What’s Up With My Data, Doc?
I can’t find the original article on the IHT website, but there’s a great piece in today’s edition on how pharmaceutical companies push their drugs by funding — I would say bribing — doctors. It’s written by Daniel Carlat, who writes a blog and publishes the Carlat Psychiatry Report.
The most interesting part of the piece is on something called prescription data-mining, where data from pharmacists on prescriptions — what patients are given what medicines — are linked to the doctors prescribing said medicines. This allows pharmaceutical companies to target doctors and get them to push their drugs by paying them to make presentations to other doctors.
Carlat himself made $30,000 in a year doing this before he saw the light. He is now a major critic of the practice, and challenges in a recent blog post the absurd industry defense of the practice of prescription data-mining that it’s all about transparency:
Today, however (on a tip from PharmaGossip), I read the most absurd argument in its defense yet, reported in yesterday’s Philadelphia Inquirer. The reporter, Karl Stark, quoted Jody Fisher, Verispan’s vice president of product management, as saying: “Doctors are trying to create a special right of privacy. I can certainly appreciate where they’re coming from. But the way the world is going is toward increased transparency of information.”
“Transparency of information”! What a wonderful Web 2.0 buzz phrase!
Of course, I’m interested because you can see in it the power of data-mining. The original pharmacist data doesn’t include the doctors’ names, only their Drug Enforcement Agency registration numbers. It’s the American Medical Association that effectively reveals the doctors’ names to Big Pharma by licensing its file of U.S. physicians, allowing data-mining companies like IMS Health and Verispan to match the numbers with the names, Carlat writes in today’s IHT piece. The AMA makes millions of dollars in this process, by the way.
Are similar things being done with our Internet-based data? Is the anonymous becoming less anonymous? If it’s not being done now, assume it will be in the future. It’s a great example of how data aren’t always valuable until they’re linked to other data, and then they’re extremely valuable.