Disproportionality Analysis for Pharmacovigilance Signal Detection in Small Databases or Subsets: Recommendations for...

Area of Science:

• Pharmacovigilance and Drug Safety
• Biostatistics
• Regulatory Science

Background:

• Individual case reports are crucial for pharmacovigilance and identifying drug safety signals.
• Disproportionality analysis is widely used for signal detection, but lacks clear guidelines on minimum database size for robust results.

Purpose of the Study:

• To investigate the impact of database size on the robustness of disproportionality analysis.
• To determine minimum database sizes to limit spurious associations in pharmacovigilance signal detection.

Main Methods:

• Created random, country-specific, and ATC-classified subsets from VigiBase.
• Calculated spuriousness rates by comparing disproportionality analysis results in original and permuted data.
• Clinically assessed drug-event combinations for labelledness at various database sizes.

Main Results:

• Spuriousness rates were generally low (<0.15) across most countries and subset sizes.
• Random subsets of 3000-5000 or more reports showed spuriousness rates close to 0.10.
• Clinical assessment confirmed a high proportion (87-100%) of labelled or known drug-event combinations.

Conclusions:

• A minimum of 500 reports is recommended for national pharmacovigilance databases.
• A minimum of 5000 reports is recommended for non-country-specific databases or subsets.
• Sensitivity is expected to be limited in smaller databases.