Data mining methods for pharmaceutical safety measures

From a pharmacoepidemiological viewpoint

Abstract

During the post-approval period, hypotheses about potentially new adverse drug reactions (ADR) have traditionally emerged from passive surveillance systems based on spontaneous case reports of suspected adverse drug reactions. With signal detection by traditional (or manual) methods, quantitative methods used on the spontaneous reporting system databases were introduced since the end of 1990's in order to detect serious ADR early as possible. In FY 2009, the Pharmaceuticals and Medical Device Agency introduced the quantitative methods (data mining method) used on Japanese SRS database to safety measures operations. This presentation introduces the recent situation about signal detection of adverse drug reactions.