Part 1 of this essay provided a wealth of hyperlinks into the Code of Federal Regulations (CFR) and FDA guidance documents with content related to placing the National Drug Code in human- and machine-readable form onto drug packages prior to November 27, 2017 (see “Is A GS1 GTIN Really Usable As An NDC For DSCSA Compliance? Part 1”). In Part 2, we will look at how the Drug Supply Chain Security Act (DSCSA) will change, or add-to, the requirements found in those earlier specifications. And finally, we will be able to answer the question in the essay title.
HOW THE DSCSA CHANGES THE NDC AND BARCODE REQUIREMENTS FOR DRUG PACKAGES
First of all, the DSCSA does not change anything about the existing linear barcode that contains the NDC as described by Part 1 of this essay. That barcode must remain on your drug packages, even after November of 2017—for now, at least. But second, you will need to add a new 2-dimensional (2D) data matrix barcode in addition to that linear barcode. This new 2D data matrix barcode must encode “…the standardized numerical identifier, lot number, and expiration date of the product” [from DSCSA Section 581(14)] (see “The DSCSA Product Identifier On Drug Packages”). For details on what it will take to meet the DSCSA serialization requirement on drug packages (and also in the European Union to meet their requirements), see “Meeting U.S. and E.U. Drug Serialization Requirements With A Single Solution”.
For the purposes of this essay, we want to focus on what the DSCSA says about the NDC specifically as it is applied to drug packages, and GS1, if at all. If you do a simple text search of the text of the DSCSA for “National Drug Code”, you will get three hits. One is where the NDC is mentioned in the definition of the term “Homogeneous Case” [DSCSA Section 581(7)], another is where it is listed as one of the data elements of the Transaction Information [DSCSA Section 581(26)] (see “DSCSA: Transaction Information”), and the third is within the DSCSA definition of the Standardized Numerical Identifier (SNI) [DSCSA Section 581(20)]. Within that definition, the NDC of the drug is one of the two constituent parts (the serial number is the other part). That’s it.
If you do the same kind of text search for “GTIN”, you won’t find it. If you search for the text “GS1”, you also won’t get any hits, but that doesn’t mean it isn’t there. In fact, it appears—in hidden form—several places. Congress did not specify GS1 explicitly or exclusively. Maybe they wanted to show the FDA how it can be done without identifying a specific standards development organization. The FDA used a comparable approach in their 2004 linear barcode rule (see Part 1). Instead, in the DSCSA, Congress simply required certain things to conform “…to the standards developed by a widely recognized international standards development organization…”. That could be HIBCC. That could be the OPCFoundation. That could be IEEE. But it will be none of those. It will be GS1, and almost everyone in the industry knows it already.
Let’s take a closer look at where GS1…ahem…where the “widely recognized international standards development organization”…is referenced in the DSCSA. It appears twice. One is in Section 582(a)(2) Initial Standards, and the other is in Section 582(h)(4) Standards for Interoperable Data Exchange. In actuality, these are covering the same standards. The first reference is where the DSCSA talks about the initial standards for data exchange that the FDA needed to provide draft guidance on initially (see “FDA Publishes Draft Guidance For DSCSA Data Exchange”), for use on January 1, 2015. In that draft document, published in November of 2014, the FDA sanctioned the use of GS1’s Electronic Product Code Information Services (EPCIS) standard for DSCSA data exchange. And the second reference is where the DSCSA talks about the FDA updating that exact same guidance—then in final form—for the Enhanced Drug Distribution Security (EDDS) phase, which starts on November 27, 2023.
None of these references are related to the use of GS1 Global Trade Item Numbers (GTINs) in meeting the NDC requirements. Does that mean it cannot be used?
HERE IS WHERE THE FDA MENTIONS GTIN AND WHERE IT MAY APPLY TO THE DSCSA
The place where GS1’s GTIN standard is mentioned and where it may apply to its use in the DSCSA is in the SNI guidance that was published by the FDA back in early 2010, well before the adoption of the DSCSA (see “FDA Aligns with GS1 SGTIN For SNDC” and “Anatomy Of An FDA SNI”). The development of that guidance was mandated by Congress in the Food and Drug Administration Amendments Act of 2007 (FDAAA) [see Section 505D(b)(2) of the FD&C]. Congress wanted the SNI to “…be harmonized with international consensus standards for such an identifier.” In the resulting SNI guidance, FDA established that the SNI for most drugs would be a “serialized National Drug Code (sNDC)”.
“The sNDC is composed of the National Drug Code (NDC) (as set forth in 21 CFR Part 207) that corresponds to the specific drug product (including the particular package configuration) combined with a unique serial number, generated by the manufacturer or repackager for each individual package. Serial numbers should be numeric (numbers) or alphanumeric (include letters and/or numbers) and should have no more than 20 characters (letters and/or numbers).”
