What a UDI is
A Unique Device Identifier is a code that identifies a medical device through distribution and use. FDA's UDI rule requires the labeler, usually the manufacturer, to place a UDI on device labels and packages, in both human-readable text and a machine-readable form such as a linear or 2D barcode. The requirement was phased in by device class between 2014 and 2022, with Class III devices first and Class I and unclassified devices last. Some devices are exempt or have exceptions.
The UDI has two parts. The device identifier, or DI, is fixed and identifies the labeler and the specific version or model of the device. The production identifier, or PI, is variable and includes whichever of these appear on the label: lot or batch number, serial number, expiration date, manufacturing date and, for certain cellular and tissue products, a distinct identification code.
Labelers submit DI information, not PI data, to FDA's Global Unique Device Identification Database, known as GUDID.
Issuing agencies and barcode formats
UDIs are created under the systems of FDA-accredited issuing agencies. In the United States those are GS1, HIBCC and ICCBBA. Most devices in commercial distribution use GS1, so it is worth understanding how GS1 encodes UDI data.
GS1 barcodes use application identifiers, short numeric prefixes that tell the scanner what each piece of data means:
- (01) Global Trade Item Number, which serves as the DI
- (10) lot or batch number
- (17) expiration date in YYMMDD format
- (11) production date
- (21) serial number
A single GS1-128 linear barcode or GS1 DataMatrix 2D code can carry several of these at once. A warehouse system that parses application identifiers can capture the DI, lot, serial and expiry from one scan instead of relying on someone to read and type them, which is both faster and far more accurate. HIBCC codes follow a different structure and should be supported too if any of your suppliers use them.
Why the PI matters more in distribution
For a distribution partner, the production identifier carries most of the value. The DI tells you which product a unit is, which the warehouse already knows from the SKU. The PI tells you which specific lot or unit it is, which is exactly what a recall, a field safety notice or a complaint investigation needs.
A common mistake is to treat UDI compliance as purely a labeling matter. Labels can be perfect while traceability still fails, because the distribution partner never records which lot or serial was shipped to whom. Labeling is the manufacturer's obligation; capturing and preserving PI data through distribution is what makes that labeling useful.
UDI data also supports customers. Many health systems record UDIs in their materials management systems and, for implants, in patient records. Shipping documentation and ASNs that carry lot and serial data help them do that accurately, and some contracts require it.
The four scan points
In a well-run medical fulfillment operation, UDI data is captured or verified at four points.
- Receiving: each inbound unit, inner pack or case is scanned so the DI, lot, serial and expiry are recorded before putaway. Discrepancies against the ASN go to hold.
- Putaway: lots are placed into locations that keep them segregated, so a picker cannot mix lots by accident.
- Picking: the picker is directed to a specific lot and must scan to confirm it. A mismatch stops the pick.
- Packing: serialized units are scanned individually, and lot-controlled items are confirmed, before the shipment record is closed.
Each scan writes to an audit trail with the user and time. The result is a shipment record that reflects what physically left the building, not what the system intended to ship. That distinction is the whole point of scan-based traceability.
Lot-controlled vs. serialized items
Not every device is serialized. Many disposables and consumables are controlled by lot and expiry only, while higher-risk or higher-value devices such as implants, powered equipment and connected monitors are often serialized.
The operational difference is significant. Lot-controlled items can be received, picked and packed by the case or inner pack, with the lot recorded once per container. Serialized items need each unit's serial captured, usually at receiving and again at packing, which takes more scans and more time.
Your item master should define the control level for every SKU: lot only, lot and expiry, serial, or serial with lot and expiry. It should also define the packaging level at which each identifier is captured. Getting this configuration right during onboarding prevents both under-tracking, which creates recall risk, and over-tracking, which adds cost without benefit.
Records a recall actually needs
When a manufacturer initiates a recall or correction, the distribution question is narrow: which consignees received affected units, how many and when. To answer it, every shipment record should contain:
- Consignee name and full ship-to address
- Order number and customer PO number
- Ship date, carrier and tracking number
- Item, lot, serial and expiry for every line
- Quantity shipped per lot or serial
With that data captured at the unit level, a recall scope is a query by lot or serial. Without it, a manufacturer is left to notify every customer who bought the product in the relevant period, which costs more, alarms more customers and makes effectiveness checks harder.
The same records also support routine needs such as answering a hospital's question about which lot was in a given delivery or reconciling consignment inventory.
Common UDI pitfalls in distribution
Most UDI problems in distribution come from a handful of causes:
- Manual keying of lot numbers instead of barcode scanning, which introduces transcription errors
- Receiving by SKU without lot capture, then trying to add lots later
- Mixed-lot cartons from suppliers that are received as a single lot
- Relabeling or overlabeling that obscures the original UDI, which can also create regulatory obligations for whoever relabels
- Returns put back into stock without re-confirming lot and serial
- Kits assembled from components without recording component lots where your records require it
Each of these can be prevented with configuration and procedure. Ask your distribution partner how they handle each one, and test the answers with a sample shipment before full launch.
Frequently asked questions
What is the difference between a DI and a PI in a UDI?
The device identifier (DI) is fixed and identifies the labeler and device version or model. The production identifier (PI) is variable and includes data such as the lot number, serial number, expiration date and manufacturing date.
Does a 3PL submit data to GUDID?
No. The labeler, usually the manufacturer, submits device identifier information to GUDID. A 3PL captures UDI data from labels to track lots and serials through distribution.
What GS1 application identifiers are used for UDI?
Common GS1 application identifiers include (01) for the GTIN, (10) for lot, (17) for expiration date, (11) for production date and (21) for serial number.