This week Mark McClellan, the former head of the FDA, said insurance data and electronic health records (EHRs) should be mined for insights into drug safety. His statement is unusual by itself. In the past, ex-FDA bosses tended to be circumspect. One can only surmise that the Bush administration and McClellan, its loyal soldier, have decided that a newly assertive approach is warranted. Could McClellan’s statement foretell some new federal safety effort?

Here’s what he said about the drug-safety data iceberg, according to the AP:

“It’s the 21st century, and we have the potential to bring that whole iceberg out of the water—or at least see it,” McClellan said. He estimated that data on about 100 million Americans rests in EHR databases of various types. Had it been analyzed, he said, the truth about Vioxx might have surfaced earlier. McClellan’s statement is a welcome endorsement of using pharmacovigilance software and systems. It’s hard to disagree with the man.

On another level, however, McClellan’s statement is mystifying.

First of all, there should be a moratorium on anyone describing simple ways society might have avoided the whole Vioxx mess. Doing so hopelessly oversimplifies the science. As our own house plants complained to us the other day, even they claimed to know all about the risks from Vioxx, way back in 1999. Our sun-loving spider plant, in particular, insists there was just something in the zeitgeist that enabled it to detect the safety signal on Vioxx before everyone else. Maybe.

And yet the California health management organization Kaiser-Permanente did peer into its database to learn things about Vioxx before the medical journals went to town. Mentioning that could have highlighted the minimal role the FDA played in analyzing and withdrawing Vioxx from the market.

Mild Quibbling

What’s more, now that we think of it, who was running the FDA during the Vioxx period? Dr. McClellan. The failure of the FDA to claim responsibility for drug safety is one reason that the public estimation of the agency is dropping. Yes, the whistleblowers at FDA have made plenty of noise. But if the FDA image is to be restored, it will take a top person at the agency to claim responsibility for the agency’s safety decisions.

It’s strange for McClellan to suggest that data-mining is some sort of futuristic, Jetson-y activity that can only be practiced in 2020 once the FDA gets more sophisticated computers. Data-mining is common in many industries. Specialized tools exist to do it in the life sciences.

When did McClellan figure out data-mining might be useful? Are any current FDA managers willing to say the same thing? Or are they just keeping their heads down? Is there a specific plan to mine safety data beyond the delayed and troubled Janus project, or is this just hand-waving? The critical question is not whether data-mining is good. Of course it is. The real issue is whether a cash-strapped Congress can divert sufficient resources to make an overdue fix at the FDA: truly modernize its applications and databases, and expedite (or kill) Janus as part of that.

Shattered Privacy

There is a third problem with McClellan’s statement. He’s assuming public support. That is not to be taken for granted in the present climate.

Expectations of privacy in our society have dropped precipitously, as has the willingness to allow more usage of sensitive personal data. How often have big companies and government agencies “lost” thousands of social security and credit card numbers? Once a week? They say “oops,” but no one is ever punished. It’s alarming. Identity theft is rampant. HIPAA or no HIPAA, the mining of electronic health records has to be marketed and executed skillfully.

ClinPage was astounded to read that when the cardiac pioneer Michael E. DeBakey needed surgery himself in 2005, his doctors didn’t sign DeBakey in under his own name. They knew medical privacy no longer exists. So (the New York Times reported) they used an alias instead. If renowned physicians don’t trust the system, why should ordinary mortals?

Non-VIP patients are also withholding information from doctors for a variety of reimbursement- and privacy-related reasons. In this climate, data-mining efforts could make the perception of the pharmaceutical industry and the FDA much worse. At the risk of stating the obvious, something of a core competency here at ClinPage, sifting through massive quantities of personal health data on millions of Americans needs to be done by people who know what they’re doing—and can quickly convince the public of same.

Skillful Messaging?

The bottom line? Better drug safety technology will require surmounting both public relations and technical hurdles. Whether the FDA can handle all that is open to question. They’re speaking in public more of late, and that’s a good sign.

Outside the FDA, Oracle could be sufficiently media-savvy and tough to take up the messaging challenge. But unfortunately many other organizations in the life sciences are too tongue-tied and uncertain to say anything that will advance drug safety technology and the uses to which it should be put.