The Medical Device Validation Handbook
167
Two issues need to be addressed with any
new labeling applications employing AIDC
technology:
• Authenticating the format of the encoded
information against application require-
ments, and
• Confirming the barcode meets quality
requirements—that is, barcode verification—
to assure reliable “reading” at the point of
use.
Labelers may use 1D or 2D symbology on labels.
The symbols need to be verified to determine
whether they are correctly configured, the label
information is accurate, and the information
corresponds to the human-readable text. It is not
practicable to inspect each label as it is printed.
Machine-vision software is available that allows
labelers to “read” the labels, segregate the fields
within the barcode, interpret the human-readable
text, and verify the form and content of both are
correct.
Printing barcodes is one thing being able to
read them is another. Symbol quality is critical
to the ability to verify barcodes. Barcode symbols
must be printed at the highest possible quality
level to maximize their ability to be successfully
decoded by the variety of barcode scanners and
cameras available today. High-quality 1D and
2D barcodes can be validated through verifica-
tion when labels are printed. They must also be
“graded” at the same time (barcode verification).
The minimum acceptable grade to guarantee
meeting industry standards and reliable decod-
ing throughout the supply chain is “C,” or 2.0.
Labelers need to understand barcode print quality.
Topics such as “edge determination,” “minimum
reflectance,” and “quiet zone” are explained in The
Layman’s Guide to ANSI, CEN, and ISO bar
code print quality documents.10
In addition to label requirements, medical
devices intended by manufacturers to be “repro-
cessed” (cleaned and disinfected, or cleaned and
sterilized) must have the UDI “direct marked”
permanently on the device itself. Laser marking
is one such method. The UDI in these cases will
be very challenging to verify and will necessitate
the use of an automated process to confirm the
label information.
Software V&V11 is critical to successful UDI
implementation. For more information on how
to validate software, be sure to read Chapter 11
on software validation.
Global Unique Device Identifier Database
Managed by the FDA, the GUDID serves as a
reference catalog for every medical device with a
DI meanwhile, the public site—AccessGUDID—
is maintained by the National Institutes of
Health. The UDI rule requires the labeler of every
device with a UDI to submit data on each to the
GUDID, unless there is an exception or alterna-
tive. There are 60-plus attributes that need to be
entered for each device (see Table 11-2).
There are two standard-based methods to
submit data:
• Structured input via an internet interface
• HL7 SPL
HL7 SPL is an XML format and uses the
FDA’s ESG as the pathway to upload data to the
GUDID. Both submission methods are one DI
record at a time. No batch option is available. For
HL7 SPL, there is one DI record per XML file.
The web-based tool will work well for up
to 200 total DI records. Any more than that
becomes overwhelming companies with that
records volume should use HL7 SPL instead.
Experience has shown it can take anywhere from
three to six minutes to enter each DI record
manually into the online interface. Third parties
also can be used to submit data. Companies also
can build their own submission tools.
The only validation in the GUDID is
based on what the database does intrinsically or
through business rules designed into the entry
program. Both submission methods mentioned
above use standardized controlled vocabularies
and business rules to validate each attribute
entered. Controlled vocabularies are:
• Data Universal Number System (DUNS)
number
• GMDN code
• FDA product codes
167
Two issues need to be addressed with any
new labeling applications employing AIDC
technology:
• Authenticating the format of the encoded
information against application require-
ments, and
• Confirming the barcode meets quality
requirements—that is, barcode verification—
to assure reliable “reading” at the point of
use.
Labelers may use 1D or 2D symbology on labels.
The symbols need to be verified to determine
whether they are correctly configured, the label
information is accurate, and the information
corresponds to the human-readable text. It is not
practicable to inspect each label as it is printed.
Machine-vision software is available that allows
labelers to “read” the labels, segregate the fields
within the barcode, interpret the human-readable
text, and verify the form and content of both are
correct.
Printing barcodes is one thing being able to
read them is another. Symbol quality is critical
to the ability to verify barcodes. Barcode symbols
must be printed at the highest possible quality
level to maximize their ability to be successfully
decoded by the variety of barcode scanners and
cameras available today. High-quality 1D and
2D barcodes can be validated through verifica-
tion when labels are printed. They must also be
“graded” at the same time (barcode verification).
The minimum acceptable grade to guarantee
meeting industry standards and reliable decod-
ing throughout the supply chain is “C,” or 2.0.
Labelers need to understand barcode print quality.
Topics such as “edge determination,” “minimum
reflectance,” and “quiet zone” are explained in The
Layman’s Guide to ANSI, CEN, and ISO bar
code print quality documents.10
In addition to label requirements, medical
devices intended by manufacturers to be “repro-
cessed” (cleaned and disinfected, or cleaned and
sterilized) must have the UDI “direct marked”
permanently on the device itself. Laser marking
is one such method. The UDI in these cases will
be very challenging to verify and will necessitate
the use of an automated process to confirm the
label information.
Software V&V11 is critical to successful UDI
implementation. For more information on how
to validate software, be sure to read Chapter 11
on software validation.
Global Unique Device Identifier Database
Managed by the FDA, the GUDID serves as a
reference catalog for every medical device with a
DI meanwhile, the public site—AccessGUDID—
is maintained by the National Institutes of
Health. The UDI rule requires the labeler of every
device with a UDI to submit data on each to the
GUDID, unless there is an exception or alterna-
tive. There are 60-plus attributes that need to be
entered for each device (see Table 11-2).
There are two standard-based methods to
submit data:
• Structured input via an internet interface
• HL7 SPL
HL7 SPL is an XML format and uses the
FDA’s ESG as the pathway to upload data to the
GUDID. Both submission methods are one DI
record at a time. No batch option is available. For
HL7 SPL, there is one DI record per XML file.
The web-based tool will work well for up
to 200 total DI records. Any more than that
becomes overwhelming companies with that
records volume should use HL7 SPL instead.
Experience has shown it can take anywhere from
three to six minutes to enter each DI record
manually into the online interface. Third parties
also can be used to submit data. Companies also
can build their own submission tools.
The only validation in the GUDID is
based on what the database does intrinsically or
through business rules designed into the entry
program. Both submission methods mentioned
above use standardized controlled vocabularies
and business rules to validate each attribute
entered. Controlled vocabularies are:
• Data Universal Number System (DUNS)
number
• GMDN code
• FDA product codes