Daily Rules, Proposed Rules, and Notices of the Federal Government
Submit written submissions in the following ways:
This rule is intended to substantially reduce existing obstacles to the adequate identification of medical devices used in the United States. By making it possible to rapidly and definitively identify a device and key attributes that affect its safe and effective use, the rule would reduce medical errors that result from misidentification of a device or confusion concerning its appropriate use. The identification system established under this rule would lead to more accurate reporting of adverse events by making it easier to identify the device prior to submitting a report. It would allow FDA, healthcare providers, and industry to more rapidly extract useful information from adverse event reports, pinpoint the particular device at issue and thereby gain a better understanding of the underlying problems, and take appropriate, better-focused, corrective action. The rule will also require dates on medical device labels to conform to a standard format to ensure those dates are unambiguous and clearly understood by device users.
The rule will fulfill the statutory requirement of section 519(f) of the FD&C Act (21 U.S.C. 360i(f)), which directs FDA to promulgate regulations establishing a unique device identification system for medical devices; this requirement was added to the FD&C Act by section 226 of the Food and Drug Administration Amendments Act of 2007 (FDAAA), Public Law 110-85.
In developing the proposed rule, FDA has been partnering with industry to conduct pilot tests to identify potential issues and generate feedback on the development of a UDI system. Throughout the pilot activities, labeler organizations from the medical device industry focused on identifying and understanding potential issues that would arise for labelers in implementing UDI and provided that feedback to FDA. The proposed rule reflects this industry input and the lessons learned from these pilot activities. FDA also solicited input through public meetings; a public workshop with stakeholders from the medical device industry, hospitals, payors and other stakeholders; and, a public request for information on a series of key questions related to the development of UDIs through which FDA received extensive input from the medical device industry and the broader healthcare community. FDA solicits comments on the proposed rule from all interested stakeholders, and is particularly interested in industry comment on whether the proposed approach reflects the lessons from the pilot activities.
Under the proposed system, the health care community and the public would be able to identify a device
By ensuring the adequate identification of medical devices through distribution and use, the rule would serve several important public health objectives—
This rule would require the label of medical devices and device packages to include a UDI, except where the rule provides for alternative placement of the UDI or provides an exception for a particular device or type of device. Each UDI would have to be provided in a plain-text version and in a form that uses AIDC technology. The UDI would also be required to be directly marked on the device itself for certain categories of devices, such as those that remain in use for extended periods of time and are likely to become separated from their labeling. The rule would require the submission of information concerning each device to a database that FDA intends to make public, to ensure that the UDI can be used to adequately identify the device through its distribution and use. The FDA database would not include patient information. The rule would also require dates on device labels and packages to be presented in a standard format.
The UDI system proposed by this rule builds on international regulatory cooperation activities and existing, internationally recognized standards relating to unique identification and data exchange. The rule would specify
Whenever a device must bear a UDI, the labeler of that device would be required to submit information concerning the device to FDA to facilitate the rapid identification of the device and the labeler, and to provide links to other FDA data. FDA will make this information available to the public through a variety of channels, including a new database, the GUDID.
The rule provides for appropriate exceptions and alternatives, ensuring that the costs and burdens are kept to a minimum.
A final rule would become effective in stages, over a period of seven years, to ensure a smooth implementation and to spread the costs and burdens of implementation over time, rather than having to be absorbed all at once.
This rule is intended to substantially reduce existing obstacles to the adequate identification of medical devices used in the United States. By providing the means to rapidly and definitively identify a device and key attributes that affect its safe and effective use, the rule would reduce medical errors that result from misidentification of a device or confusion concerning its appropriate use. The identification system established under this rule would lead to more accurate reporting of adverse events by making it easier to identify the particular device involved prior to submitting a report. It would also allow FDA, healthcare providers, and industry to more rapidly extract useful information from adverse event reports, pinpoint the particular device at issue and thereby gain a better understanding of the underlying problems, and take appropriate, narrowly-focused, corrective action.
