Data Mining
Since the adverse events in AERS are contained in a database, they can be analyzed by a variety of methods including data mining. Data mining is a statistical technique to search large databases to detect associations that occur at higher than expected frequencies.
Data mining enables OSE to find and examine associations or patterns that would not otherwise be obvious. The goal is to detect something that occurs at a higher frequency than expected based on reports of the same event for other drugs in the database.
It is important to understand that the frame of reference here is the database. Data mining does not outline the intrinsic rate of an adverse event in the population. However, it can detect, for example, a rash that is reported more commonly with one drug than is reported relative to another drug in the database.