Definition

A statistical method comparing the observed frequency of a drug-event combination against the expected frequency based on overall reporting patterns in a database.

Example

PRR (Proportional Reporting Ratio) of 3.5 for Drug X and hepatitis means this combination is reported 3.5 times more frequently than expected based on database patterns.

Regulatory source

EU GVP Module IX, Pharmacoepidemiology

Common mistakes

Interpreting disproportionality as proof of causation. It indicates a statistical signal requiring clinical evaluation, not confirmed causality.

Inspector question

What disproportionality methods do you use? How do you set thresholds for signal detection?

Related terms