Few weeks ago, there was an interesting
presentation about the device database.
(available here: http://accessgudid.nlm.nih.gov
)
It was about Unique Device Identifiers
(UDI) and what are the different parts of this identifier.
A UDI is a composite of UDI = DI + PI
The bar code is also using “(01)” notation
to identify different subparts of the UDI.
There was some discussion about fixed
length vs. dynamic length for a part of an identifier.
I would like start a discussion in the
forum as a follow up of the presentation.
Here is the structure of the UDI
This is analogous to the 9 digit and 11 digit
NDCs. It would be ideal if as researchers we could avoid in UDIs having the
same issues with zero padding in NDCs codes.
I would definitely want to always be able
to extract (in SQL) reliably the DI part from a column in the database that
stores the UDI.