Last fall, the FDA got more power and funding to monitor drugs once they're on the market. Last month, Commissioner Andrew von Eschenbach promised Congress big news on that monitoring, a project that since has been dubbed "Sentinel Initiative." Yesterday, the agency delivered: FDA is teaming up with the Center for Medicare and Medicaid Services on data-mining. FDA will be able to probe CMS's patient databases, actively looking for potential safety problems. Next, the data-mining will move into the private arena via deals with insurance companies; already, the FDA announced a similar partnership with WellPoint.
As you know, the FDA's current after-marketing surveillance is haphazard at best. The agency relies on doctors, drugmakers, and patients to report suspected side effects, but often, doctors and patients don't recognize a problem as a drug side effect or simply don't report their suspicions. Since 2005, there's been talk of some sort of data-mining operation, but it's only now coming to fruition.
Not that scrutinizing databases will make for a perfect system. Medicare claims data can be misleading, and Medicare patients use lots of meds--an average of 28 per year--so teasing out which drug might be to blame for a problem can be, well, a problem. Realizing this, FDA is initially going to use the system to get better data on side effects that already are well known.
Another potential problem, for drugmakers at least: The system will most likely boost the number of drug-safety warnings, at least at first. Some of those warnings could be false alarms. CDER chief Janet Woodcock promises to work to minimize that possibility; we'll see how it plays out.