Drug-drug interactions in informatics

• Martin

  English

  Author

Public health is increasingly at risk from drug-drug interactions (DDIs). According to recent 
estimates, DDIs result in approximately 74,000 ER visits and 195,000 hospital admissions 
annually in the United States. Phase IV clinical trials and post-marketing surveillance are examples 
of current DDI discovery methods that are inadequate for identifying many DDIs and do not notify 
the public of potentially harmful DDIs prior to a drug's release onto the market. Modern statistical 
and computational techniques have been used to the DDI problem in recent work. Here, we 
examine recent advancements that cover a variety of informatics techniques in this field, from 
building databases for effective searching of known DDIs to predicting new DDIs using 
information from electronic medical records, adverse event reports, scientific abstracts, and other 
sources. We also discuss the reasons for the difficulty of detecting DDIs and the prospects for 
informatics-based methods of DDI discovery. 

• pdf