Ascent: An Agentic System over the Model Context Protocol for Real-World Clinical Data Analysis
On the map: LLM agents and tool use → Medical and scientific domain agents
Abstract (arXiv)
Answering epidemiological questions from real-world clinical data requires medical coding, schema-aware SQL, and validation of implicit choices about populations, denominators, and time. We present Ascent, an agentic system that exposes medical coding, question answering, and cohort analysis through a shared Model Context Protocol tool surface for standardized and native schemas. We introduce EpiTrap, a dataset testing whether systems avoid recognized pharmacoepidemiological errors, and compare a fixed pipeline with agents across models and orchestrators. With capable models, agents improve accuracy over the fixed pipeline by an average of 27 and 20 percentage points on native and standardized schemas, respectively. These gains require more tool calls and longer runtimes. Experience from real projects highlights the system's value for feasibility assessment, diagnostic iteration, and expert-guided analysis.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical and scientific domain agents, one of 16 directions in LLM agents and tool use. Builds on 17 of its 51 references that are in the map. First posted on arXiv 21 Sep 2026.
Builds on
17 of 51 references.
Closest ancestors · 4
Same direction, newest first
- 2026CodeClinic: Evaluating Automation of Coding Skills for Clinical Reasoning Agents10 May
- 2025From Conversation to Query Execution: Benchmarking User and Tool Interactions for EHR Database Agents27 Sep
- 2025Infherno: End-to-end Agent-based FHIR Resource Synthesis from Free-form Clinical Notes16 Jul
- 2024ReflecTool: Towards Reflection-Aware Tool-Augmented Clinical Agents23 Oct
Other references · 13
- 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
- 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 965
- 2026DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligencecited by 877
- 2023C-Pack: Packed Resources For General Chinese Embeddingscited by 556
- 2020Self-Alignment Pretraining for Biomedical Entity Representationscited by 233
- 2024When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMscited by 230
- 2024A Preview of XiYan-SQL: A Multi-Generator Ensemble Framework for Text-to-SQLcited by 46
- 2025XiYan-SQL: A Novel Multi-Generator Framework For Text-to-SQLcited by 40
- 2024Retrieval augmented text-to-SQL generation for epidemiological question answering using electronic health recordscited by 8
- 2025EHR-MCP: Real-world Evaluation of Clinical Information Retrieval by Large Language Models via Model Context Protocolcited by 5
- 2025Generating patient cohorts from electronic health records using two-step retrieval-augmented text-to-SQL generationcited by 4
- 2026Patient-Similarity Cohort Reasoning in Clinical Text-to-SQLcited by 3
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026ClinLens: Towards Long-Horizon Coding Agents for Longitudinal Multimodal Clinical Data Science
- 2025Agentic Medical Knowledge Graphs Enhance Medical Question Answering: Bridging the Gap Between LLMs and Evolving Medical Knowledge
- 2026Can LLM Agents Generate Real-World Evidence? Evaluating Observational Studies in Medical Databases
- 2026EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering and Reasoning
- 2026A knowledge-guided agentic framework for mitigating patient-context ambiguity in health queries
- 2026FastOMOP: A Foundational Architecture for Reliable Agentic Real-World Evidence Generation on OMOP CDM data
- 2025MedBrowseComp: Benchmarking Medical Deep Research and Computer Use
- 2025FHIR-AgentBench: Benchmarking LLM Agents for Realistic Interoperable EHR Question Answering