3
Regulatory Framework for Postmarket Surveillance of Medical Devices

“As I see it, what the Senator from New York is doing in this particular case is the same thing as if the Congress of the United States should attempt to say by law that calling a sheep’s tail a leg would make it a leg … if he desires to legislate against these mechanical devices he ought to do it in the open instead of by indirection and attempting to define as a drug something which palpably is not a drug.”

Senator Bennett Champ Clark (79 Cong. Rec. 4841, April 2, 1935)

The extension of U.S. Food and Drug Administration (FDA) oversight to devices has been uneven and sometimes has relied on stretching the definition of drugs.1 Until 1976 when Congress added the Medical Device Amendments (P.L. 94–295) to the Federal Food, Drug, and Cosmetic Act (P.L. 75–717), federal officials had limited explicit authority to regulate the safety or effectiveness of medical devices. In this legislation, Congress provided for additional regulatory scrutiny of medical devices while creating a regulatory framework that recognized certain differences between drugs and devices, particularly the substantial variability in the risk posed by different types of devices.

Virtually the entire regulatory framework for medical devices is general, that is, it applies to devices whether their primary or exclusive use is with adults or children. One exception is that when medical devices are tested with children in studies that will be submitted to FDA, they are usually subject to regulations for the protection of human research subjects that provide special, additional protections for child subjects. Also, in meeting its regulatory responsibilities, FDA may take special notice of children, for example, by limiting the labeled indications for the use of a device to

1  

As described in Chapter 1, the Supreme Court in 1969 sustained FDA’s categorization of a laboratory screening device (an antibiotic sensitivity disk) as a drug subject to premarket review (United States v. Bacto-Unidisk, 1969).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 73
Safe Medical Devices for Children 3 Regulatory Framework for Postmarket Surveillance of Medical Devices “As I see it, what the Senator from New York is doing in this particular case is the same thing as if the Congress of the United States should attempt to say by law that calling a sheep’s tail a leg would make it a leg … if he desires to legislate against these mechanical devices he ought to do it in the open instead of by indirection and attempting to define as a drug something which palpably is not a drug.” Senator Bennett Champ Clark (79 Cong. Rec. 4841, April 2, 1935) The extension of U.S. Food and Drug Administration (FDA) oversight to devices has been uneven and sometimes has relied on stretching the definition of drugs.1 Until 1976 when Congress added the Medical Device Amendments (P.L. 94–295) to the Federal Food, Drug, and Cosmetic Act (P.L. 75–717), federal officials had limited explicit authority to regulate the safety or effectiveness of medical devices. In this legislation, Congress provided for additional regulatory scrutiny of medical devices while creating a regulatory framework that recognized certain differences between drugs and devices, particularly the substantial variability in the risk posed by different types of devices. Virtually the entire regulatory framework for medical devices is general, that is, it applies to devices whether their primary or exclusive use is with adults or children. One exception is that when medical devices are tested with children in studies that will be submitted to FDA, they are usually subject to regulations for the protection of human research subjects that provide special, additional protections for child subjects. Also, in meeting its regulatory responsibilities, FDA may take special notice of children, for example, by limiting the labeled indications for the use of a device to 1   As described in Chapter 1, the Supreme Court in 1969 sustained FDA’s categorization of a laboratory screening device (an antibiotic sensitivity disk) as a drug subject to premarket review (United States v. Bacto-Unidisk, 1969).

OCR for page 73
Safe Medical Devices for Children adults or directing that pediatric questions be examined in studies following the approval of a device. This chapter provides a descriptive foundation for the later discussion of the adequacy of FDA’s program of postmarket surveillance to protect children. The chapter begins with a brief overview of organizational responsibilities for medical device regulation. It then reviews the premarket regulatory responsibilities of FDA as context for the following description of the agency’s responsibilities for postmarket surveillance of medical devices.2 The last section describes some agency programs and activities, for example, inspections of manufacturers that cover both premarket and postmarket arenas. ORGANIZATION OF FDA FOR MEDICAL DEVICE REGULATION Within the Food and Drug Administration, primary responsibility for regulating medical devices (and radiation-emitting electronic products) resides with the Center for Devices and Radiological Health (CDRH). The Center for Biologics Evaluation and Research (CBER) is responsible for regulating medical devices related to blood and cellular products (e.g., kits to test blood for HIV). For combination products that involve a drug and a device or a biological product and a device, the primary regulatory responsibility is assigned to CDRH if the main mode of action of the product is not biological or chemical and does not depend on being metabolized. Within CDR, the Office of Device Evaluation is responsible for the clearance or approval of medical devices that require premarket review. (The Office of In Vitro Diagnostic Device Evaluation and Safety handles reagents and in vitro diagnostic products.) Another unit, the Office of Science and Engineering Laboratories, contributes to the development of standards and methods for product assessments, performs laboratory evaluations and analyses, and conducts research and testing relevant to medical devices or radiation-emitting electronic products. This unit provides technical support for the development of the device-specific guidance documents as discussed later in this chapter. Although other offices have roles related to postmarket surveillance, the key unit is the Office of Surveillance and Biometrics (OSB). It has three divisions, the Division of Biostatistics, the Division of Postmarket Surveillance, and the Division of Surveillance Systems. Within OSB, the Division of Postmarket Surveillance oversees the adverse event reporting program, which includes the analysis and investiga- 2   As noted in Chapter 1, in referring to premarket and postmarket rather than premarketing and postmarketing activities, this report follows the legislative language that provided for this study and the usual (but not invariable) practice of FDA in describing activities that occur prior to or following the entry of a medical product into the market.