The rule will fulfill a statutory directive to establish a unique device identification system. Section 226 of FDAAA amended the FD&C Act to add new section 519(f), which directs FDA to promulgate regulations establishing a unique device identification system for medical devices: “
Under the system that would be established by this proposed rule, two tools would be used together to identify a device: A UDI on the label and packaging of a device (represented both in plain text and through automatic identification and capture technology), and a new database, the GUDID, containing device identification information for each UDI. Certain devices, such as those that remain in use for extended periods of time and are likely to become separated from their labeling, would also be directly marked with a UDI, allowing accurate identification even when the device is no longer accompanied by its label or package. The principles behind the rule's requirements and exceptions regarding UDI labeling and the GUDID are discussed in more detail in section I.B of this document.
By requiring adequate identification of medical devices through distribution and use, the rule would serve several important public health objectives—
Device-related medical errors are a serious problem. The presence of a UDI that is linked to identifying information in the GUDID database will facilitate rapid and accurate identification of a device, thereby removing a cause of confusion that can lead to inappropriate use of a device (e.g., confusing a sterile version or model with a version or model that is not sterile and which requires sterilization prior to use, or failing to recognize that a particular device is the subject of a recall or enforcement action). The presence of AIDC technology as part of a UDI would make it possible to “scan” a device at a patient's bedside and rapidly compare the device attributes reported to the GUDID with the medical order and the patient's history, thereby improving the accuracy of device use and providing greater assurance that a device is appropriate for the patient.
The proposed system would allow anyone to use a device's UDI to look up identifying information in the GUDID
The rule would require some devices to be directly marked with a UDI, so that it will always be possible to positively identify the device, regardless of how long the device remains in use. These devices, by their intended or customary use, are typically separated from the labeling that accompanies delivery of the device to users:
• An implantable device;
• A device that is intended for more than one use and to be sterilized before each use; and
• Stand-alone software.
These devices involve unique risks to patients, and consequently it is particularly important to ensure the adequate identification of such devices throughout the entire product life cycle. For example, a device that is intended for more than one use, but which must be sterilized before each use, might be used over several years; during that time, the device package, with its label and any package insert, might be lost, leaving the user of the device uncertain as to whether the device needs to be sterilized, or just given a routine cleaning, and if sterilization is required, what type of sterilization process should be employed. The same is true for implanted devices and stand-alone software—loss of the device package and accompanying labeling can leave the user uncertain as to how to use the device, how to monitor its performance, or what actions should be taken in particular circumstances.
FDA intends to provide Internet access to all data in the GUDID. Furthermore, once data concerning a device has been submitted to the GUDID, it will remain available long after production and marketing of the device has ceased. The GUDID will include information important to the identification of the device, but will not include patient information.
UDIs, particularly when provided through AIDC technology, would allow rapid and accurate data acquisition, recording, and retrieval. The use of UDIs in patient records, particularly electronic patient records, would help avoid confusion among similar devices during an extended treatment period and where more than one health care provider is involved in the administration of a course of treatment. The use of UDIs in computerized physician order entry systems will help ensure that the intended device will be used in the treatment of a patient, rather than some similar device that may not fully meet the requirements of the health care professional who ordered the use of the device.
An essential prerequisite to resolving adverse events is the timely and precise identification of the particular device or devices that may have a connection with an adverse event. The proposed UDI system would make this possible. From 2005 through 2009, FDA received an average of more than 492,000 adverse event reports involving devices each year. During this 5-year period, more than 17,700 reports involved a death, and more than 283,000 reports involved an injury.
Because reports come from multiple sources—manufacturers, device user facilities, importers, and voluntary reports from physicians and other concerned individuals—we often receive more than one report of a particular death or injury. Reviewing a significant number of reports, seeking essential missing information, and resolving inconsistencies among reports are major challenges, particularly when trying to identify recurring problems involving a particular device. Although we do not have precise statistics, many initial reports do not provide a precise identification of the specific device the report concerns and require extensive FDA follow-up to identify the specific device involved. The inclusion of UDIs in adverse event reports would lead to greater accuracy in reporting, and eliminate uncertainty concerning the identity of devices that are the subject of reports.
