PREMARKET NOTIFICATION AS A KEY ELEMENT OF US REGULATION OF MEDICAL DEVICES
To gain a foundation on which to build its discussions, the committee commissioned two former Center for Devices and Radiological Health (CDRH) staff to draft a background paper on the 510(k) or premarket notification process, explaining the concept of substantial equivalence and discussing the scientific integrity of the process, its strengths, its weaknesses, and its flexibilities.1 Coauthor Philip J. Phillips, of the Phillips Consulting Group and formerly the deputy director for science and regulatory policy at CDRH, presented an overview of the paper that he drafted in collaboration with Larry Kessler, of the University of Washington School of Public Health, who previously served as director of the CDRH Office of Science and Engineering Laboratories and director of the CDRH Office of Surveillance and Biometrics.
Breadth of FDA Responsibility
FDA regulates a complex and broad array of medical devices and device-related products. The agency has authority over general-purpose articles, such as glassware and reagents, that are used in laboratories; although these are not labeled specifically for medical purposes, their use is medical. FDA regulates device components and parts and the accessories that go with them. Even cases for holding spectacles and contact lenses technically are medical devices. Some devices are custom-built to meet the specific needs of
The complete commissioned paper is available as Appendix C.
an individual patient or an individual health-care practitioner. A new industry that has emerged in the United States reprocesses single-use disposable devices and reintroduces them into interstate commerce as new, and these are subject to the same FDA regulatory review processes as the original devices. Similarly, there is an industry that remanufactures durable medical equipment, including some class III medical devices, and reintroduces them into interstate commerce.
The system of identifying generic types of devices and classifying them is critical for the successful regulation of such diverse products. It is an efficient means for the agency to allocate its resources and maximize FDA’s effect on public health.
Safety and Effectiveness
The same definitions of safety and effectiveness apply to all medical devices, whether they are in class I, class II, or class III. How safety and effectiveness are determined is outlined by statute and regulation. Factors that are taken into consideration include the intended patient population, the conditions of use that are communicated through labeling and advertising, the balance of health benefits and risks associated with use of a device, and the reliability of the device.
For FDA to allow a product to be marketed in the United States, there must be
Reasonable assurance of safety. The probable benefits derived from the use of the product must outweigh the probable risks; there must be an absence of unreasonable risk of illness or injury associated with the use of the product.
Reasonable assurance of effectiveness. Use of the product must yield clinically significant results in a significant portion of the target population.
The approach to ensuring safety and effectiveness depends heavily on a device’s classification. Class I devices are subject to general controls, class II to special controls, and class III to premarket approval.
Phillips pointed out that premarket notification is the general control that gives the agency an opportunity to determine whether a new product falls into an existing generic type of device or should be considered as a different generic type (it could be in class III, or it could be the subject of a de novo or reclassification effort and be in class I or class II).
Generic type of device is defined by regulation and means a group of
devices that do not differ substantially in purpose, design, materials, energy sources, function, or any other feature related to safety and effectiveness. The agency must conclude that similar regulatory controls are sufficient to provide reasonable assurance of safety and effectiveness. If a new device is put into class I in accordance with an existing regulation, the general controls applicable to that generic type of device ensure its safety and effectiveness. If a device is put into class II, the general controls and the applicable special controls that have been developed ensure the product’s safety and effectiveness. More than 1,800 generic types of devices are the subjects of classification regulations.
In the early days, the initial 510(k) submissions were simple to review, Phillips said, because there were no substantial differences between new devices going to market and older devices that had been on the market. However, the challenges associated with substantial equivalence quickly emerged. There was an expectation at the time that the Bureau of Medical Devices would start to develop performance standards for class II medical devices, but FDA did not have the resources to do it. Differences between new devices and old devices rapidly became more pronounced, and there were changes that would affect the intended use of a device and technologic changes. Another factor that challenged the system was that any kind of substantial shift toward using the premarket approval (PMA) process more frequently would seriously drain FDA resources. The PMA process was so burdensome and demanding that there was pressure in the agency to bolster the concept of substantial equivalence, which would allow products to go to market quickly with appropriate safeguards. Substantial equivalence evolved to compensate for regulatory realities.
A device is “substantially equivalent” to a predicate device if it has either of the following two groups of characteristics.
has the same intended use as the predicate and
has the same technologic characteristics as the predicate.
has the same intended use as the predicate, and
has different technologic characteristics, and the information submitted to the agency does not raise new questions of safety and effectiveness, and
is demonstrated to be at least as safe and effective as the legally marketed device.
