Data preprocessing scripts and standardized prompts used in the study: Enabling interoperability across disparate health data sources using Large Language Models

This ZIP file contains: (1) the exact standardized prompt submitted to all three LLMs across all 36 runs (3 LLMs × 4 schemas × 3 repetitions), (2) the complete set of source-to-target mapping pairs used for preprocessing and standardizing CDM element names across PCORnet, OMOP, Sentinel, and i2b2, and (3) the complete mapping results from the three LLMs for all four Common Data Models (PCORnet, OMOP, Sentinel, i2b2) to FHIR across iterations.