Notice the reference to 21 CFR Part 207, which we scrutinized in Part 1 of this essay, and where it is established that the NDC is a 10-digit code.
Be aware that the sNDC is not the only SNI that can appear on a prescription drug product. According to the SNI guidance:
“blood and blood components and certain minimally manipulated human cells, tissues, and cellular and tissue-based products (HCT/Ps), do not currently use NDC numbers. Examples of HCT/Ps that do not use NDC numbers include allogeneic placental/umbilical cord blood, peripheral blood progenitor cells, and donor lymphocytes for infusion. Instead, such products currently use other recognized standards for identification and labeling, such as ISBT 128, which creates a unique identification number for each product package. […] The SNI for these products should be the unique identification number created for each package under these other recognized standards, such as ISBT 128.”
Finally, in answering the question “Is the SNI that FDA is recommending compatible with international standards” as Congress mandated, the SNI guidance says:
“…[A]doption of the sNDC as the SNI for most prescription drugs, and of other recognized standards, such as ISBT 128, for certain biological products, satisfies the requirement in 505D(b)(2) that the SNI developed by FDA be harmonized, to the extent practicable, with internationally recognized standards for such an identifier. Specifically, use of an sNDC is compatible with, and may be presented within, a GTIN, which can be serialized using an Application Identifier (AI) (21) to create a serialized GTIN (sGTIN) for use with RFID or for certain barcodes. GTIN is a global standard for item and object identification, established by GS1, a consensus-based, not-for-profit, international standards organization that works with manufacturers, distributors, retailers, and others in the drug supply chain. …”
IS A GS1 GTIN REALLY USABLE AS AN NDC FOR DSCSA COMPLIANCE?
We finally have what we need to answer the question, is a GS1 GTIN really usable as an NDC for DSCSA compliance? The text of the DSCSA is not enough to answer it, but when combined with the SNI guidance from 2010, I think we have everything we need. According to that guidance as quoted directly above, the NDC may be represented within a GTIN, which can be combined with a GS1 serial number using GS1 Application Identifier (AI) 21 to create a serialized GTIN (sGTIN) ”for use with RFID or for certain barcodes”. Unwinding this back up through the DSCSA, the “2-dimensional (2D) data matrix barcode” mentioned in DSCSA Section 582(a)(9) can clearly be implemented with a GS1 Datamatrix barcode. (Notice that the term “data matrix” refers generically to any barcode that conforms to the ISO/IEC 16022:2006 standard, where the term “Datamatrix” distinctly refers to the GS1 standards-conformant variation of the ISO/IEC standard.)
It appears clear to me that the FDA will accept the NDC encoded into a GTIN for the purpose of combining it with other data elements, like serial number, lot and expiration date, when creating a 2D barcode that complies with the DSCSA (see “Depicting An NDC Within A GTIN”). It is much less clear that they would accept a human-readable GTIN to represent an NDC. In fact, my interpretation would be that they would not accept a GTIN as the human-readable NDC. Of course, the FDA will have the last word on that, if we could just get them to respond to questions like this.
Ignoring this trail of references from the CFR and FDA guidance, I also think it is simply illogical for the FDA to accept the GTIN in place of an NDC because the human-readable contents of a drug label is intended to be read and used by the FDA, patients and medical professionals. For the last 40 years, the 10-digit NDC is the code that identified a given drug in a specific packaging level. It doesn’t make any sense to introduce a GS1-specific formatting of that 10-digit code on drug packages, even if the 10-digit code is visible in the middle of the 14-digit GTIN. Patients and medical professionals should not be expected to understand the GTIN formatting and how to extract the specific 10 digits from the 14 of the GTIN, if it were printed on the drug in place of the NDC. This would be unsafe. (But also see “An Open Letter To The FDA, RE: Please Fix The National Drug Code Soon”.)
Likewise, the formatting of the human-readable expiration date should not be formatted in the same way that GS1 requires it to be formatted inside their barcodes. GS1 requires a format of “YYMMDD” inside the barcode, but the FDA recommends drug manufacturers follow US Pharmacopoeia (USP) guidance for formatting the human-readable expiration date on drug packages. That guidance shows examples as “Month/Year” for the expiration date in human-readable form and this has been the norm for decades.
In short, the formatting of the data dictated by GS1 standards when encoding a GS1 barcode should not be followed when formatting the human-readable contents that appear next to that barcode (see also “I’ve Changed My Mind, Follow GS1’s HRI Specification”). If you feel that you must put a GTIN in the human-readable next to the barcode, then also put the 10-digit NDC above it, and in a slightly larger font so that patients and medical professionals will see it first. Label it as “NDC” and label the GTIN as “GTIN”.
Disagree? Leave a comment below.
Dirk.