Operating Manual

A regulatory-grade signal-detection instrument for spontaneous adverse-event data.

PRISM applies the four canonical disproportionality measures — Reporting Odds Ratio, Proportional Reporting Ratio, Pearson χ², and the Bayesian Information Component — to the FDA Adverse Event Monitoring System (AEMS, formerly FAERS) via the openFDA API. Outputs follow READUS-PV guidance and are intended for pharmacovigilance research, hypothesis generation, and safety surveillance.

Caution Disproportionality analysis identifies statistical reporting patterns in spontaneous adverse event data. It does not establish causality, incidence rates, or comparative drug safety. AEMS contains adverse event reports, medication error reports, and product quality complaints voluntarily submitted to the FDA. Detected signals require further clinical evaluation, biological-plausibility assessment, and real-world evidence validation. This tool is not intended to inform individual patient treatment decisions.

Getting Started

Three query modes, one analytical core

MODE 01 · SINGLE

Full drug profile

Scan the top-N most frequently reported preferred terms for a drug. Returns ROR / PRR / χ² / IC, a forest plot of top signals, demographics, and temporal trends.

MODE 02 · COMPARE

Head-to-head

Run two drugs against the same AEMS background and inspect shared and unique events side-by-side. Useful within therapeutic class — interpret with caution.

MODE 03 · REVERSE

ADR → Drug

Rank the drugs most frequently reported as primary suspect for a specified adverse event. Supports exact MedDRA Preferred Terms or broader substring search.