At present, all biosurveillance systems that monitor OTC sales data monitor category data, not individual products. There are at least 9000 individual products intended for self-treatment of the symptoms of infectious disease, and many distinctions among these products, such as flavor or bottle size, are not relevant to biosurveillance. Aggregating products with similar uses into analytic categories reduces variability in sales data due to promotions and other causes.
A prerequisite to monitoring of sales at the product category level is a set of categories. The GTIN standard facilitates the creation and maintenance of product categories. A product-category mapping is an assignment of each GTIN code to a category.
How to best aggregate GTINs into categories for biosurveillance is, ultimately, a research question. Studies about the informational value of OTC sales data, such as the ones we have already discussed, are the best way to address this question. One study, however, grouped products into categories with a statistical analysis of historical OTC data sales alone (Magruder et al., 2004). A limitation of this approach is that it finds some associations between sub categories that are coincidental. It also does not provide information about which categories are best to monitor for which disease outbreaks.
Was this article helpful?