Versioned Crosswalks for Harmonizing Public Administrative Hospital Discharge Data

This repository contains the harmonized, record-level hospital discharge dataset produced by a crosswalk pipeline. It provides the output data only — pipeline code and crosswalk generation logic are maintained separately. Coverage The dataset harmonizes public hospital discharge records from five jurisdictions: Korea (NHIS), Brazil (DATASUS/SIHSUS), Mexico (SAEH/SSA), Chile (DEIS-MINSAL), and New York State (SPARCS), spanning 2001–2025 and totaling approximately 562 million source records. Structure Each source dataset is mapped to a common 14-field data model, with a CNTRY_CD column (values: kr, br, mx, cl) identifying the country of origin. Files are organized by source and year, following the versioned crosswalk (DOI: https://doi.org/10.5281/zenodo.20812649).