We then ran a paired within-model experiment in which each base model was queried with and without guideline access through a transparent, openly released retrieval pipeline, and scored every response with a condition-blind automated judge alongside deterministic retrieval metrics computed from the logs. Secondary outcomes included hallucination rates and guideline citation behavior. Inter-rater reliability was assessed using intraclass correlation coefficients (ICCs). Results: In a paired within-model design that held each base model fixed, adding transparent guideline retrieval improved accuracy—significantly in the three weaker open-weight models (Mistral, Qwen3, and GPT-OSS) and directionally in the already-strong DeepSeek and GPT-5 bases. Because a pre-specified blinding check found that experts could still identify retrieval-augmented answers with 98.5% accuracy, we anchored causal interpretation on measures that do not depend on the human raters, ranked by their independence: deterministic, log-derived retrieval metrics first, and then an automated, condition-blind LLM judge, whose agreement with the experts (Spearman ρ = 0.81, 95.7% within-one agreement) establishes shared calibration rather than independence from their bias. Deterministically from the retrieval logs, citation groundedness rose from 0% to 51–89% and retrieval recall@5 was 92%. On the judge, content-level hallucination fell from 42% to 4% and accuracy rose by a pooled +0.64 points (95% CI 0.47–0.80); the accuracy gain persisted after adjustment for response length (+0.48, 95% CI 0.22–0.73), which retrieval shortened rather than lengthened. The accuracy gain was large for weaker base models and small or non-significant for already-strong ones, whereas the hallucination and auditability gains were consistent across all models. The human ratings reproduced the judge’s accuracy effect (+0.61, 95% CI 0.49–0.74), and GPT-5 run through the transparent pipeline showed no significant difference from the proprietary Custom GPT (judge accuracy 4.48 vs. 4.58). Conclusions: Guideline retrieval yields a reproducible, largely base-independent improvement in the safety and auditability of LLM answers to clinical guideline questions, with accuracy gains concentrated in weaker base models. Because retrieval-augmented answers are recognizable to experts, rigorous evaluation should rely on rater-independent measures, and residual hallucination continues to require human oversight.

Large language models in spine care and research.

BACKGROUND Large language models (LLMs) have emerged as powerful transformer-based systems capable of capturing long-range dependencies and complex semantic relationships in clinical language. In this review paper, we first examine the technical foundations of medical LLMs, including transformer architecture, attention mechanisms, training paradigms, and retrieval-augmented generation. RESULTS We then survey documented applications in spine surgery and spinal care, highlighting moderate guideline concordance (46-67%) for diagnostic support, automated generation of operative notes and discharge summaries for administrative workflows, LLM-assisted literature review and manuscript drafting for research support (with ~68% novelty accuracy), and translation of complex surgical concepts into patient-friendly materials at a seventh-grade reading level. We next explore emerging multimodal models that integrate text, imaging, laboratory, and genomic data via cross-modal attention, demonstrating superior performance in holistic diagnostic and prognostic tasks. DISCUSSION Finally, we discuss key implementation challenges, including model accuracy and hallucinations; computational, privacy, and regulatory constraints under HIPAA/GDPR; and bias mitigation, to outline strategies for safe, effective, and equitable deployment. CONCLUSION By mapping technical capabilities to clinical and research use cases, this review highlights the promise of LLMs in enhancing decision support, workflow efficiency, research productivity, and patient communication in spine care, while emphasizing the need for interdisciplinary collaboration, robust evaluation metrics, and governance frameworks that prioritize patient safety and equity.

Retrieval-augmented generation for medical question answering: a multi-metric performance evaluation

The proposed framework offers a practical and scalable approach to mitigating hallucinations without requiring task-specific fine-tuning, highlighting the potential of retrieval-augmented approaches for trustworthy artificial intelligence (AI)-assisted healthcare applications.

Model and Task-Aware Test-Time Scaling Strategies for Large Language and Vision-Language Models in Medicine: Evaluation Study

Abstract Background Test-time scaling has emerged as a promising method to enhance the reasoning capabilities of large language models (LLMs) and vision-language models (VLMs) during inference without additional training. While foundational studies established scaling paradigms in general domains, their applicability to the unique complexities of medical AI remains underexplored. Objective This study aims to conduct a comprehensive investigation of test-time scaling in the medical domain. We evaluate the impact of scaling across different model sizes and task complexities. Furthermore, we seek to identify domain-specific bottlenecks and assess model robustness against user-driven perturbations, such as misleading clinical authority. Methods This study evaluated a diverse set of general and medical-specific LLMs and VLMs. Experiments used five textual medical benchmarks comprising over 5500 questions and two multimodal benchmarks comprising 7000 samples. Performance was measured under three scaling conditions: increasing token budgets, iterative sequential scaling, and parallel scaling. Robustness was tested by embedding misleading hints with varying tones and levels of simulated clinical expertise into prompts. Results For nonreasoning LLMs, accuracy saturated quickly, with token usage often remaining under 500 tokens regardless of budget increases. Reasoning models demonstrated significant performance gains on complex tasks as token budgets increased. Notably, we identified distinct domain-specific behaviors. First, current VLMs showed a structural bottleneck in integrating visual clues and experienced limited benefit from token expansion. Second, medically fine-tuned LLMs excelled in clinical question answering but exhibited degraded scaling efficiency on calculation tasks compared to general-domain models. This reflects a disparity between qualitative clinical alignment and procedural logic. Third, while optimal scaling improved robustness, models exhibited a cognitive vulnerability by readily abandoning correct reasoning when confronted with misleading expert physician hints. Regarding scaling strategies, parallel scaling outperformed sequential scaling on easier tasks. Conversely, extended sequential scaling or increased budgets proved essential for complex problem-solving. Conclusions Test-time scaling rules from general domains do not perfectly translate to medical AI. Longer reasoning is not universally beneficial. Concise reasoning with parallel scaling is optimal for simpler tasks. An extended chain of thought via sequential scaling or increased budgets is required for complex problems. Furthermore, safe clinical deployment requires addressing fundamental vision-language alignment, balancing clinical and procedural reasoning, and mitigating vulnerabilities to perceived clinical authority.

Small but Specialized: A Domain-Adapted Retrieval-Augmented LLM Outperforms Frontier Generalists in Pediatric and Adolescent Gynecology

MedLLM: An Open Medical Language Model at the Sub-Billion Scale

Across medical benchmarks, MedLLM shows a pattern visible only at sub-billion scale: medical competence does not degrade uniformly under compression but splits by task type and dissociation is masked at 7B, where both capabilities are present, and surfaces only when capacity is scarce.