The inclusion of UDIs in adverse event reports would allow manufacturers and FDA to more rapidly review and analyze reports and identify the particular device at issue. This would permit more rapid isolation and identification of the underlying problems, and development of an appropriate solution to a particular concern. UDIs would also allow FDA, manufacturers, and the healthcare community to more accurately target safety alerts, recalls, and other corrective actions on the specific devices that are of concern. UDIs, particularly when provided using AIDC technology, would allow device user facilities and health care professionals to identify those devices more rapidly and with greater assurance, and prevent further patient exposure. At the same time, devices not implicated by the problem would be less likely to be “swept up” in an over-broad attempt to remove potentially hazardous devices.
Currently, locating all devices subject to a recall is a time- and labor-intensive process. Manufacturers, distributors, and healthcare facilities often do not know how many recalled devices they have in stock, do not know exactly where those devices are located, and are sometimes uncertain which of several similar devices is the subject of a recall. Consequently, delays in identifying recalled devices can result in the continued use of those devices on patients in a variety of settings (e.g., hospitals, long-term care facilities, homecare environments) and involves an increased risk for patient harm. A device labeled with a UDI can be identified more rapidly and with greater precision than a device that does not bear a UDI. The use of AIDC technology, such as a bar code, would allow increased use of automation to speed efforts to identify specific devices that are the subject of a recall. The more rapidly a recall is implemented and completed, the more rapidly the risks presented are reduced and eliminated.
A class 1 recall is the most serious type of recall, and involves a situation where there is a reasonable probability that use of the device will cause serious injury or death. It is particularly important, therefore, that a class 1 recall be completed as rapidly as possible. The absence of a system that allows rapid
Often, a recall must cite more than one descriptor to identify the specific devices subject to the recall. For example, a September 22, 2009, class 1 recall of the Penumbra, Inc., Neuron 6F 070 Delivery Catheter required reference to both the product catalog number and the lot number to determine whether a particular catheter was subject to the recall, and a June 17, 2009, class 1 recall of Abbott Vascular-Cardiac Therapies/Guidant Corp. POWERSAIL Coronary Dilatation Catheters referred to product designation, product number, lot number, and expiration date. Recalls would be expedited and simplified if a single descriptor, such as the proposed UDI, could serve to adequately identify all devices.
There is no uniformity in the placement or formatting of the descriptors presently used to identify devices, and no assurance that different companies are using a given term in the same way. The inconsistency in methods used to identify a recalled device complicates efforts to identify such devices that remain in possession of a patient, physician, or in a hospital's inventory and to complete the remedial action that would mitigate or eliminate the risk of further harm. These problems would be significantly reduced by the presence of UDIs on the labels and packaging of devices and the inclusion of UDIs in recall notification information. The inclusion of AIDC technology, such as a bar code or a RFID tag, would permit inventories to be checked more rapidly and would result in the more accurate detection and removal of recalled devices.
By citing a device identifier, or a range of UDIs, FDA would be able to more precisely focus a safety alert, public health notification, or other communication on the particular device that is the subject of the alert, eliminating confusion with similar devices. Health care professionals and patients would be able to take responsive action more rapidly, and users of similar devices that are not the subject of the safety alert would not be faced with the uncertainty of not knowing whether they have been exposed to, or are affected by, a problem or risk.
While not required, inclusion of device identifiers in informational materials, such as package inserts, could provide a quick and useful means for patients and health care professionals to obtain additional information concerning a device, without having to provide that information in the document. This could allow the document to focus on its important core messages without the distraction of greater complexity, while a reader who wants those additional details could use the UDI to obtain information from the GUDID.
