I remember hearing opposing experts in the Purdue v. Endo patent infringement trial[Oxycontin] argue their respective positions re the accuracy of IMS data. Another data source, supposedly more accurate, was mentioned by both experts, but I no longer recall its name.