Signal detection of methylphenidate by comparing a spontaneous reporting database with a claims database

Data mining is critical for signal detection in pharmacovigilance systems. In this study, we compared signals between spontaneous reporting data and health insurance claims data for a socially issued drug, methylphenidate. We implemented data-mining tools for signal detection in both databases: Reporting Odds Ratios (ROR), Proportional Reporting Ratios (PRR), Chi-squared test, and Information Component (IC), in addition to a Relative Risk (RR) tool in the claims database. The claims database generated 15, 15, 36, 1, and 1 adverse drug reactions (ADRs) by ROR, PRR, chi-square, IC, and RR, respectively. The World Health Organization (WHO) spontaneous database generated 91, 91, 137, and 96 ADRs by ROR, PRR, chi-square, and IC, respectively. We found seven potential matching associations from the claims and WHO databases, but only one of them was present in the Korean spontaneous reporting database. In Korea, spontaneous reporting is still underreported and there is a small amount of data for Koreans. Signal comparison between the claims and WHO databases can provide additional regulatory insight.