Analysis code and result data for "Non-specific between-drug differences in the FDA Adverse Event Reporting System are large and ubiquitous: an empirical-null audit of 26 drug pairs"

This record contains the complete analysis pipeline and every result file underlying the study"Non-specific between-drug differences in the FDA Adverse Event Reporting System are large andubiquitous: an empirical-null audit of 26 drug pairs". Contents: S_code/ (47 Python scripts and a pinned requirements file, implementing FAERS ingest anddeduplication, empirical-null panel construction, random-effects and false-discovery-rate testing,bootstrap calibration, time-to-onset analysis, and the JADER replication); S_data/ (48 CSV tables ofpanel estimates, formal test results, sensitivity analyses and JADER replication outputs); README.md(re-analysis guide); and REPLICATION_PROTOCOL.md (the pre-specified JADER replication plan writtenbefore the data were accessed). The underlying source data are public and are not redistributed here: the FDA FAERS quarterly extracts(faers_ascii_2016q1.zip through faers_ascii_2026q2.zip, downloaded 2026-09-17) fromhttps://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html, and the PMDA JADER releasepmdacasereport202609 (1,054,239 cases) from https://www.info.pmda.go.jp/fukusayoudb/CsvDownload.jsp. Code is released under the MIT License; the CSV result tables are released under CC-BY 4.0.