After a long wait, the FDA announced the proposed UDI rule today helpful resources. In 165 pages, it actually says very little.
Basically, manufacturers have to assign an identifying number to a medical device model and put this number on the product or product label in human readable and machine readable form. (It does get interesting for devices that require direct part marking however).
The part about machine readable is dubious because the FDA refrained from boldly specifying the technology to be used, such as bar code or rfid tag. This means that no one (read: hospitals) could invest in reading equipment. Especially with the many varieties of rfid tag now available and being developed for the future, I can not imagine any hospital justifying the purchase of reading equipment. And if this is the case, then why should FDA require a machine readable number at all?
In contrast to this decision, the FDA boldly specified a bar code for unit dose drug labeling, and a linear bar code at that. In doing so, they opened the door for hospitals to confidently invest in the lowest cost scanners (since they did not need to be capable of reading 2D matrix codes as many had argued for) and immediately benefit from the presence of these bar codes on drugs. Patient safety followed as patients and drugs could be matched at the bedside.
Arguably, the same benefit will not result from reading model numbers on medical devices used in hospitals. Who needs to record the exact model of a thermometer every time it is used? However, one would think that traceability of implants, down to the serial number would be critical.
The announcement to day is only a draft of the proposed rule, and the FDA invited comments from the public. I hope that some of those comments would address this concern and ask for definitive guidelines on what kinds of machine readable technology (1D bar code, 2D bar code, rfid tag, etc.) must be used so that systems can be put in place to use it.
by Glenn Spitz