Leading the Way
When it comes to implementing item-level serialization, most of the pharmaceutical supply chain seems to be waiting for FDA’s next move, or for California’s deadline for electronic pedigrees to get closer. Nonetheless, Purdue Pharma and Pfizer keep forging ahead. These two pharmaceutical manufacturers have been at the forefront of the pharmaceutical industry in piloting and adopting automatic identification technologies at the unit package level to combat counterfeiting and diversion. Starting in 2004–2005, these manufacturers began employing serialized RFID and/or 2-D bar codes to achieve internal efficiencies and a degree of supply-chain visibility.
|Pfizer's new unit label for Viagra.|
At AIM Expo in Las Vegas, Harry Ramsey, manager, package design & development at Purdue Pharma, and Murugan Govindasamy, senior manager, global manufacturing services at Pfizer, spoke about their journeys to automatic item-level serialization in “Item-Level Pharma Serialization: The Role of RFID and Data Matrix Bar Code Technology.”
At the event, Govindasamy explained how Pfizer is modifying its end-to-end serialization pilot to comply with California’s law. Pfizer began its pilot in 2005 for serializing Viagra, testing both RFID and bar codes on three packaging levels: units, cases, and pallets. Its program progressed to develop sound practices for aggregation, and it has since evolved to follow GS1 standards at all coding levels.
“Meeting patient safety was our number-one driver,” says Govindasamy. “As the industry leader and manufacturer of one of the popular medicines being counterfeited, we had the obligation to take the lead in developing the serialization program and also test the technology and the supply network. We also believed that it would help to build confidence in the track-and-trace technology among the industry partners and the regulatory agencies.”
Pfizer’s Viagra packaging line in Amboise, France, has been modified to include only 2-D bar codes at the unit level. “Auto ID technologies continue to evolve with or without pharma industry. But the pharma industry has some unique requirements, which have been evolving in the last few years,” says Govindasamy. “For example, DataMatrix has been in use for more than 20 years, but print quality verification guidelines and technologies were developed only after the pharma industry started using online printing systems to print 2D codes on cartons and labels at the packaging lines. The focus on this symbology only happened after healthcare companies asked GS1 to include it in their standards as a true 2-D code. From there, existing standards were further refined in the GSMP process of GS1. When we started our pilot, we had developed internal standards and adopted our own guidelines based on our knowledge on the possible future regulations. As we had committed to support the development of the standards, we were prepare to adapt to those standards as time moved on.”
In 2004, Purdue Pharma began implementing ultra-high frequency RFID for Oxycontin, starting the program mainly in response to Walmart’s requests. But the company saw further utility. “We were aware of the benefits track and trace could offer to fight diversion of Oxycontin, and management supported us fully,” explained Ramsey during an earlier AIM Expo session, “Serialization in the Pharmaceutical Supply Chain: Lessons Learned and Remaining Challenges,” moderated by Bob Celeste, GS1 US.
The company started with bottle tagging using Gen1 UHF RFID, then progressed to Gen2 testing in June 2006 with case-level tagging and shipping in June 2007. “We started with a bolt-on approach using Class 0 Gen1 RFID tags, eventually upgrading to Gen2 tags. We did not aggregate at first, but did so with the Gen2 upgrade. As of May 2007, every bottle of Oxycontin has an RFID chip, and to date we have hit more than 20 million bottles. RFID has worked perfectly within our own four walls.”
PFIZER’S NEW LABEL
Having reworked its Viagra 30-count bottle labeling line in France, Pfizer is introducing the new unit label with a shaded box featuring a 2-D DataMatrix that encodes the GTIN (GS1 Application Identifier 01) and the serial number (AI 21); also in the box below the 2-D code are the human-readable lot number, expiration date, and serial number. These data are printed in white against the color background. A linear bar code encoding the GTIN along with a human readable GTIN are printed in black just outside the shaded box, running vertically to the left. A strip of color-shifting ink featuring Pfizer logos is also included on the label.
|See the sidebar "Achieving Necessary Control"|
At the case level for 48 bottles, Pfizer is employing two linear GS1-128 bar codes and one 2-D Data Matrix bar code to represent a host of data that include the GTIN, case serial number, batch/lot number, expiration date, and item counts. The case label also employs an RFID tag encoded under the SGTIN-96 format. At the pallet level, a linear bar code encoding the pallet ID number and an RFID tag encoded under the SSCC-18 format are employed. All data is aggregated at each packaging level through device, line, and plant level systems under the TIPS Guardian SPT server by Systech, which sends data to SAP Aii.
“Adopting global standards is important and achievable,” Govindasamy told the AIM Expo audience. However, “implementation takes considerable time. Implement building blocks for serialization in phases, and develop a flexible, modular, and configurable solution. Selecting technology with a long-term strategy in mind is very important, and partnering with technology solution providers is critical.”
“Clear definitions and usage methods can be shared with all levels of users,” continues Govindasamy. “Equipment vendors should understand clearly so that they can develop technologies to meet these standards. Also, standards should evolve constantly in relation to current technology capabilities. For example, printing and coding and inspection technologies should be more robust.”
The hardest part of serialization is readying a 20-year old line not designed for serialization with very minimal capital investment,” adds Govindasamy. “The easiest part is replicating proven technologies.”
For Ramsey, getting RFID vendors to work together to prove that the technology could work was the easiest part of serialization employing RFID, he told the audience.
To date, however, RFID has not gained significant traction in the supply chain. “Less-than-full totes have been problematic for distributors,” says Ramsey. But it is surprising that pharmaceutical packagers haven’t been more supportive. “The processes for RFID and bar coding are so different from each other. You can tailor RFID to your packaging line, but you have to tailor your packaging line to bar coding for complete line of sight,” he told the audience during an earlier AIM Expo session, “Serialization in the Pharmaceutical Supply Chain: Lessons Learned and Remaining Challenges,” moderated by Bob Celeste, GS1 US.
Purdue continues to employ RFID for item-level serialization. “We always planned to use 2-D Data Matrix for redundancy, but we wonder whether it will become the sole carrier of the serial number. We are looking at how we can meet California’s 2015 deadline, and we hope that if a federal requirement emerges, it will look like California’s.”
When asked during AIM Expo whether Purdue will continue to use just RFID for units or move toward 2-D bar coding for either redundancy or as the sole data carrier, Ramsey wondered whether the pharma industry itself might make its way back to RFID, given some of the weaknesses of Data Matrix.
Pfizer and Purdue along with a handful of other forward-thinking pharma companies stepped up early to test item-level automatic identification, well before industry delivered standards and FDA made its intentions clear. These two companies have learned much as they perfected the use of auto-ID within their own four walls.
Ramsey advises pharma companies just getting started to not to get “hung up on costs when looking for solutions. Instead, look for solutions that will work.”
Serialization will take time, adds Govindasamy. “It is almost too late to start now for 2015,” he says. Ramsey agrees, saying that “supplier capacity to support manufacturers’ serialization efforts will get absorbed soon.”
Concludes Govindasamy: “Find partners for the long term.”