|From Information Silo to Bridge, Part 1: The State of UDI|
COVER STORY: PACKAGING
FDA mandates that
the labeling of all
devices include the manufacturer name, the product name,
and the place of
manufacture and directions for use.
As New York Times author Barry Meier recently observed, “even as the use of expensive devices, like artificial knees and defibrillators expands rapidly, patients and doctors get less information about products that are implanted in their bodies than consumers get on the safety and performance of their cars.”1
When it comes to devices, comparisons with other products are inevitable. For example, the Advancing Patient Safety Coalition recently compared medical devices to peanut butter. Specifically, in a letter to congressional members dated June 18, 2007, the coalition stated, “We can simply and quickly identify each and every jar of peanut butter that might have salmonella and remove them from store shelves in hours, but we cannot do that reliably today with potentially life-threatening defective medical devices.”2 Similarly, during a 2006 CDRH public meeting, Larry Kessler said that the medical device industry is “probably a decade or more behind the grocery industry” in terms of product tracking.3
The parallel nature of these quotes is indicative of the problem at hand: costly medical devices, which have the potential to significantly affect patient health, are tracked with less vigilance than the food supply. All too frequently, information specific to the device is lost, changed, or confused as it traverses the supply chain. Each participant finds the information that it needs, creates a system for handling those data, and passes the device to the next player in the chain, which, in turn, creates its own system. This generates a system of information silos, in which information is stored but not shared.
Figure 1. (click to enlarge) Entities in need of information about a given medical device. These supply-chain players physically handle the product.
The ideal system for sharing information would be more like a series of bridges that enable information to be effectively disseminated from one end of the chain to the other and back again. Figures 1 and 2 provide examples of the complex entities that need information about a medical device. Some handle the device, others do not; yet all require information.
FDA is looking to break down the silo mentality by creating a system that stimulates the flow of standardized information between the partners so that all parties can easily, quickly, and effectively have access to information that is accurate, relevant, and timely. As such, every manufacturer or packager of medical devices needs to know one acronym: UDI. A unique device identifier (UDI) system is likely to be implemented for many devices in the near future. It would be akin to what is required of drug manufacturers in the form of bar coded National Drug Code (NDC) numbers. In fact, it is probable that the new device system will surpass the requirements for drug products by requiring more information, such as lots or serial numbers, to be encoded.
Part 1 of this article examines the current state of industry and identifies the difficulties and benefits of implementing a UDI system. Part 2 will discuss required data, data management systems, and data carriers that can help create a functional UDI system.
In 2004, FDA published a final rule, titled “Bar Code Label Requirements for Human Drug Products and Biological Products” (21 CFR Parts 201, 314, and 601). The rule required that drug products and biologics (other than blood, blood components, and devices regulated by the Center for Biologics Evaluation and Research) have their NDCs encoded in a linear bar code format; specifically, either the Health Industry Business Communication Council (HIBCC) standard or that of GS1 (formerly the Uniform Code Council).
The impetus behind bar coding for drugs was to reduce medication errors, an issue that had received significant scrutiny after publication of the Institute of Medicine (IoM) study, “To Err is Human: Building a Safer Health System.”4
At the time of the rule, FDA requested public comment on whether the bar code requirements should include medical devices. Ultimately, the decision was made to exempt medical devices from the requirement, largely because of the lack of a standard, unique identifying system like the NDC.5 Nonetheless, the implementation of a system that employs a UDI and leverages automatic identification for medical devices promises wide-ranging improvements in the safety, security, and efficacy of medical devices. Those who are familiar with FDA's mission will realize that implementation of such a system is a very logical move.
Twenty-six members of the U.S. Congress recognized this principle and drafted a letter to FDA commissioner Andrew von Eschenbach. The letter, dated May 7, 2007, urged FDA to require a national UDI system “as soon as possible,” indicating that such a system would benefit manufacturers, improve patient safety, and promote the success of electronic health records and regional health information organizations.
Figure 2. (click to enlarge) Supply-chain players that do not in general physically handle the product, but nonetheless need information about a medical device.
Further evidence of the impending nature of a UDI system comes from a Manager's Amendment to the Medical Device User Fee and Modernization Act, put forth by Representative Darlene Hooley (D–OR) and approved by the House Energy and Commerce Committee on June 21. The amendment requires the Secretary of the Department of Health and Human Services to “promulgate regulations establishing a unique device identification system for medical devices requiring the labeling of devices to bear a unique identifier,” and is intended to “create an increased sense of purpose and urgency behind the FDA's efforts to implement this important system.”
