As deadlines for meeting electronic pedigree (ePedigree) requirements approach, a growing number of technology choices help deliver supporting data. The foundation of a drug ePedigree, serialization, is the application of a unique code at the item level. These unique codes can be used to authenticate products and enable tracking and tracing.
In a March 2010 guidance document, Standards for Securing the Drug Supply Chain—Standardized Numerical Identification (SNI) for Prescription Drug Packages, the US Food and Drug Administration recommended that each sellable unit carry a unique serial number. In most cases, according to FDA, this identification could consist of the National Drug Code for the product plus a unique serial number with a maximum of 20 characters, printed in both machine- and human-readable forms.
Serialization begins on the packaging line with the application of a two-dimensional DataMatrix code or a radio-frequency identification (RFID) tag. Since a DataMatrix code is easy and inexpensive to apply, it's now more commonly used than RFID. Many processes reproduce DataMatrix codes, including inkjet, thermal-transfer, and laser coding.
The hardware and software associated with code capture and recording also must accommodate events on the line such as rejects or quality-assurance sampling to ensure that no discrepancies occur between the products and the data.
Line-management software sends serialization data, including all good numbers (i.e., commissioned products), bad numbers (i.e., decommissioned products), and parent–child relationships to higher level systems. "Ultimately, this data goes into a system of record that becomes the master historian," says Joe Ringwood, chief operating officer at Systech International (Cranbury, NJ).
Implementing serialization requires hardware such as printers and vision systems as well as software for the line, site, and historian levels. Integration is required to make everything work together. The complexity of the integration depends on what systems are already in place and the approach taken by the manufacturer.
In general, commercial, configurable serialization software will take less time to implement than line, plant, and enterprise systems that must be customized to perform serialization functions. Choosing specific, standard devices also simplifies implementation, particularly if multiple lines or locations are involved. It's also helpful to specify device-neutral software that is compatible with virtually any brand or model of equipment.
"Serialization means you have to know what happens to each serialized package and requires new workflows and processes," says Ringwood. "So the better job done assessing current workflows and defining 'to-be' workflows, the smoother the implementation," he predicts.
Ringwood also advises looking at the total cost of ownership instead of focusing solely on the expenses associated with implementation. "Many decisions have implications down the road," he explains.
Implementing serialization requires months of planning to identify software needs, qualify vendors, perform pilot projects, and integrate hardware and software. Serialization teams should include representatives from automation, information technology, engineering, and plant-floor teams. In fact, says Roebles, the most successful implementations involve and train the plant-level personnel who work with the system daily. Actual implementation schedules vary, but vendors agree that an 18-month validation process is common. "Companies that haven't started planning better start soon or they won't be ready when FDA or a country they sell to sets [ePedigree] rules," warns Roebles.