Intended use, as defined by regulation (21 CFR 801.4), refers to the intent of persons legally responsible for the labeling. However, this particular definition is not geared to the premarket determination of intended use. It is a postmarket regulation that gives FDA the authority to determine whether someone is distributing a product and promoting it in a way that is consistent with what FDA views as its legal intended use. In the context of a 510(k) submission, when FDA considers issues of intended use, it focuses primarily on indications for use, on whether a product is intended for use by licensed health-care practitioners or laypersons, and on whether the product is intended for single or multiple use. Changes in the indications for use present the biggest regulatory challenges, and the related agency decisions can be the most difficult to explain.
Indications for use is not defined specifically within the confines of 510(k). For PMA, it refers to the disease or condition that a product is used for and the patient population for which the product is intended. For 510(k) devices, indications for use may be functional (that is, what the device does). In the case of very simple devices, such as scalpels, it is not necessary to name all the patient populations or all the diseases and conditions in which the devices may be used. It is not reasonable to describe use to that level.
Two basic changes in indications for use appear in labeling. One is expansion of the patient population. For example, cardiovascular diagnostic catheters may be proposed for use in cardiac ablation procedures. In that case, a cardiovascular mapping catheter that is used for diagnosis is a class II medical device, but the same product promoted and labeled for therapeutic purposes is a class III device subject to PMA requirements.
The other is a change from general to specific indications for use and vice versa. For example, carbon dioxide lasers, which are very old devices that have not changed much from a technologic standpoint, have had considerable changes in indications for use over the years. Carbon dioxide lasers that are labeled for such procedures as photocoagulation, cutting, and ablation of soft tissue are now being proposed specifically for the removal of tattoos. A guidance document available for general and specific intended uses, Phillips noted, outlines the criteria that FDA uses in making a decision about intended use.2
Phillips suggested that the IOM committee consider the following definition to understand FDA’s 510(k) approach to intended use:
Guidance is available at http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm073944.htm.
In the context of 510(k), intended use is a regulatory concept that determines the boundaries of use for a generic type of device, and is constructed to encompass the widest breadth of use where the regulatory controls for the device type continue to ensure safety and effectiveness.
According to the statute, in the face of different technologic characteristics, the agency asks whether the differences raise new safety or effectiveness questions. If the answer is yes, the device is found to be not substantially equivalent. The agency has not applied the statutory language exactly but asks whether the new technology raises new types of questions about safety and effectiveness. The reason that that word types was added to the program guidance is that any change in technology can raise a new question. By asking about new types of questions, the agency has greater discretion in making some of its regulatory decisions, Phillips said. Similarly, when considering whether a product is as safe and as effective as another product, the agency asks whether the risks that are inherent in the new technology can be mitigated. In other words, if the technology and the effect of the change on the actual use of the product are well understood, the agency does not automatically reclassify the device to make it subject to PMA; instead, it looks to mitigate risks.
An example is wireless technology. During the last 15–20 years, many devices have been updated to function through wireless mechanisms. Rather than reassess every device, the agency provided requirements for performance testing and labeling to ensure that the risks associated with each wireless-technology device had been mitigated.
Scientific Integrity of the 510(k) Process
Regardless of a common misperception, a 510(k) submission always contains data, Phillips said. The manufacturer of a product that has the same intended use and the same technologic characteristics as a predicate device must submit descriptive data, including side-by-side comparisons of the new device and the legally marketed device with which it is compared, and performance data when descriptive data do not ensure performance. If there are differences in intended use between a new device and a predicate that do not constitute a “new intended use” and there are different technological characteristics, the submission will contain descriptive and performance data, as well as bench, animal, and clinical data to assess those differences and mitigate any associated risks.
The 510(k) process is not only for new manufacturers and new devices that are coming to market, Phillips said. It also applies to existing manufacturers that are modifying their already-marketed devices. Change is inevitable. Whenever there is a change that could substantially affect safety
and effectiveness, the manufacturer must obtain a new clearance from FDA before it can introduce the product into commercial distribution. It is the manufacturer’s responsibility to know when modifications and changes require submission of a 510(k), and there is FDA guidance for industry in addressing this.3
Ultimately, the integrity of the 510(k) process requires verifying that FDA is receiving the necessary submissions and that in all cases the changes have undergone the necessary verification and validation.
