Selected control events and reporting odds ratio in signal detection methodology

Area of Science:

• Pharmacovigilance
• Drug Safety Signal Detection
• Regulatory Science

Background:

• Traditional pharmacovigilance relies on spontaneous reporting systems (SRs) to detect adverse drug events.
• Identifying less frequent or masked adverse events can be challenging with standard methods.

Purpose of the Study:

• To evaluate the utility of the reporting odds ratio (ROR) using "control events" for detecting drug safety signals.
• To assess if this method can identify signals obscured by more common adverse events.

Main Methods:

• Utilized drug use investigations (DUIs) from 1970-1998 and spontaneous reports (SRs) from 1998-2008.
• Converted event terms to Medical Dictionary for Regulatory Activities (MedDRA) preferred terms.
• Calculated ROR using 20 selected "control events" with stable incidence proportions and compared ROR with risk ratios (RR) from DUIs.

Main Results:

• "Control events" constituted 18.2% of all reports.
• ROR with "control events" identified some hidden signals, particularly for drugs with lower-than-average "control event" proportions.
• The median ratio of ROR (control events) to RR was approximately one, suggesting "control events" generally reflect exposure distribution, though individual RORs varied significantly.

Conclusions:

• The ROR method incorporating "control events" serves as a valuable adjunct to traditional signal detection techniques.
• This approach aids in discovering safety signals that might be masked by more prominent adverse event reports.