ABSTRACT

Some kinds of data are collected as free text but must eventually be grouped together into like terms for review and analysis. The most common examples of this kind of data are adverse events, medications, and medical history. In all of these cases, the investigator reports a term with little or no guidance on terminology or wording. Yet, to make assessments of the safety of the drug, like terms must be counted or classi-—ed together. For example, the adverse event terms headache, mild headache, and aching head should all be counted as the same kind of event. Similarly, the drugs reported as Tylenol and acetaminophen should be classi—ed as the same drug.