Predictive analytics—lumped under the faddish banner of Big Data—is the high-profile set of techniques that tame numeric deluges to deduce that a new epidemic is starting to break or that a last-minute steal of an airfare has just popped up. The best uses for Big D may be yet to come, though.
The FDA just issued a solicitation document, a request for quotation, seeking a contractor to comb drug makers' expenditures, everything from phone calls to educational promotional materials, that would ferret out instances when companies have been peddling drugs for uses that have not received regulatory approval. The contractor that wins approval would furnish this information in real-time to the FDA's Office of Criminal Investigations. There's a latent demand for these services, as the FDA, even without the ability to marshal the desired level of analysis, has discovered some whoppers in recent years: GlaxoSmithKline had to pay $3 billion for pushing drugs off-label: the antidepressant Wellbutrin for weight loss and sexual dysfunction is one example.
This is where Big Data gets interesting.