OCR for page 73
Safe Medical Devices for Children tion of reports. It also conducts epidemiologic research on the safety and use of medical devices and supports the development of epidemiologic methods for medical device research. As described in Chapter 5, the division has recently assumed responsibility for monitoring postmarket studies required at the time a device is approved for marketing. The Division of Biostatistics provides statistical support for both premarket and postmarket programs and also conducts and collaborates in original research on the health effects of device use. The Division of Surveillance Systems takes the lead in planning, developing, implementing, and maintaining OSB databases and information systems. Additional support in evaluating device problem reports may be provided by the Division of Device User Programs and Systems Analysis (which is part of CDHR’s Office of Communication, Education, and Radiation Programs and which includes FDA’s human factors program). The Office of Science and Engineering Laboratories provides technical assistance for premarket as well as postmarket programs. In addition to its central offices in Maryland, FDA has more than 160 field offices, laboratories, and other sites throughout the country. These sites house most of the agency’s Office of Regulatory Affairs employees who are responsible for various enforcement activities (e.g., seizures of adulterated foods or medical products) and for inspections of medical product, food, and cosmetic manufacturers (FDA, 2003g). Table 3.1 shows CDRH budget authority and total funding levels for FY 1994 to FY 2004. TABLE 3.1 Budget Authority and Total Program Level Funding History for Center for Devices and Radiological Health FY 1994–2005 (in millions) Year Budget Authority (Center and Field) Total Program Level (Budget Authority plus User Fees) 1994 $159 $159 1995 157 170 1996 144 152 1997 147 160 1998 144 156 1999 147 159 2000 158 170 2001 173 186 2002 180 194 2003 193 217 2004 191 222 NOTE: Program level funding includes user fees from inspections of mammography facilities and, beginning FY 2003, fees provided for under the Medical Device User Fee and Modernization Act of 2002. SOURCE: FDA Congressional Justification submissions (data compiled by CDRH staff).

OCR for page 73
Safe Medical Devices for Children The fiscal year 2004 budget for OSB postmarket surveillance activities was approximately $15 million, approximately half of which covered costs for about 70 full-time equivalent staff positions, including approximately 50 positions devoted to postmarket surveillance specifically (personal communication, Thomas P. Gross, M.D., Director, Division of Postmarket Surveillance, CDRH, October 1, 2004, and April 6, 2005). About one-third of the budget involved the MedSun program, which is described later in this chapter. The budget includes no funds for the analysis of outside data sources, such as Medicare databases or professional society registries. BASICS OF PREMARKET REGULATION OF MEDICAL DEVICES The premarket regulatory processes of FDA include evaluations, decisions, and other actions that occur prior to the marketing of a medical product. In some cases, requirements are established before or at the time of marketing approval for actions that will take place after marketing, for example, when further clinical studies are specified as a condition of FDA’s approval of a device. When Clearance or Approval of a Device Is Required According to FDA, approximately 20,000 American and foreign firms produce about 80,000 brands and models of medical devices for the U.S. market (FDA, 2002b). Before they can be marketed in the United States, roughly 55 to 60 percent of medical devices require FDA clearance or approval. The requirements that must be met for a device to be legally marketed in the United States depend in considerable measure on its risk classification. The Medical Device Amendments of 1976 provided that devices be classified—in ascending order of risk—as Class I, II, or III devices (21 USC 360c). FDA completed the basic process of classifying existing devices into the three groups by 1988 (Merrill, 1994). In 2004, the three classes accounted for about 43 percent, 44 percent, and 13 percent of classified devices, respectively (personal communication, Donna-Bea Tillman, Ph.D., Deputy Director, Office of Device Evaluation, CDRH, January 18, 2005; see also FDA, 2004n).3 Table 3.2 provides examples of common pediatric-use devices in the three classes. 3   The 13 percent figure includes Class III devices that require premarket approval.

OCR for page 73
Safe Medical Devices for Children TABLE 3.2 Examples of FDA Class I, II, and III Devices Device Class Examples I Bassinets Nursing bottle nipples Infant caps Mechanical toothbrushes Circumcision trays Mechanical wheelchairs II Neonatal incubators Dialysis catheters Circumcision clamps Powered wheelchairs Apnea monitors Transcutaneous electrical nerve stimulators Ventriculoperitoneal shunts III Implantable insulin pumps Implantable cardiac pacemakers Ventricular assist devices Cochlear implants Deep brain stimulators Class I Devices Class I devices are considered to present relatively low risk to patients. As specified in statute (21 USC 360c(a)(1)(A)), this class covers devices for which certain “general” controls (e.g., standards for good manufacturing practices) provide reasonable assurance of the safety and effectiveness of the device or devices that are not intended or represented (“purported”) to support or sustain life or play an important role in preventing impairment or that are not expected to pose an unreasonable risk of illness or injury. General controls apply to Class II and Class III as well as Class I devices. These controls are discussed further below. Nearly all Class I devices and some Class II devices may be marketed without FDA clearance or approval. Examples of Class I devices that are not exempt from review under the notification procedures outlined below are dental mercury, mechanical wheelchairs, and surgeon’s gloves (FDA, 2005e,f,g). If a device that is normally exempt from FDA premarket clearance is to be marketed for a new intended use or involves a new fundamental technology, it would require a premarket clearance.

