alexcpn/fda-faers-parquet

--- license: cc0-1.0 task_categories: - tabular-classification tags: - pharmacovigilance - drug-safety - fda - faers - adverse-events - signal-detection pretty_name: FDA FAERS Adverse Event Reports (Parquet) --- # FDA FAERS Adverse Event Reports Pre-processed [FDA Adverse Event Reporting System (FAERS)](https://www.fda.gov/drugs/fdas-adverse-event-reporting-system-faers) data in Parquet format for pharmacovigilance signal detection. ## Tables | Table | Description | |-------|------------| | DEMO | Demographics and case admin (deduplicated: latest CASEVERSION per CASEID) | | DRUG | Drug/medication info (drugname, prod_ai, role_cod, route, dose) | | REAC | Adverse reactions (MedDRA preferred terms) | | OUTC | Patient outcomes (death, hospitalization, disability, etc.) | | RPSR | Report sources | | THER | Therapy dates | | INDI | Indications for use | All tables linked by `primaryid`. Drug role codes: PS=Primary Suspect, SS=Secondary Suspect, C=Concomitant, I=Interacting. ## Current Coverage - **2024 Q4** (410,849 unique cases) ## Usage with DuckDB ```python import duckdb conn = duckdb.connect() conn.install_extension("httpfs") conn.load_extension("httpfs") demo = conn.execute(""" SELECT COUNT(*) as total_cases FROM read_parquet("hf://datasets/alexcpn/fda-faers-parquet/data/DEMO.parquet") """).fetchdf() ``` ## Disclaimer FAERS is a voluntary reporting system. Report counts do not establish causation. Disproportionality signals (PRR, ROR) indicate statistical associations that require clinical validation. Not intended for direct clinical decision-making. ## Source [FDA FAERS Quarterly Data Extract Files](https://fis.fda.gov/extensions/FPD-QDE-FAERs/FPD-QDE-FAERs.html)