A comparison of disproportionality analysis methods in national adverse drug reaction databases of China.

Objective: Several disproportionality analysis methods are widely used for signal detection. The goal of this study was to compare the concordance of the performance characteristics of these methods in spontaneous reporting system of China.
Methods: Algorithms including reporting odds ratio (ROR), proportional reporting ratio (PRR) and information component (IC), a composite criterion previously used by Medicines and Healthcare Products Regulatory Agency (MHRA) were compared. Kappa coefficient was used as the gauge to test the concordance. Reports received in the year 2004 and 2005 were extracted for analysis in this study.
Results: After data processing, 361,872 reports representing 52,769 combinations were analysed. The analysis generated 24,022, 22,646, 5637 and 5302 signals of disproportionality by PRR, ROR, MHRA and IC, respectively. The kappa coefficient increased with the threshold of number of drug-adverse drug reactions (ADR) combination, and the coefficient exceeded 0.7 when the number of suspected drug-ADR exceeded 2.
Conclusion: This study shows that different measures used are broadly comparable in spontaneous reporting system in China when two or more cases per combination have been collected.