OCR for page 73
Safe Medical Devices for Children Class II Devices Devices categorized as Class II present more risk than Class 1 devices. For a Class II device to be legally marketed, the manufacturer must usually submit a notification of intent to market and receive FDA clearance under “510(k)” provisions (referring to the applicable section of the Federal Food, Drug, and Cosmetic Act, 21 USC 360(k); see also FDA, 2004w). These 510(k) provisions cover devices that are “substantially equivalent” to a “predicate” or “pre-amendment” device, which is one that was either marketed before May 28, 1976 or one that has been shown (through the notification and clearance process) to be substantially equivalent to such a device.4 FDA considers a device substantially equivalent if it has the same intended use and the same technological characteristics as the predicate (pre-amendment) device. A device may also be considered substantially equivalent when it has the same intended use but different technological characteristics if these differences do not raise different questions of safety and effectiveness and if information (which can include clinical data) is provided to show that the device is as safe and effective as a legally marketed device. This latter definition allows FDA “the flexibility to clear some fairly novel devices through the 510(k) process” (Kahan, 1996, p. 89). In assigning devices to Class II, FDA has determined that (1) general controls are not by themselves sufficient to provide reasonable assurance of safety and effectiveness, but (2) sufficient information is available to develop special controls for that purpose. These controls are discussed below. For certain Class II devices, FDA issues guidance documents that describe what kinds of bench, animal, and clinical data should be submitted to show that a device is substantially equivalent to a predicate (preamendment) device. In 2002, for example, the agency issued guidance for manufacturers of carbon dioxide and oxygen monitoring devices about the kinds of information they should submit to document safety and effectiveness as part of a 510(k) submission (FDA, 2002d). Submission of clinical data is required for about 10 to 15 percent of devices covered by the 510(k) process (Tillman and Gardner, 2004). FDA also has the authority to require further studies for devices that are covered by regulations authorizing postmarket surveillance studies as discussed below. 4   The provisions also apply to pre-amendment devices that have been classified as Class III devices but for which FDA has not yet issued regulations calling for premarket approval applications. A few Class II devices are exempt from the provisions, including pediatric hospital beds, enuresis alarms, and hematocrit measuring devices (FDA, 1998j).

OCR for page 73
Safe Medical Devices for Children Class III Devices Class III devices are intended to support or sustain life or play an important role in preventing impairment or are considered to pose an unreasonable risk of illness or injury. For devices in this class, FDA has determined that general controls are inadequate to reasonably assure safety and effectiveness and that available information is insufficient to develop adequate special controls.5 As indicated by percentages cited earlier, many more devices enter the market through the clearance process than through the approval process (more than three times as many in 2004). Usually, manufacturers of Class III devices must submit premarket approval (PMA) applications to FDA. As part of such applications, they must present the results of investigations—including data from clinical studies—that support the device’s safety and effectiveness for the use or uses proposed. After a device is approved, changes in the device, its labeling or packaging, or its manufacturing may require approval under a supplemental PMA application if the changes affect safety or effectiveness. Such supplemental applications also cover changes or other actions related to any postmarket studies that were required as a condition of approval of a device. The 1997 legislation provided that manufacturers have an opportunity to meet with FDA to discuss their clinical investigation plan prior to submitting a PMA application. They may request a “determination meeting” to discuss what kind of scientific evidence (e.g., a randomized clinical trial) FDA considers necessary to demonstrate that a device is effective for its intended use. The resulting determination is binding on the agency, unless it is subsequently judged to be “contrary to public health” (FDA, 2001d, p. 1). In addition, the legislation provided the opportunity for an “agreement” meeting to those planning a PMA application or a 510(k) submission for certain devices. The purpose of such a meeting is to reach agreement on the main elements of the investigational plan, including the clinical protocol. The results of an agreement are again binding in most circumstances. An alternative to the PMA application is the product development protocol (PDP), which was provided for by the 1976 Medical Device Amendments but not implemented until the 1990s (FDA, 1999e). The PDP pathway allows a manufacturer, with FDA agreement, to consult with FDA 5   A new device may also be automatically classified as a Class III device because no predicate device exists, that is, the new device is not substantially equivalent to any other Class I or Class II device marketed before May 28, 1976, or to any device that was placed into Class I or Class II after that date. If a device is automatically classified into Class III because no predicate device exists to which it can be claimed equivalent and if the device presents a low risk to patients, the FDA Modernization Act of 1997 allows FDA to reclassify the device into Class I or II under a “de novo” or “risk-based” procedure (FDA, 1998k).

