Conference Paper

Comparing SQL and Graph Database Query Methods for Answering Clinical Trial Questions with LLM-Powered Pipelines

Jaime Yan

PharmaSUG 2025 · 2025 · SI-342 · sole author

Abstract

Should clinical data exploration use SQL or graph databases? This paper benchmarks both query paradigms behind LLM-powered natural-language pipelines on 150 representative clinical trial questions over synthetic ADaM data, comparing accuracy, latency, and the classes of questions each paradigm handles well. The proposed RagQL-Nav framework — query decomposition, intelligent routing, and dual-query validation — reached 91% accuracy on complex queries, roughly 12 points above single-system approaches, offering practical guidance on when graph approaches earn their complexity.

Keywords

SQL · graph database · LLM pipeline · benchmark · clinical data · ADaM · query routing

Cite

Jaime Yan. "Comparing SQL and Graph Database Query Methods for Answering Clinical Trial Questions with LLM-Powered Pipelines." PharmaSUG 2025, 2025. doi:10.5281/zenodo.22182920.