Screening of ICSR databases for statistical associations between products and events whereby the observed frequency is different than the expected frequency. For ADRs caused by one drug (called drug-event pairs), common measures of disproportionality are:
- Frequentist statistics
- PRR (Proportional Reporting Ratio)
- ROR (Reporting Odds Ratio)
- Bayesian statistics
- IC (Information Component)
- EGBM (Empirical Bayes Geometrical Mean).
For ADRs caused by the interaction of two drugs (called drug-drug-event triplets), the Omega (Ω) measure of disproportionality can be used.