Strengths, Weaknesses, and Flexibilities
A key strength of the 510(k) process is FDA’s ability to apply knowledge that has been gained in the review of one 510(k) submission to the review of later 510(k) submissions. That does not mean that confidential, proprietary information is shared externally. Rather, the agency can use it internally so that it does not have to ask redundant questions.
A weakness in the program is the lack of device-specific guidance documents and use of special controls for many devices. The agency could do a much better job of recognizing international and national consensus standards and of leveraging them in the review processes, Phillips said. And it is difficult to address safety and effectiveness issues related to legally marketed devices—a problem that is not peculiar to the 510(k) process but is also relevant to PMA products.
The agency has flexibility in that it can ask for whatever information its reviewers believe is necessary. The agency can also use the work of standards-development organizations in developing voluntary consensus standards. There is an opportunity, for a class I or II medical device to go to one of the dozen or so FDA-accredited third-party review organizations. Some may consider that having outside reviewers do FDA’s work poses a vulnerability but, Phillips said, the work product comes to the agency for final review of a submission.
Premarket Notification vs Premarket Approval
Phillips stressed that the premarket notification, or 510(k), process is a classification process, whereas PMA is a determination of safety and effectiveness that leads to approval. Thus, the programs cannot be directly compared. For devices in class I and class II, safety and effectiveness are ensured by conforming with all the general controls. There must be verification that companies are in conformance with the controls and vigilance in
Guidance is available at http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm080235.htm.
monitoring the performance of a device once it is in commercial distribution to determine whether it is performing as it was designed to. For the 510(k) process, devices are cleared for marketing, not approved, and devices may not be marketed as “approved by FDA.”
PMA of class III devices involves device-by-device assessment. From concept to obsolescence, class III medical devices are subject to FDA conditions and requirements that their manufacturers are obliged to meet. These devices do receive approval by FDA.
There is tremendous competition in the medical device industry, Phillips said. There is an emphasis on cost containment, and the practice of medicine places demands on the industry. Those environmental factors prompt the continual evolution of medical devices with regard to intended use and technology. That results in substantial regulatory challenges. An example is combination products, such as medical devices that are associated with a drug or biologic.
Phillips reiterated that ensuring the safety and effectiveness of medical devices is a complex task and that the 510(k) process is only one part of it. He urged the committee not to be misled by the concept of substantial equivalence; the issues are broader. The 510(k) process has strengths and weaknesses, but it makes important contributions to public health. Whether 510(k) is maintained as is, changed, or abandoned, the system of medical device regulation must be flexible enough to accommodate the constant and rapid change associated with the medical device industry and must have the integrity to withstand criticism. It is difficult enough for FDA staff to make decisions without having to operate in an environment in which people are second-guessing the underlying regulatory framework in which the decisions are being made.
Given adequate resources, Phillips said, changes could be made in how devices are classified, reviewed, and managed by FDA. The current framework in which the agency has to operate is sound. There are, however, major resource issues that create gaps, for example, in verifying compliance with general controls. Many manufacturing facilities of class I and class II devices are not inspected as frequently as they should be for good manufacturing practice. With regard specifically to 510(k), there needs to be additional emphasis on developing guidance for industry or special controls and a roadmap for bringing new and innovative products to market and removing the guesswork. The agency does not have the resources needed to look at the
500 or 600 class II medical devices each year and develop special controls for all of them or to develop guidance for all of them. Phillips stated that the system in its current form is fundamentally sound but that it could be improved in a number of ways to fill in the gaps and provide a higher level of assurance of safety and effectiveness.
THE FOOD AND DRUG ADMINISTRATION’S COMPLIANCE INFRASTRUCTURE
In the context of medical devices, compliance is simply conforming to the law, said Tim Ulatowski, director of the Office of Compliance at CDRH. FDA seeks a voluntary commitment to compliance by industry and others that must conform to the law but stands ready to enforce the law when it is necessary.
Some of the functions that fall under the heading of compliance are manufacturing-facility inspections, promotion and advertising, import control, bioresearch monitoring (for example, the evaluation of the conduct of clinical investigators, institutional review boards, and clinical sponsors), registration and listing (ensuring that all manufacturers are registered and that products are listed in the FDA system), recall monitoring and classification, premarket manufacturing review, enforcement actions, and education.