OCR for page 73
Safe Medical Devices for Children to develop and implement a mutually acceptable device development testing protocol for a device. The manufacturer can then secure approval by submitting and having FDA accept a notice that it has fulfilled the requirements of the protocol.6 FDA considers this pathway appropriate for “those devices in which the technology is well established in industry” (FDA, 2003c, unpaged). To comply with the FDA Modernization Act of 1997 (P.L. 105–115), FDA revised procedures for expedited review of PMA applications. Expedited review is allowed for devices that are intended to treat or diagnose life-threatening or irreversibly debilitating diseases or conditions and that also represent (1) breakthrough technologies, (2) technologies for which no approved alternatives exist, (3) technologies that offer significant advantages over existing approved alternatives, or (4) technologies the availability of which is in the best interest of patients (FDA, 2003i). Expedited review may involve advance consultation with FDA. (Provisions for expedited review also exist for products requiring premarket clearance.) When a device is approved for marketing, FDA may impose requirements for further study of or reporting about a device to expand knowledge about its safety or effectiveness (21 CFR 814.82). These required studies are often referred to as condition-of-approval or post-approval studies. These studies and their monitoring by FDA are discussed further in Chapter 5. Investigational Devices For certain devices that have not been approved or cleared for marketing or that are being tested for indications not previously approved or cleared, use of the device during testing occurs under an “investigational device exemption” or IDE (21 USC 360j(g); 21 CFR 812.2(c)). An IDE is required for a “significant risk” device, which regulations define as one that presents a potential for serious risk to the health, safety, or welfare of a research participant (21 CFR 812.3(m)).7 The IDE regulations specifically 6   The protocol is to include a description of (1) the device, including modifications; (2) any preclinical or clinical studies completed, underway, or planned; (3) manufacturing methods, facilities, and controls; (4) applicable performance standards, if any; (5) proposed labeling; and (6) other information deemed necessary by FDA. The manufacturer must also submit progress reports and information on studies described in the protocol. 7   FDA may allow clinical use of unapproved devices in other situations, including certain emergency situations and certain situations in which a clinical study has been completed but the marketing application has not yet been approved. In addition, under so-called “compassionate use” provisions, FDA may allow use of an investigational device when it might benefit a patient who does not meet criteria for inclusion in research but who has a serious medical condition and no satisfactory alternative (FDA, 2003e).

OCR for page 73
Safe Medical Devices for Children mention the potential for serious risk related to implants, life-supporting or life-sustaining devices, and devices that are substantially important in preventing impairments in health. An IDE application must include information on preclinical studies and any already available clinical data. The sponsor must also submit an investigational plan that describes the research design and analytic methods to be used. The study cannot proceed until the IDE is approved by FDA and an Institutional Review Board (IRB).8 For studies involving significant risk devices, FDA and investigators or sponsors may engage in extensive communication and negotiation about the characteristics and objectives of research studies to support claims of product safety and effectiveness. An IDE application is not required for a “non-significant risk” device study. The sponsor must, however, comply with certain recordkeeping and reporting requirements. In addition, even if an IDE is not required, research involving human participants is still subject to certain other requirements, including IRB review (see, e.g., FDA, 2003d). A study involving a “nonsignificant” risk device is said to have an “abbreviated IDE” or “deemed approved IDE.” Humanitarian Use Devices In addition to the clearance and approval processes described above, Congress has allowed devices to be approved for marketing under a Humanitarian Device Exemption (HDE). To qualify, a device must be intended for patients with a rare disease or condition for which no comparable device is available that has a 510(k) clearance or an approved PMA application for the proposed indication (21 USC 360j(m)). “Rare” is defined to mean that the condition affects or is manifested in (causes symptoms in) fewer than 4,000 individuals in the United States per year. Among other requirements, manufacturers seeking an HDE must present evidence that (1) provides a reasonable assurance of product safety when the device is used as proposed and (2) indicates that the probable health benefits of the device outweigh the potential for harm, taking into account the risks and probable benefits of available alternative therapies. Evidence of effectiveness is not required. Granting of an HDE allows a company to market a device as a Humanitarian Use Device (HUD). Except in certain emergency situations, such a 8   An IRB is a group of qualified individuals charged under federal regulation with protecting the rights and welfare of people involved in research in accord with federal regulations. IRBs review and approve plans for research involving humans.

OCR for page 73
Safe Medical Devices for Children device can only be used in a health care facility following IRB approval and continuing review. (The IRB may approve use of the device on a case-by-case basis or under a research protocol or without any further restrictions.) As is the case for a PMA application, FDA may order a manufacturer to conduct further studies as a condition of approval for an HDE (21 CFR 814.126(a), 814.82(a)(2)). For example, when FDA approved an HDE for the use of a left ventricular assist device with children, it required that the first 50 children receiving the device be followed to heart transplantation, death, or other outcome. Least Burdensome Approach The FDA Modernization Act of 1997 specified that the procedures and information required of manufacturers to demonstrate substantial equivalence for 510(k) clearance involve the “least burdensome means” for such demonstration (21 USC 360c(i)(1)(D)). Likewise, for devices requiring premarket approval, regulators are to “consider, in consultation with the applicant, the least burdensome appropriate means of evaluating device effectiveness that would have a reasonable likelihood of resulting in approval” (21 USC 360c(a)(3)(D)(ii)). Although the statutory provisions involve premarket clearance and approval processes, FDA has said it will apply the least burdensome concept to postmarket and other activities as well (FDA, 2002q, unpaged). In attempting to put the least burdensome concept into practice, FDA has stated that it would apply these basic principles. First, “[t]he basis for all regulatory decisions will be found in sound science and the spirit and the letter of the law.” Second, “[i]nformation unrelated to the regulatory decision should not be part of the decision-making process.” Third, “[a]lternative approaches to regulatory issues should be considered to optimize the time, effort, and resources involved in resolving the issue consistent with the law and regulations.” Fourth, “[a]ll reasonable measures should be used to reduce review times and render regulatory decisions within statutory timeframes.” (All text quoted from FDA, 2002q.) User Fees In the Medical Device User Fee and Modernization Act of 2002 (P.L. 107–250), Congress authorized FDA to charge a fee for the review of 510(k) submissions and PMA applications. One major objective was to help speed the clearance or approval of devices by augmenting FDA resources. In contrast to the provisions for prescription drug user fees, device user fees are expressly allowed to be used for evaluating condition-of-approval postmarket studies and identifying safety and effectiveness issues

