Automatic drug adverse reaction discovery from parenting websites using disproportionality methods

作者：Hadzi Puric Jelena^*; Grmusa Jeca

来源：IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, 2012-08-26 to 2012-08-29.

DOI：10.1109/ASONAM.2012.144

• drug adverse reactions
• text-mining
• pediatric health care
• disproportionality analysis

The pediatric fact that children "are not just little adults" reflects special drug class, dosing, pharmacokinetic parameters that must be considered when planning safe and effective health interventions, because evidence of drug safety and efficiency cannot be extrapolated from adults to children. There is significant interest in utilization of observational health care data in pediatric drug safety research. We gathered data from eight different parenting websites in order to explore automatic drug adverse reaction discovery. In recent years, Web forums offer an opportunity for parents and doctors to share healthcare information regarding prescription drugs. Our aim is to discover drugs and adverse reactions and quantify their association using statistical methods based on different measures of disproportionality. Also, applied disproportionality methods have the potential to scale well to large databases such ours.

• 出版日期2012