Aggregation of saleable drug packages to shipping cases and pallets is not required by the US Drug Supply Chain Security Act (DSCSA) or the EU Falsified Medicines Directive (FMD) or the related Delegated Regulation (EUDR). But certain business processes in the EU under the FMD will be difficult to accomplish without it, and after November 2023, the operation of the supply chain in the US will not be efficient without it (see “Aggregation: The Achilles’ Heel of Pharma Supply Chain Operation Under A Serialization Regulation”, “EU FMD: Aggregation Is Not Mandated, But It Will Be Necessary” and “Pharma Aggregation: How Companies Are Achieving Perfection Today”). Absent a mandate, companies need to recognize, themselves, just how vital aggregation is to their businesses and prepare to generate it and/or make use of it.
Toward that end, last month GS1 Healthcare published a short “Discussion paper on aggregation in the pharmaceutical supply chain”. I think it is in response to the recent convening of a panel of experts in the EU to analyze the sometimes hidden need for aggregation and propose a way to produce and exchange it. GS1’s new discussion document (available here) is brief, but it provides a good definition of the often confusing terms “aggregation”, “trade item” and “logistics unit”, and how to use GS1’s Electronic Product Code Information Services (EPCIS) to document and share aggregation data (see “EPCIS Explained”).
This new document is for anyone needing a high-level introduction to aggregation, but it has its limitations. My biggest complaint is that it doesn’t spend enough space explaining why aggregation is important and what you use it for. The need for aggregation only follows from the need for inference. Nobody decides to capture aggregation information just for the fun of it. So to understand aggregation, you first need to have a good appreciation for what inference is and why it is so important. A study on aggregation should always start with inference (for an example, see “Pharma Aggregation: How Companies Are Achieving Perfection Today”). But the GS1 paper dedicates only a single short paragraph to inference, and it’s on the last page of text:
“Suppliers share their Aggregation Events with the intended recipient(s) of those objects. This allows downstream partners to practice inference, by which the recipient only needs to verify the identifiers of the highest-level objects in the packaging hierarchy. The full packing hierarchy (including all levels of aggregated children) from prior packing steps is inferred to be intact.”
I know they were trying to create a short paper, but I think that’s too brief to fully convey what the purpose of aggregation is (see “Inference in the Pharmaceutical Supply Chain”). Maybe the paper should have been focused on inference, and then include a brief explanation of aggregation, instead of the other way around.
Back in 2010, GS1 US published a white paper called “The Practice of Inference in the U.S. Pharmaceutical Supply Chain”, which explained it well, but they removed it from the internet at some point in recent years. Even the current GS1 US DSCSA Implementation Guideline version 1.2 relies on a reference to that white paper to explain the concept. It was kind of old, so I’m guessing that inference paper must have spoken too much about the now preempted California Pedigree regulation so maybe that’s why GS1 US has withdrawn it (see “California Pedigree Law: Historic Change to Commerce” and “Preemption: What Does It Mean?”). Hopefully they will update it and refocus it on the DSCSA, FMD and other pharma serialization regulations around the world. The recent regulations, including Russia, Brazil and several in the Middle-East, include aggregation as part of their mandates so that inference can be used.
There are a few other, less serious issues with the new discussion paper:
- The same picture is used to depict both, a “trade item, next level above secondary pack” and a “logistics unit”. That’s going to be really confusing for people who are new to the concept of aggregation because it is a picture of a corrugated box with a linear barcode on its side and tape across the top. To most people, that looks a lot like a typical logistics unit, not a trade item. Pictures are worth a thousand words, unless they aren’t intuitive, and then they just get in the way to understanding.
- In the document they refer to something they call a “registered trade item”, and something else they call a “trade item”, as if there are some trade items that are “registered” and some that are not. If they aren’t going to explain the difference (they don’t—the paper is too short, remember), then they should just use one term or the other.
- They don’t define the term “bundles”, although they use it as if we already know what it is. Maybe most pharmaceutical packaging professionals will know that a bundle is just an intermediate non-trade item grouping of trade items used to make the packing of a logistics unit easier to accomplish. The paper claims that a bundle can be optionally identified by either a GTIN or an SSCC, but I would argue that if it’s identified by a GTIN then it’s not a bundle, it’s a trade item. Labeling bundles with an SSCC is usually just to make the packing process easier, it’s not necessary for shipping purposes. A bundle is not a shipping container. Don’t use a GTIN unless you plan to offer that bundle grouping as an orderable trade item (like a multi-pack).
The page in the document called “Aggregation with EPCIS” subtly exposes one of the shortcomings of using EPICS just for aggregation—its complexity. They acknowledge that you can document aggregation without EPCIS, but then they try to imply that all you need is an EPCIS Aggregation Event to do it with EPCIS. And then they step into quicksand by talking about the “unpacking” business step—something no one in the pharma supply chain is likely to use—and they are stuck. Oops. It’s starting to sound a little complex.
And then they have to say that their supply chain drawing—which includes packing at the manufacturer, shipping, receiving by a distributor, unpacking, packing, shipping, receiving by a dispenser and unpacking—is “…intentionally simplified and does not include the beginning/end of life Commissioning and Decommissioning steps which are critical to ensuring the end-to-end traceability of each serialized product.” Thus basically proving that if all you want to do is document aggregation (like they will want to do in the EU), EPCIS is a really complex way to do it. It will be fine here in the US in 2023 because we will likely need to use EPCIS to document those other supply chain events in addition to aggregation. It probably won’t be a good use for the EU or those other countries that are mandating aggregation data because those countries are defining their own data format for upload, and it’s not likely to be EPCIS.
Of course, as usual, I’m being overly critical here. The GS1 discussion paper on aggregation isn’t all bad. It’s a quick read. See what you think.