OCR for page 73
Safe Medical Devices for Children for devices (see 21 USC 379i(5)(j) and (k) and 21 USC 379j(h)(2)(A)(ii) for devices compared to 21 USC 379g(6) for drugs). The legislation provided a complete waiver of fees for certain reviews involving pediatric use of a device. Specifically, if a company seeks clearance or approval of a device solely for pediatric use, the fee may be waived. If the company later seeks to add an adult indication, the user fee would be assessed at the regular level for a PMA review. If, however, a company has a device that has been cleared or approved for an adult indication and then seeks clearance or approval for a use that involves only a pediatric population, the fee may be waived. According to FDA, 32 applications for FDA approval or clearance were exempted from user fees in FY 2004 (personal communication, Heather Rosecrans, Director, Premarket Notification, Office of Device Evaluation, CDRH, January 18, 2005). Two applications involved premarket approvals; the other 30 were applications for clearance under 510(k) procedures (7 of which involved a determination that the product in question was not a device). User fees do not apply to requests for Humanitarian Device Exemptions. Off-Label or Unlabeled Use of Devices A typical FDA letter granting approval of a PMA application states the indications for use of the device. Sometimes the approval letter may note limitations, for example, that the use is for those over a certain age. Likewise, each 510(k) clearance letter includes an accompanying “indications for use” page that states the cleared indications and any limitations on use (FDA, 2002f). Once a device is approved or cleared, physicians may use the device for indications that are not mentioned in the device’s labeling but are not specifically restricted. Such use is sometimes called “off-label” or “unlabeled” use (see, e.g., FDA, 1998c, 2002f). It is considered part of the practice of medicine, which FDA—by statute—does not regulate (21 USC 396; FDA, 1998h). When General or Special Controls Apply General controls apply to all three classes of medical devices (FDA, 1998d). They include requirements for actions both prior to and after a device reaches the market. General controls require device manufacturers to register each manufacturing location with FDA; list their marketed devices with FDA; comply with device labeling regulations;

OCR for page 73
Safe Medical Devices for Children analyses of the design of devices associated with multiple adverse event reports. Examples include glucose meters, infant apnea monitors, ventilators, and infusion devices (Wiklund, 2003). Although not expressed in terms of design changes per se, in 2002 FDA issued a guidance document for apnea monitors that identified risks to health associated with the monitors and recommended device features or testing procedures to mitigate each risk (FDA, 2002c). In addition to considering whether design problems have contributed to reported adverse events, FDA also seeks to prevent safety problems related to design deficiencies. Quality system regulations, which apply to the development and manufacture of devices (and, thus, apply both before and after a device is marketed), emphasize the detection and correction of problems during the manufacturing process. The goal is to prevent the problems that would trigger a recall or Public Health Notification.17 Responses to Familiar Problems As noted above, the focus of adverse event reporting and other postmarket surveillance activities is on the detection of unexpected, serious problems or recognized problems occurring more frequently than expected. Sometimes, however, the agency targets familiar, well-recognized problems for special initiatives and collaborations with other public and private groups. For example, in 2001, FDA issued a notice on reducing the radiation risk from computed tomography for pediatric and small adult patients (Feigal, 2001b). The agency acknowledged that the recommendations were not new, but it decided it was important to emphasize that radiation doses for small patients should be kept as low as possible, consistent with achieving clinical objectives. On another front, FDA has created a home health care committee to review what FDA has done and might do to respond to problems with the use of complex medical devices in the home (see, FDA, 2004c). The agency has recognized that the use of sophisticated medical devices in the home, while an accepted and necessary part of modern health care, “adds an additional level of risk of unintended adverse events” (Arcarese, 2002a, unpaged). Based on discussions with a range of interested parties, the home health committee has decided to focus in particular on safe use of infusion pumps in the home. It 17   The relevance of design controls to the prevention of safety problems is suggested by this FDA description of what might be required of a manufacturer planning a new defibrillator for use by hospital and emergency medical personnel. “Designers would have to consider all aspects of use in both settings … [including] storage temperatures in the ambulance, road shock and vibration, two-way radio interference, electrical noise generated by the siren and many other factors” (FDA, 1996a, unpaged).

