KMID : 1011320110040010015
|
|
Journal of Pharmacoepidemiology and Risk Management 2011 Volume.4 No. 1 p.15 ~ p.21
|
|
Development of Signal Detection Program for Adverse Drug Reactions
|
|
Seong Jong-Mi
Choi Nam-Kyong Ahn So-Hyeon Park Byung-Joo
|
|
Abstract
|
|
|
Objective: There is a growing need for detecting signals of adverse drug reaction (ADR) from large automated databases efficiently and effectively. We developed a data-mining program that performs signal detection of ADR automatically.
Methods: This software was developed with Microsoft Studio 2008 and Microsoft SQL 2005 Express. Adverse events (AEs) database in Microsoft Excel included information on suspected drug names and AEs were imported to the software for data-mining analysis. The algorithms included proportional reporting ratio, reporting odds ratio and information component by Bayesian Confidence Propagation Neural Network. The results of signal detection were displayed as a tabular format. Its feasibility and capacity for signal detection was tested with ADR reports collected from 24 June 2009 to 28 February 2010 by PharmacoVigilance Research Network.
Results: We developed the automated system that can carry out data import and analysis to detect drug-ADR signals. User-friendly interfaces were developed, which allowed users to easily view and analyze AES database. Using this program 10,352 drug-AE pairs were detected as signals among total 17,482 reports and 59,397 drug-AE pairs.
Conclusion: This automated data-mining system can be a promising tool for detecting drug-ADR signals using AEs database, which can be useful to the healthcare policy makers for making decision on pharmacovigilance and also to the physicians and pharmacists for their patients care.
|
|
KEYWORD
|
|
Signal, Adverse drug reaction, Adverse event, Pharmacovigilance
|
|
FullTexts / Linksout information
|
|
|
|
Listed journal information
|
|
|