Ascent system improves clinical data analysis with smarter agents
Ascent: An Agentic System over the Model Context Protocol for Real-World Clinical Data Analysis
Artificial IntelligenceInformation Retrieval
Summary
Understanding medical data to answer health questions is tricky because it involves coding diseases, writing careful database queries, and making sure the right patient groups and timeframes are used. The authors created Ascent, a tool that helps experts with coding and analyzing medical data using a common interface. They also made a test dataset called EpiTrap that checks if systems avoid common mistakes. Their experiments show that using intelligent agents with Ascent improves accuracy a lot but takes more time and effort. This system helps with checking if studies are possible, finding mistakes, and guiding experts in real medical research.
What this means in practice
- •For hospital data teams: Improve accuracy and flexibility in clinical data queries by replacing fixed pipelines with agent-based analysis over native and standardized medical schemas.
- •For health insurance analysts: Use Ascent’s protocol to validate and refine population definitions in claims data studies for better epidemiological insights.
Authors
Angelo Ziletti, Leonardo D'Ambrosi, Melanie Tuchardt, Tim Kondziella
Abstract
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.