OCR for page 73
Safe Medical Devices for Children has asked pump manufacturers to provide basic information and use instruction for every infusion pump marketed since 1984. The information will be included on the committee’s website to allow ready access to such information by patients, family members, and home care nurses.18 CONFIDENTIALITY OF INFORMATION OBTAINED BY FDA Much information that FDA receives is confidential or is treated as confidential. The scope of agency confidentiality requirements or practices has been a prominent focus of criticism during recent controversies about the availability of information from postmarket studies of drugs. As discussed in Chapters 5 and 6, the committee’s efforts to learn more about the status or findings of postmarket studies ordered by FDA were limited by agency confidentiality policies (as well as by the agency’s lack of an adequate study monitoring system). Confidentiality protections for information submitted to FDA are provided by three key federal statutes: the Federal Food, Drug, and Cosmetic Act (21 USC 301 et seq.), the Freedom of Information Act (5 USC 552), and the Trade Secrets Act (18 USC 1905). Implementing regulations clarify how these protections apply to study protocols and to preclinical and clinical study data, as well as to information relating to product design, product composition, and manufacturing methods and processes. Confidentiality provisions of the Privacy Act (5 USC 552a) also restrict the disclosure of information about individuals, including patients and health care personnel. Confidential information can, of course, be leaked or released mistakenly, but the provisions described here normally operate as intended to restrict disclosure of covered categories of information. FDA personnel are prohibited from disclosing (or from using to their own advantage) information acquired under their statutory authority that concerns “any method or process which as a trade secret is entitled to protection”19 (21 USC 331(j)). This provision covers information acquired by FDA under investigational device exemption applications, 510(k) premarket notifications, and premarket approval applications, or otherwise acquired under authority of various statutory provisions. The provision does not expressly include Section 522, which authorizes FDA to require 18   The home care committee has already issued a pamphlet for consumers on blood glucose monitors and prepared a checklist for consumers to promote safe use of devices in the home (FDA, 2003m). 19   This provision does not authorize the withholding of information from the U.S. House of Representatives, Senate, or a committee or subcommittee thereof with jurisdiction over the specific subject matter.

OCR for page 73
Safe Medical Devices for Children postmarket surveillance and requires a manufacturer to submit a plan for the required surveillance (21 USC 360(l)). FDA, however, treats such plans as protected from public disclosure under provisions of the Freedom of Information Act and FDA’s implementing regulations. Interim and final reports of postmarket surveillance studies or data are treated similarly, although limited findings may be made public in connection with a safety advisory, the approval of a new indication for a product’s use, a labeling change, or similar action. Data and information relating to postmarket studies conducted under an IDE would be protected consistent with the IDE confidentiality regulations. More broadly, the federal Trade Secrets Act prohibits federal employees from disclosing information which “concerns or relates to the trade secrets, processes, operations, style of work, or apparatus or to the identity, confidential statistical data, amount or source of any income” or other financial information of any person or company (18 USC 1905). The Trade Secrets Act is a criminal statute. Likewise, a violation of disclosure provisions of the FDA statute (21 USC 331(j)) is subject to criminal penalties (21 USC 333(a)). Congress can, however, request or subpoena information including trade secret and confidential commercial information, and that is not subject to 331(j) nondisclosure. In addition, FDA can require that a summary of safety and effectiveness information be submitted to document the basis for premarket approval, and the agency can release the summary to the public once it has issued an approval order. Such a summary “shall include information respecting any adverse effect on health of the device” (21 USC 360j(h)(2)). Much of the committee’s understanding of the kinds of studies submitted to support PMA approvals was obtained by reading individual approval summaries, which are posted on the CDRH website. Otherwise, by statute, information contained in a PMA application is to be held confidential and is not to be used by FDA to approve another manufacturer’s PMA application, to establish a performance standard or special control, to reclassify a device, or to approve a product development protocol (21 USC 360j(c)). An exception is that information from preclinical or clinical tests or studies that demonstrate safety or effectiveness can be used by FDA for these purposes 6 years after the approval of the PMA application, but this exception does not cover information regarding “methods of manufacture and product composition and other trade secrets” (21 USC 360(j)(h)(4)). An FDA regulation provides confidentiality protections for PMA filings (21 CFR 814.9). It provides that the existence of a PMA filing will not be disclosed by FDA if its existence has not been disclosed by the applicant, up until the time the PMA application is approved or denied approval. An exception permits FDA to disclose a summary of portions of the safety and

OCR for page 73
Safe Medical Devices for Children effectiveness data prior to approval “if disclosure is relevant to public consideration of a specific pending issue” (21 CFR 814.9(d)(1)). After approval, FDA can disclose certain information in the PMA application, except for trade secret or confidential commercial information. FDA can disclose: (1) safety and effectiveness data “previously disclosed to the public”; (2) a protocol for a test or study “unless the protocol is shown to constitute trade secret or confidential commercial or financial information” protected under the Freedom of Information Act and implementing regulations; (3) adverse reaction reports, consumer complaints, and similar data and information—but only after deleting trade secret or confidential commercial or financial information and deleting personnel, medical, and similar information the disclosure of which would constitute an unwarranted invasion of personal privacy; (4) assay methods and other analytical methods, unless they do not serve a regulatory purpose and they are trade secret or confidential commercial information; and (5) a list of components previously disclosed to the public. FDA cannot disclose: (1) safety and effectiveness information not previously disclosed to the public that constitute trade secret or confidential commercial information; (2) manufacturing methods or processes, including quality control procedures; (3) quantitative or semi-quantitative formulas; and (4) production, sales, distribution, and other similar data and information. Data and information contained in an IDE is handled in accordance with the PMA regulation in Section 814.9 (21 CFR 812.38(d)). Similar confidentiality provisions apply to data and information contained in a 510(k) notification and disclosure of the existence of a 510(k) notification prior to its clearance (21 CFR 807.95). An exception exists under all these regulations that permits FDA to disclose a summary of portions of the safety and effectiveness data prior to approval or clearance “if disclosure is relevant to public consideration of a specific pending issue.” The Freedom of Information Act directs federal agencies to make information in agency files available to the public, but certain information is exempted from public disclosure. Among the types of exempted information are (1) “trade secrets and commercial or financial information obtained from a person and privileged or confidential,” (2) “personnel and medical files and similar files the disclosure of which would constitute a clearly unwarranted invasion of personal privacy,” and (3) information specifically exempted from disclosure by statute (5 USC 552(b)(3), (4), and (6)). FDA has issued regulations implementing these exemptions. Under the regulation applicable to trade secrets and confidential commercial information (21 CFR 20.61), FDA has defined these terms as follows: (a) A trade secret may consist of any commercially valuable plan, formula, process, or device that is used for the making, preparing, com-