FDA has concluded that a UDI system has the potential to provide additional benefits. For example, we expect UDIs could be used by other Federal agencies, such as the Centers for Medicare and Medicaid Services, the National Institutes of Health, the Centers for Disease Control and Prevention, the Department of Defense, the Department of Homeland Security, and the Department of Veterans Affairs, for a wide variety of purposes, ranging from management of the Strategic National Stockpile, inventory management, and the provision of high-quality medical services. Other benefits include facilitating the development of more useful electronic patient records by allowing providers to electronically capture and record important information concerning the use of a device on a patient. UDIs could help identify similar devices in the event of a shortage, and could reduce the potential for injury from counterfeit devices by offering a better way to detect a counterfeit product and remove it from the market.
The UDI system would provide a basic infrastructural element, which would allow unambiguous identification of medical devices throughout their lifecycle and would provide the foundation for a host of benefits. These may include improved device traceability, improved postmarket surveillance, and better security of devices through more effective detection and removal of counterfeit devices, and other
Through our work with the Global Harmonization Task Force (GHTF) and foreign regulatory partners, we envision that the UDI system would support global public health initiatives with which FDA is concerned, including more efficient and effective cross-border identification of devices, adverse event reporting and postmarket surveillance, and would improve our ability to communicate and respond to issues and concerns about devices used not only in the United States, but in other nations as well.
The full benefits of UDI require that hospitals and other healthcare facilities concurrently adopt information technology (IT) to fully realize the enhanced ability to identify devices throughout distribution and use. In order to realize its full potential benefits, UDI users must be able to store UDI information in various administrative, clinical and payment information systems, including EHRs. Though many such systems exist today, changes will need to be made in the systems to accommodate UDI.
The use of electronic health technology to reduce medical errors in healthcare facilities would require the use of scanners (many of which are already in place) and standard operating procedures for using newly developed systems that link critical patient information (such as latex sensitivity) with specific medical device information. Hospitals and other health-care facilities will choose to make investments in the new technology and methods if they expect it to be a cost-effective method to reduce errors and improve patient safety involving medical devices.
Putting a standardized unique device identifier on a device label is one step in creating systems that could reduce device related medical errors. The proposed rule would create a platform that would enhance the value of the new electronic health technologies and thereby encourage their development. But the proposed rule does not require hospitals and other health care facilities to make these changes.
In developing our proposed system for identification of devices, FDA first developed several general objectives, or principles, that we then applied throughout the drafting of our proposed rule. Each of these principles is identified in this section I.B, with a brief discussion of how they are resolved in the proposed rule.
The healthcare community needs to identify a wide range of medical devices in every medical specialty. When fully phased-in, the rule will apply to all three device classes; however, we are proposing to exempt class I devices from production identifiers and proposing full exceptions from UDI labeling and data reporting for certain very low risk devices and other categories of devices; see proposed §§ 801.30, 801.35, and 801.128(f). Although we are not aware of compelling reasons for other exemptions based on the device class or medical specialty, for example, we seek comments on this issue.
Basing the UDI system on existing, accepted standards ensures that all UDIs will be unique, broadly compatible, and broadly accepted for use by the U.S. healthcare community and in international commerce. By incorporating these existing standards into our proposed system, we avoid the confusion, inconsistency, and inefficiency that would result if every labeler created their own device identifiers without regard for the needs of the healthcare community. Therefore, the UDI system we are proposing would incorporate by reference four international standards: International Organization for Standardization/International Electrotechnical Commission (ISO/IEC) 646:1991, Information technology—ISO 7-bit coded character set for information interchange; ISO/IEC 15459-4:2006(E), Information technology—Unique identifiers—Part 2: Registration procedures; ISO/IEC 15459-4:2008, Information technology—Unique identifiers—Part 4: Individual items; and ISO/IEC 15459-6:2007, Information technology—Part 6: Unique identifier for product groupings. See proposed § 830.10. In addition, all widely-used AIDC technologies—e.g., bar codes, RFID tags, and near-field communication are based on established, broadly-supported standards. (Ref. 1) A multiplicity of nonstandardized systems would impose excessive costs on device user facilities and others, would provide no assurance that identifiers would be unique, would run counter to efforts to achieve international harmonization with regard to the identification of devices, and would greatly complicate FDA efforts to identify and resolve adverse events and other problems involving devices.