In addition to the flurry of legislative and regulatory activity around the issue, legitimate actors in the healthcare supply chain, from the beginning to end, worldwide, are calling for consistent and accurate information about healthcare products. The list of UDI supporters is long, but among them are the Advancing Patient Safety Coalition, the Association for Healthcare Resource and Materials Management (part of the American Hospital Association), and the American Nurses Association. They want information to be readily available in a standard format that can be accurately read and used to improve patient health. These groups indicate that such a system would
The Current Structure for Tracking
Currently, FDA mandates that the labeling of all devices include the manufacturer name, the product name, and the place of manufacture and directions for use (although the requirement for directions is exempted in certain instances). The Safe Medical Devices Act of 1990, later amended by the Food and Drug Administration Modernization Act, requires implementation of a “tracking system” for devices that are likely to have failures that result in serious health consequences. These devices are identified when FDA issues an order for the tracking of a device. FDA does not require a specific method, but it does insist that manufacturers have standard operating procedures (SOPs) that will produce mandated information within timeframes specified in 21 CFR 821 Subchapter H. These requirements continue until the device is returned, destroyed, explanted, or the patient dies.
Regardless of whether tracking is ordered by FDA, there are inevitably instances, such as recalls, for which devices must be located. The current system enables the supply-chain actors (see Figures 1 and 2) to use, sanitize, and add layers of proprietary information in numerous, nonstandard ways that are not open to purchasers or shared across manufacturers. (With the devices that have tracking ordered by FDA, each manufacturer creates its own SOP to meet the spirit of the requirement and has limited control over the record keeping of downstream partners. With those that are not under order, information is shared even less consistently).
Frequently, varying numbers are assigned internally for use with differing systems—the silo effect. Sharing information from the chain, beginning to end, is difficult if not impossible. In some cases, even within a single hospital, databases from one department may not effectively exchange data with other departments. Additionally, if multiple ID numbers or bar codes are used as the device traverses the chain, it can be confusing for those needing information to know what is appropriate for their system.
Even with all that is depicted in Figures 1 and 2, they present an oversimplification of those in need of information. For example, consider a typical hospital structure. Within the hospital, purchasing, stockroom, patient care units, and the provider at the point of care all require varying information about the device being used. Beyond the point of care, information is needed as well (i.e., adverse events, recalled products, and billing and reimbursement). However, in the current system, not all players are able to accurately and quickly get the information that they require. The isolated charts, files, and databases may not be complete enough to provide needed information. The system creates various problems that a UDI system could avoid.
Determining Causal Paths. Under the current system, determining causal relationships is difficult, if not impossible. For instance, in 2005, a group of orthopedic surgeons who had noted a number of failing hip surgeries approached the Agency for Healthcare Research and Quality requesting information. The surgeons suspected that the device itself was failing, and were looking for data to evaluate the validity of their suspicions. The records, however, were not complete enough to prove or disprove their hypothesis. The absence of data precluded the opportunity to take decisive action regarding the procedure.6
Notices and Recalls. This was also the case when a hospital improvement alliance, Premier, attempted to locate a specific surgical mesh that was the subject of a Class I recall. Steven Stemkowski, manager in product management and marketing for Premier's comparative hospital data warehousing and healthcare consulting services group, explained the difficulties encountered by Premier during one of FDA's discussions of UDI in 2006.7 He indicated that of data examined from 400 hospitals, information from only about 40 was sufficient to identify the recalled product. Additionally, within the set of records that were complete, Premier estimated that 50 patients had the product administered to them after the recall had been issued.
There are obvious health and safety ramifications associated with these problems, but also significant costs. In its report to FDA, the Eastern Research Group (ERG) estimated that a fully functioning UDI system would enable hospitals to save half of the time that it takes at present to identify recalled devices. John Eyraud, executive vice president of ERG, estimates that this increased recall efficiency would save hospitals approximately $35 million.8
Evidence of Efficacy. Implementation of a UDI system not only offers the promise of improved notification and recall, but also the ability to assess patient outcomes. Marcel Salive, MD, of CMS has noted that a UDI system would provide the benefit of “comparative evidence.”9 By linking specific devices to patient outcomes, CMS could maximize the influence of the healthcare dollars that they spend. Not only does more specific, accurate, and timely information have the potential to improve the decision-making process of the payers, it can also play a key role in the trend toward evidence-based medicine and affect patient outcomes by providing evidence of causal linkages.
Obstacles to a Standardized System
There is general consensus among the supply-chain entities that UDI would prove beneficial in the ways previously listed. So, why does no such system exist? Simply put, because it is not easy. Decisions must be made while considering a variety of different factors that are briefly described in the following sections.
Product Disparity. Products of the device industry are extremely varied. They range from small, simple items to capital equipment. Some are meant for mass markets, while others are niche items. Some are packaged individually and others are packaged in boxes of hundreds or thousands. In addition, others come in kits that contain multiple devices in a single package. Products can be reprocessed, disposable, or used for a lifetime. Some carry a significant risk; others do not. Increasingly, devices are combined with drugs or biologics to form combination products. Such product disparity makes it difficult to decide what should be tracked and how the devices should be marked for UDI.
Rapid Pace of Auto-ID. A predominant change taking place in global healthcare value chains is the focus on technologies that promise a positive effect on the chain itself, namely faster movement with a lower expense and increased information accuracy. Automatic identification (auto-ID) technologies look to provide system benefits as a result of standardization, interoperability, and mass serialization. These principles are precisely what the UDI system is interested in achieving.10–12 The primary technologies to implement these changes are bar codes and RFID.