OCR for page 73
Safe Medical Devices for Children pounding, or processing of trade commodities and that can be said to be the end product of either innovation or substantial effort. There must be a direct relationship between the trade secret and the productive process. (b) Commercial or financial information that is privileged or confidential means valuable data or information which is used in one’s business and is of a type customarily held in strict confidence or regarded as privileged and not disclosed to any member of the public by the person to whom it belongs. FDA’s regulation applicable to medical information provides that names and “information which would identify patients or research subjects in any medical or similar report, test, study or other research project” shall be deleted prior to public disclosure of the record (21 CFR 20.63(a)). “The names and any information that would identify the voluntary reporter or any other person associated with an adverse event” shall not be publicly disclosed by FDA or by a manufacturer who reports such an event (21 CFR 20.63(f)). The names of entities required by statute to make adverse event reports are not protected from disclosure. FDA’s definitions of trade secret and confidential commercial information were drafted to be consistent with judicial determinations that “any technical or scientific information developed by a company may be considered a trade secret where it is not generally known or readily ascertainable and when it is protected and maintained as confidential by the developer and is of value to him” (FDA, 1974, p. 44613). Testing data, including protocols used for testing the product and test results, can be protected from disclosure under the Freedom of Information Act (e.g., Heeney v. FDA, 2001). Sometimes other statutes may affect the confidentiality protections usually afforded to confidential commercial or financial information under the Federal Food, Drug, and Cosmetic Act and the Freedom of Information Act. For example, public companies subject to regulation by the Securities and Exchange Commission (SEC) are required to disclose certain information that might be “material” to decisions by investors and potential investors. To comply with these SEC requirements, a public company might publicly disclose in SEC filings or press releases certain information about the existence of a clinical trial, the results of a clinical or preclinical study, or the fact that a PMA application or 510(k) submission has been submitted to FDA. Companies that are privately funded (such as by venture capital or private investors) might not disclose such information if they are not subject to disclosure requirements. As discussed above, FDA is bound by confidentiality requirements to the extent the information has not been previously disclosed to the public. Companies provide the highest confidentiality protections to device design information, manufacturing processes and methods, and quality con-

OCR for page 73
Safe Medical Devices for Children trol information. Protocols for clinical studies and preclinical testing are also typically protected from public disclosure. Aggregate results of a safety and effectiveness study may be disclosed in summary form, but raw data or site-specific data are often protected as confidential, unless published in a scientific journal. As applied in the context of postmarket surveillance, FDA would be required by the three statutes discussed above to protect from public disclosure information that constitutes trade secrets or confidential commercial information. For example, a postmarket study protocol and the results of condition-of-approval studies would be submitted as supplemental PMA applications and thus would be governed by the confidentiality requirements of the Federal Food, Drug, and Cosmetic Act and Section 814.9 of FDA’s regulations. Under the postmarket surveillance provision in Section 522 of the Act, a manufacturer is required to submit a plan for FDA-required postmarket surveillance, but FDA treats such plans as protected from public disclosure under provisions of the Freedom of Information Act and FDA’s implementing regulations. Interim and final reports of postmarket surveillance studies or data are treated similarly, although limited findings may be made public in connection with a safety advisory, the approval of a new indication for a product’s use, a labeling change, or similar action. Data and information relating to postmarket studies conducted under an IDE would be protected consistent with the IDE confidentiality regulations. FDA PROGRAMS THAT CROSS THE PREMARKET/POSTMARKET BOUNDARY In addition to premarket and postmarket programs, FDA has programs that cross the market approval boundary to promote device safety both before and after a device is marketed. This is consistent with the agency’s analysis of its range of activities as they relate to a product’s total life cycle from initial concept to obsolescence (see Figure 1.2). Research, Analysis, and Methods Development As described above, CDRH has active research programs to evaluate elements of device technologies or their effects, to support the development of standards or guidance, and otherwise to build the knowledge base for device design, testing, manufacture, regulation, and clinical use. CDRH’s 2003 Annual Report described its epidemiological research program, which provides consultative services on topics or problems requiring epidemiological expertise (e.g., literature reviews, risk assessments, design of obser-