FDA is aware of two existing device identification systems that are based on the ISO/IEC standards discussed in the preceding paragraph. The International not-for-profit association known as “GS1” operates a system that uses a Global Trade Identification Number (GTIN) to identify a device; GS1 also operates the Universal Product Code (UPC) system that is used to identify most items sold by retail establishments in the United States. (Ref. 2) The Health Industry Business Communications Council (HIBCC) operates a system that encodes an identifier in a Health Industry Bar Code (HIBC) to identify a device. (Ref. 3) We believe roughly 35 to 50 percent of all medical devices used in the United States are already labeled with device identifiers that conform to one of the systems operated by these two organizations (a 2005 ECRI Institute report, “Automatic Identification of Medical Devices,” cited survey data suggesting bar codes were currently found on 25 percent of class I devices, 44 percent of class II devices, and 50 percent of class III devices) (Ref. 4). These existing systems are providing valuable services to device user facilities (hospitals, nursing homes, and other facilities) and to health care professionals. These systems have proven to be successful in creating unique identifiers that are in widespread use in systems used by hospitals, healthcare professionals, and industry.
Because these existing systems include tightly-integrated functions that go far beyond simply identifying devices—functions such as inventory management and enabling commercial transactions that are not part of FDA's public health responsibilities and are outside our statutory authority—FDA believes it would be inefficient and counterproductive to try to replace the existing systems with a single, FDA-designed system of device identifiers. Because any FDA system would necessarily have a narrow scope limited to the adequate identification of devices, labelers would have to continue to use the existing systems as well as the FDA system, which would result in duplication of effort, substantial
We have honed our proposed data submission requirements to minimize overlap and avoid inconsistency with other existing FDA regulatory requirements, such as establishment registration and device listing. We are proposing to require the submission of fewer types of data than those identified and discussed in the public meetings (Ref. 5) that influenced development of this proposed rule. See proposed § 830.310. We are requesting comments on whether we have adequately minimized overlap and inconsistency, and whether we should require or permit the submission of additional data that may be useful to the healthcare community.
The proposed rule would require each UDI to be provided in both a plain-text form and a form that uses AIDC technology. See proposed § 801.45. FDA would not require use of any particular technology for the AIDC form of the UDI. The system would permit the use of any type of bar code, RFID tag, near-field communication, or any other technology, whether existing at the present time or developed in the future. This would allow for technological evolution and advancement without prior FDA approval. FDA expects that a new technology would be deployed only after considerable consultation among issuing agencies, device user facilities, healthcare professionals, and device manufacturers, and we believe such decisions are best left in the hands of the healthcare community.
We have taken care to avoid conflict and minimize overlap with existing regulatory requirements, and we have included several conforming amendments to existing regulatory requirements to ensure UDIs are integrated in our regulatory processes wherever appropriate and feasible. For example, Part 810—Medical Device Recall Authority, Part 820—Quality System Regulation and Part 821—Medical Device Tracking Requirements.
Pursuant to the proposed tiered effective dates, UDI requirements would be phased in over seven years following publication of a final rule (see table 7 of this document). This would allow all participants—FDA, industry, the health care community, and other government agencies—ample time to become familiar with and phase-in the rule's labeling and data submission requirements. This approach also provides FDA the opportunity to identify unforeseen weaknesses or problems in our implementation of the UDI system and to make appropriate mid-course corrections within the scope and authority of this rule, if finalized. We are proposing to phase in the rule's requirements by class because this will allow us to focus first on devices that have higher risks.
The proposed rule would allow for accreditation of multiple issuing agencies, see proposed § 830.100, so that the varying needs of labelers and users of different types of devices can be met by different systems with differing levels of complexity and function. Because all issuing agencies would have to employ systems based on the same technical standards, and would have to meet the same accreditation requirements, each system would still be broadly compatible with other systems. Furthermore, all systems would employ the FDA-administered GUDID database, which would serve as the single authoritative source of information for the positive identification of any device labeled with a UDI. We will maintain a list of all FDA-accredited issuing agencies on our Internet site.
