Applying Data Mining to Detection of Adverse Drug Reactions
Daisuke Koide, Kazuhiko Ohe · Studies in health technology and informatics · 2001
To detect adverse drug reactions (ADRs) from a hospital information system efficiently, data mining was applied to prescription and laboratory data. From several techniques, market basket analysis was done. The association of targeted drugs and tests, which appeared on the Emergency Safety Information (ESI), were easily identified. (GPT vs. Benzbromarone, and PLTD vs. Ticlopidine) Further evaluation requires determining ADRs, however, data mining has a possibility to facilitate drug safety monitoring.