The hierarchy of decision-making related to compliance activities starts with the commissioner at the head of the agency, who is directed by the president and the secretary of health and human services. Within the commissioner’s organization is the Office of Chief Counsel (reporting directly to the secretary), the Office of Regulatory Affairs (ORA), CDRH, and other centers.
ORA manages the regional and district offices and the field laboratories across the country. It is a well-staffed, well-organized office that consists of thousands of people who conduct facility inspections, among other tasks. Recall coordinators interact with companies as they conduct recalls of products. The field laboratories sample-test products for enforcement purposes. ORA manages import operations at the points of importation by interacting with US Customs and Border Protection, evaluating entries to determine whether they are legally able to enter the marketplace in the United States. ORA also coordinates overall FDA compliance and enforcement policy and has a criminal-investigations group that evaluates possible felonies and misdemeanor criminal cases.
Within CDRH, the Office of Compliance establishes the medical device and radiologic health compliance and enforcement policy and procedures. The office determines resource allocations in collaboration with ORA—for example, how to allocate inspection staff, how much will be spent on manufacturing inspections, how much time will be spent on mammography
inspections, how much time will be allocated for inspections that might need to occur on the spur of the moment because of a particular problem, and time spent on bioresearch monitoring inspections. The Office of Compliance classifies recalls, reviews communications issued by the manufacturers, and drafts and issues communications from the agency regarding recalls. It also develops strategies for inspections and inspection assignments, coordinates foreign assignments, reviews inspections and enforcement cases, and enforces promotion and advertising law.
The Office of Chief Counsel (OCC) evaluates the cases brought by ORA and CDRH for legal sufficiency. Cases then move to the Department of Justice, to be brought before a court. OCC also evaluates regulations, guidance, new initiatives, and advisory letters.
FDA can take a variety of advisory actions. At the time of inspection, a manufacturer may be advised of potential violations. If the violations are serious enough, warning letters are issued to notify companies or clinical investigators that they are subject to legal action if the violations recur. If another inspection identifies the same violations, FDA moves into the penalty phase of enforcement action.
Product-specific seizure is one enforcement mechanism. A product may be seized, and a seized product may not be moved until the violation of the product is removed or, most often, the product is destroyed or reconditioned. FDA may seek an injunction against a facility or company, which more often than not will be resolved through consent decrees rather than litigation. The company and FDA enter into a court-approved consent order for corrections, monitoring, and oversight of the company as it resolves the violations. Another option is civil money penalties. Finally, there are misdemeanor or felony prosecutions in cooperation with the FDA Office of Criminal Investigation.
Unlike PMA devices, 510(k) products are not subject to preclearance inspections. There is an exception to that, under 513(f)(5), for cases of “substantial likelihood that the failure to comply with such regulations will potentially present serious risk to human health.”
Manufacturers that submit 510(k)s to FDA for clearance are subject to quality system inspections, bioresearch monitoring inspections, or medical device reporting–related inspections under FDA’s surveillance inspection program. There are also for-cause inspections in cases of allegations, reports, or other signals that come to the agency’s attention. Ulatowski noted that even if a company does not have to submit a 510(k) for a particular change in a device, it still must document changes, and those records are subject to review on inspection.
FDA compliance staff monitor information on the Web and review letters sent to FDA, most often by competitors who provide information about a product that is potentially being marketed without FDA clearance
or approval. The agency also looks for substantial new claims that are made without clearance. Investigational device inspections are conducted for clinical studies that may be submitted in a 510(k). FDA evaluates problems with imported products, including products subject to 510(k) clearance, and can detain products at the border, particularly if there is a violative inspection of a foreign facility. Most recalls are voluntary, whether of 510(k) or PMA devices, but on occasion FDA has exercised its mandatory recall authorities under the law. Additional 510(k)-related compliance activities in CDRH include sampling and testing of 510(k) products as needed.
Other components of CDRH monitor and evaluate the postmarket environment, not just from a manufacturing-environment standpoint but with reference to the clinical environment, trying to use information from all sources to gain an understanding of how products are performing in the marketplace. Ulatowski noted that FDA sometimes moves from surveillance to action, but the integration between these functions is not optimal.
FDA is trying to prevent problems, not only to react to them, and to identify issues related to risk as they emerge and deal with them strategically in a coordinated effort, not only with compliance activities but with educational efforts. Nevertheless, much time is spent in reacting to issues that come to light.