Oversight is necessary to ensure that all device identifiers are unique and meet the proposed requirements, and that all system users are treated fairly. FDA is proposing to require that any organization that wishes to issue UDIs be accredited by FDA. See proposed § 830.20(a). We have included accreditation criteria and information submission requirements designed to ensure that only a well-qualified organization that would issue identifiers that comply with the proposed rule would be permitted to serve as an issuing agency. See proposed §§ 830.100 and 830.110.
Where possible, we have included reasonable flexibility in our proposal. For example, certain categories of devices would be excepted from UDI requirements, see proposed § 801.30, and labelers may request an exception or propose an alternative that would, for example, provide for more effective identification of a device, see proposed § 801.35. Direct marking requirements would apply only to certain narrow categories of devices and there would be some flexibility in how this requirement may be satisfied, see proposed § 801.50. We seek comment on whether these flexibilities achieve the appropriate balance.
We seek to do this in two ways. First, a business can choose to use any system provided by any accredited issuing agency, which will give the labeler a choice among a range of services at a range of fees. We anticipate that the participation of multiple issuing agencies will also lead to competition that will help ensure fees are reasonable. Second, FDA may act as an issuing agency if we find that a significant number of small businesses will be substantially harmed by the fees assessed by all accredited issuing agencies, see proposed § 830.200. If FDA acts as an issuing agency, any business would be permitted to use the FDA system and, under current law, there would be no fee, see proposed § 830.210. We expect this provision will encourage issuing agencies to be sensitive and responsive to the needs of small businesses.
It is important to understand that a UDI is simply a numerical or alphanumerical code and on its face is not itself intended to communicate any information directly concerning a device; you would not, for example, be able to parse out a segment that indicates that the device is a cardiovascular device, or that the device is packaged sterile, or that the device is marketed under a particular FDA premarket submission. Instead, the UDI would function as a reference number allowing you to find data concerning the device in an FDA database, the GUDID. The real value of a UDI is derived from its connection to corresponding information identifying the version or
In order to serve the public health purposes discussed in section I.A of this document, the UDI system requires a GUDID that is freely and easily accessible to all—hospitals and other device user facilities, health care practitioners, patients, other government agencies, academia, industry, and the general public. None of the information that we are proposing to collect would constitute trade secret information, confidential commercial information, or personal privacy information, and public disclosure of this information would not be prohibited. Open access to the GUDID would also encourage the integration of UDI data into healthcare delivery support systems, electronic medical records, and procurement, inventory management, and accounting systems, and would allow those systems to work together more effectively and efficiently.
Since the issuance of the final bar code rule, various entities have asked that we revisit the issue of bar coding for medical devices to improve patient safety, quality of care, and cost effectiveness of health care, e.g., by improving delivery and supply chain efficiency. In response to this, in 2005 FDA met with various stakeholders, including device manufacturers and distributors, hospital associations, and other Federal agencies to solicit information and comments about employing a uniform system for the unique identification of medical devices. As a result of these meetings, FDA believes the majority of stakeholders support the development of a uniform system of unique identifiers as a way to improve patient safety and recognize other ancillary benefits such as better management of the purchase, distribution, and use of medical devices. However, there were a variety of experiences and opinions about how best to implement such a system. In 2006, we commissioned a report from Eastern Research Group, Inc. (ERG), concerning the benefits, costs, and issues with developing and implementing a UDI System. (Ref. 6) Thereafter, we published a notice in the
We used the comments responding to the August 2006
FDA held a public workshop on February 12, 2009, to discuss issues relating to establishment of a UDI system (see 74 FR 2601, January 15, 2009). (Ref. 5) We asked device identification standards organizations to discuss the development and use of UDI standards, including the use of production identifiers. We asked device manufacturers to discuss the use of standards and the marking of devices with UDIs. We also discussed the potential development and use of a UDI database in general and with respect to particular attributes, as well as issues relating to implementation of a UDI system by interested stakeholders (e.g., distributors, hospitals, payors). We asked device manufacturers to describe their current practices for applying standards to medical devices, including identifiers on medical device labels, and managing medical device identifier data. We also requested information regarding the difficulties and costs involved in adding a UDI to a device's label, including effects on manufacturing and labeling processes and expected capital and operating costs. We asked device user facilities (hospitals, nursing homes, and clinics) to describe how a UDI system could be used, the costs involved, whether a UDI system would require any change in operations, and how UDIs would affect adverse event reporting and recall management. We asked all interested persons to submit comments, including answers to any of these questions, to a regulatory docket, FDA-2008-N-0661, CDRH 200866—Unique Device Identification System; Public Workshop. Comments received by the docket may be reviewed at
We carefully reviewed and considered all comments during our development of this proposed rule.