However, the fluid state of auto-ID technologies has also made standardization challenging. Decisions must balance the attractiveness of widely understood, easy-to-use, proven technologies with those that are emerging and promise unrealized benefits.
Provider and Patient Disparity. Additionally, although discussion here has primarily been limited to U.S. hospitals, devices are used in a variety of settings in an increasingly global marketplace. Ambulatory care settings, home health settings, telemedicine, and the like must be considered as decisions are made regarding the UDI system.
Integration with Other Systems. To achieve maximum benefits from the UDI, the system must fully integrate with a variety of other systems that are currently in development. These systems include substance registration systems, electronic health records, and health information technology systems. Additionally, the final system must integrate with supply-chain management systems such as enterprise resource planning and warehouse management systems.
And, as if these weren't enough, the global nature of the marketplace cannot be ignored. Regulations and standards regarding label requirements and auto-ID technologies must also play a role in the creation of the UDI. Bridges must span the globe to link all separated silos. Imagine having governments around the world work in concert to recall problematic products with a single information system.
Creating a UDI System
The current medical device supply chain consists of a series of information silos, and people are noticing. Startling comments are being made in written and oral formats; legislators, regulators, and citizens are calling for something to be done, and done quickly. Succinctly, UDI will happen. The questions are when and how?
As the multiple considerations are weighed and the UDI system is developed, it must be done, of course, while considering costs and benefits. With all of the complexities, it is unlikely that a one-size-fits-all solution will be implemented. A more likely scenario is one where devices that are most directly linked to patient safety will have more rigorous requirements for information and more aggressive time frames for implementation.
Regardless of what the device is or how it is used (and possibly reused), there are issues for both data and data carriers. These boil down to
In addition, decisions must be made regarding the infrastructure of the system that will interface with the data carrier to manage the data and inform the players in need of information, a data management system. Such data management systems, as well as data carriers, will be discussed in the second part of this article, scheduled for the November 2007 issue of MD&DI.
The substance, recommendations, and views set forth in this presentation are not intended as specific advice or direction to medical device manufacturers and packagers, but rather are for discussion purposes only. Medical device manufacturers and packagers should address any questions to their own packaging experts and have an independent obligation to ascertain and ensure their own compliance with all applicable laws, regulations, industry standards and requirements, as well as their own internal requirements.
The authors would like to thank Ulrike Kresya and the GS1 organization for providing funding to examine the many issues associated with global data standards for healthcare. The authors also wish to acknowledge the contributions of Jay Crowley, senior advisor at CDRH, and Ed Dzwill, manager of packaging technology at Johnson & Johnson, for their review of an early version of this document.
The authors have been working on a funded project with the GS1 organization to investigate the business case for global standardization for healthcare products.
Laura Bix is an assistant professor at Michigan State University 's School of Packaging . She can be contacted at email@example.com. Robb Clarke is an associate professor at Michigan State University . He can be contacted at firstname.lastname@example.org.
1. Barry Meier, “A Choice for the Heart,” New York Times (June 23, 2005).
2. Advancing Patient Safety Coalition, “Letter to Congressional Members” [online] (18 June 2007 [cited 20 August 2007]): available from Internet: www.aamc.org/advocacy/library/teachhosp/corres/2007/061807.pdf.
3. Larry Kessler, in transcripts of Center for Devices and Radiological Health Public Meeting on Unique Device Identification (25 October 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/UDI-Meeting.html.
4. Institute of Medicine, To Err is Human: Building a Safer Healthcare System, (Washington, DC: National Academy Press, 2000).
5. Eastern Research Group, “ERG Final Report: Unique Identification for Medical Devices,” (22 March 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/erg-report.html.
6. Jon White, “Panel Discussion on the Benefits and Costs of a UDI System,” in transcripts of Center for Devices and Radiological Health Public Meeting on Unique Device Identification (25 October 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/UDI-Meeting.html.
7. Steve Stemkowski, “Panel Discussion on the Development, Maintenance and Use of a Repository for UDI-Related Information,” in transcripts of Center for Devices and Radiological Health Public Meeting on Unique Device Identification (25 October 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/UDI-Meeting.html.
8. John Eyraud, “Panel Discussion on the Benefits and Costs of a UDI System,” in transcripts of Center for Devices and Radiological Health Public Meeting on Unique Device Identification (25 October 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/UDI-Meeting.html.
9. Marcel Salive, “Panel Discussion on the Benefits and Costs of a UDI System,” in transcripts of Center for Devices and Radiological Health Public Meeting on Unique Device Identification (25 October 2006 [cited 20 August 2007]); available from Internet: www.fda.gov/cdrh/ocd/udi/UDI-Meeting.html.
10. Healthcare Distribution Management Association (HDMA), Medication Errors and Patient Safety: The Bar Code Connection, (Reston, VA, HDMA Healthcare Foundation: 2004).
11. HDMA, Adopting EPC in Healthcare: Cost & Benefits, (Reston, VA, HDMA Healthcare Foundation: 2004).
12. ECRI, “Guidance Article: Radiofrequency Identification Devices” Health Devices 34, no. 5 (2005): 149–159.