Towards epidemiology

The FDA is already helping to sponsor research to see if the UDI could help play a role in device epidemiology.


For instance, at Mercy headquarters in a St. Louis suburb, Drozda's team is working on a demonstration project for the Medical Device Epidemiology Network, or MDEpiNet, which would use a database filled with searchable attributes linked to the UDI to help doctors or researchers watch products through the supply chain. For researchers, this means more powerful post-market studies and comparative effectiveness research, Drozda said. They can start to compare one company's device, say a lead wire, against another company's, but by having access to attributes across all devices and manufacturers, you could design queries for the database just focusing on one attribute. For lead wires, for instance, you could see whether leads under a certain diameter were more prone to fracturing, Drozda said. "The attributes let you do a lot more sophisticated analyses."

They also give the FDA a better grasp on relative safety, by giving both adverse events for a device and what's known as "exposure," or the amount of people who are known to have used the device or have had it implanted. "So they can see we've gotten four (adverse event) reports with this out of 1,000 of those things implanted," Drozda said.

The system could also help doctors at the point of care. According to Drozda, ideally, based on their demonstration project, it would work like this: an item comes into the cath lab or operating room, the doctor or nurse scans it, and then based on the UDI, the staff could access an attributes database linking to relevant clinical characteristics that could even help drive care decisions.

"They'd have a lot of information available to them, simply by scanning the bar code that has the UDI on it, " Drozda said.

As part of the implementation of the UDI, the FDA is creating its own database, the Global UDI Database, but Drozda said because of regulatory implications, it could be "cumbersome" to add or remove device attributes to the database that might have clinical significance and would be interesting to a doctor or researcher. The attribute database being developed by Drozda, however, would be more flexible.

Right now, in the demonstration project, Drozda's team is building a database only for stents, which includes at least nine attributes, such as length, material, and whether it's impregnated with a drug. The reason for starting with stents is simple, Drozda says. "It's a finite universe you can get your arms around." There are only three manufacturers (at least in the United States), the science is fairly well established, and manufacturers are often willing to share data. But later, Drozda said they'd hope to expand it to areas with potentially more safety and comparative effectiveness issues, such as implantable cardiac defibrillators, ICD wires and hip and knee replacements.

More Industry Headlines