The core requirements summarized here provide context for the more detailed discussions that follow:
• Proposed § 801.18 provides for standardized formatting of dates on medical device labels, eliminating any possibility of confusion from date formats that might be interpreted in more than one way.
• The labeler of each device would be responsible for meeting labeling and data submission requirements under this proposal. The labeler would, in most instances, be the manufacturer of the device. The term “labeler” is defined at proposed § 801.3, and is discussed in section II.B.1 of this document.
• Unless the device is excepted, the label of a medical device, and a device package, marketed in the United States would be required to bear a UDI; this requirement would be phased in over 5 years. See proposed § 801.20.
• The UDI would have to be provided in two forms: easily-readable plain-text and AIDC technology. See proposed § 801.45. These two forms ensure that the UDI of a device would be readily discernable to patients and health care professionals and to automated systems used to identify and manage devices.
• The proposed rule provides several categorical exceptions, proposed § 801.30, as well as case-by-case
• Direct marking would be required for certain categories of devices, with exceptions. For each device subject to direct marking, this requirement would go into effect two years after the base UDI labeling requirement goes into effect for that device. See proposed § 801.50.
• Whenever a device must be labeled with a UDI, the labeler (the person who causes the label to be applied to the device) would have to submit data concerning that device to the GUDID database. See proposed § 830.320. This information would have to be submitted no later than the date the label of the device must bear a UDI, and would have to be updated when changes occur. See proposed § 830.330. Exceptions are identified in the detailed discussion of part 830. This data would be freely available to the public and would provide the information necessary to identify a device labeled with a UDI.
• UDI labeling requirements would also apply to—
○ Certain combination products;
○ In most instances, to the device constituent parts of combination products;
○ Convenience kits; and
○ A device included in a convenience kit, except for a single use device.
The terms “combination product” and “convenience kit” are defined at proposed § 801.3 and are discussed in section II.B.1 of this document.
• UDIs would be issued under systems operated by FDA-accredited “issuing agencies” and conform to certain international standards, incorporated by reference at proposed § 830.10. A different UDI would be required for each version or model of a device. These terms are defined at proposed § 830.3.
• In order to provide for efficient implementation of this rule, we propose to phase in its requirements over several years. Table 7 of this document, Effective Dates of UDI Regulatory Requirements, summarizes how we would phase in the requirements proposed in this rule.
Part 801 (21 CFR part 801) provides FDA's general medical device labeling requirements. All devices are subject to subparts A through E of part 801, while subpart H provides special requirements for specific devices; subparts B, F, and G are presently reserved. FDA provides additional labeling requirements in subpart B of part 809 that apply only to in vitro diagnostic products. FDA is proposing amendments to part 801 to provide UDI labeling requirements for devices. The changes we are proposing to part 801 provide a new definitions section, see proposed § 801.3; a new provision standardizing the format of dates provided on medical device labels, see proposed § 801.18; new subpart B, Labeling Requirements for Unique Device Identifi