ROR (Reporting Odds Ratio)
What is it?
ROR is the most straightforward of the disproportionality measures: it compares the odds of a specific event being reported for a given drug against the odds of that same event being reported for every other drug in the database. A ROR well above 1 means that pairing is showing up more than the rest of the database's background pattern would suggest.
Its simplicity is also its limitation. With very few reports, a ROR can swing sharply from one new case alone, which is exactly the scenario EBGM's Bayesian shrinkage was designed to smooth out. In practice, safety teams tend to look at both together rather than relying on ROR in isolation.
Drag the report count
Same rare drug-event pair, nothing else about it changes. Drag from 2 reports up to 50 and watch what ROR does purely because more evidence arrived.
Drag the slider to add reportsDashed line = the true rate. Dot = ROR as reports pile up.
n = 2 reports
Barely any evidence yet. This number is not to be trusted.