OCR for page 73
Safe Medical Devices for Children vational studies). The report listed studies that produced journal publications and conference presentations on a broad range of topics, including allergic reactions to platinum in breast implants, breast implant rupture, tampon-associated toxic shock syndrome, gender differences in pulmonary artery rupture, and uses and outcomes associated with transmyocardial revascularization (a procedure sometimes used to relieve chest pain). Other studies focused on methodology or process issues, for example, the use of the National Electronic Injury Surveillance System (NEISS) to assess the frequency of injuries due to medical devices. CDRH epidemiologists have also participated, on an exploratory basis, in premarket approval assessments to help determine whether and what kind of postmarket evaluations would be appropriate. Concern about deficiencies in device design has prompted FDA to direct more attention to general principles and strategies for safe device design and use, including human factors engineering. Human factors engineering analyzes how people employ technologies and how user characteristics (e.g., cognitive capacities, expectations) interact with characteristics of their environments (e.g., workload, lighting) to affect the safe and effective use of technologies (FDA, 2003s). Such analysis can be applied to adverse events involving medical devices and potential means to prevent them. Quality Systems Regulations A major boundary crossing program involves quality system regulations, which encompass good manufacturing practices. These regulations are among the general controls described earlier in this chapter. Congressional concern about manufacturing practices dates back at least to the 1938 Federal Food, Drug, and Cosmetic Act, when Congress specified that manufacturing methods, facilities, and controls be “adequate” for regulated products. The first FDA guidance about adequate manufacturing processes dates to the early 1940s; it followed a drug manufacturing mishap that left dozens of people dead or injured (Swann, 1999). FDA issued the first requirements for good manufacturing practices for medical devices in 1978 (FDA, 1978). These requirements—which remained essentially the same until 1996—covered methods, facilities and controls related to the manufacture, packing, storage, and installation of medical devices. The Safe Medical Devices Act of 1990 expanded FDA authority to include control related to device design prior to actual production. That legislation also encouraged FDA to work with other countries toward commonly recognized good manufacturing practices. In 1996, FDA published the Quality System Regulations, which it described as “revising the current good manufacturing practice (CGMP) requirements for medical devices and incorporating them into a quality sys-

OCR for page 73
Safe Medical Devices for Children tem regulation” (FDA, 1996b, p. 52602). By intent, the regulations are very similar to the International Organization for Standardization (ISO) provisions (ISO, 2000), which have recently been updated. (ISO is described briefly below.) The quality system regulation is broad in scope as indicated in Box 3.4. The quality system regulations cover a very large array of devices. For that reason, the regulation “provides the framework that all manufacturers must follow by requiring that manufacturers develop and follow procedures and fill in the details that are appropriate to a given device according to the current state-of-the-art manufacturing for that specific device” (FDA, 1996b, p. 52603). As noted earlier, special control guidance documents for particular categories of devices (e.g., apnea monitors) may be much more specific. BOX 3.4 Topics in Medical Device Quality Systems Manual: A Small Entity Compliance Guide The Quality System Regulation Quality Systems Design Controls Process Validation Personnel Buildings and Environment Equipment and Calibration Device Master Record Document and Change Control Purchasing and Acceptance Activities Labeling Product Evaluation Packaging Storage, Distribution, and Installation Complaints Servicing Quality Systems Audits Factory Inspections Appendix Appendix 1: The Quality Systems Regulation Appendix 2: Application of the Medical Device GMPs [Good Manufacturing Practice] to Computerized Devices and Manufacturing Processes SOURCE: FDA, 1999c.

OCR for page 73
Safe Medical Devices for Children Design Controls As described in the quality system regulations (21 CFR 820.30), the design controls requirements for Class II and III devices, Class II and III investigational devices, and certain Class I devices cover considerable ground. Design controls involve creating plans that cover design and development activities and assign responsibility for implementing them; specifying design input requirements, that is, the physical and performance requirements for a device design that are appropriate given the device’s intended uses and users; developing the design output, meaning the results of the design effort at each stage, including the finished design effort (the device, its packaging and labeling, and the device master record); verifying that the design output is consistent with the design input requirements; conducting periodic design reviews to assess the adequacy of the design requirements, evaluate whether the design will meet the requirements, and identify problems; validating through tests of production units under actual or simulated conditions and other means that the device (including software) meets objectives for intended uses and users; correctly translating the device design into production specifications; controlling changes in design during the design process and after the device is marketed; and documenting the design process in the design history file. FDA requires that applications for premarket approval include descriptions of design controls and other quality systems information, and it evaluates compliance during a pre-approval facility inspection (FDA, 2003q,r). The agency also evaluates compliance during routine quality systems inspections for all devices covered by the design control requirements. Corrective and Preventive Actions In guidance on the quality system inspections of manufacturers, one focus is what FDA terms the Corrective and Preventive Actions or CAPA subsystem (FDA, 1999b). A major component of this subsystem consists of procedures to detect, understand, and correct problems during the manufacturing process. The objective of this aspect of quality system regulations is to prevent defective devices from reaching the market. Other components of the CAPA subsystem reviewed by FDA are the manufacturer’s confor-

OCR for page 73
Safe Medical Devices for Children mance with adverse event reporting regulations, recall and corrective actions, and procedures for any required tracking of a medical device. Inspections As noted above, FDA inspections of manufacturing facilities may occur before or after a medical device is approved for marketing. A pre-approval inspection is usually required as part of the PMA process. FDA staff also conduct “directed” or “for cause” inspections when they are investigating a specific problem or following up to assure that corrective actions from a previous inspection have been implemented. A program of bioresearch monitoring includes on-site inspections and data audits of sites involved in FDA-regulated research. As described by FDA in its 2002 Performance Plan, the FDA Modernization Act of 1997 has allowed firms to declare conformity to standards or quality systems requirements as part of steps to streamline the premarket clearance process (FDA, 2001h). This has increased the burden on FDA’s inspection process, which as discussed in Chapter 7, falls short of meeting statutory requirements that FDA inspect facilities that manufacture Class II and III devices every 2 years (21 USC 360(h)). FDA’s quality system inspections focus on particular subsystems of manufacturing quality controls, specifically management, design (see above), corrective and preventive actions (see above), and production and processes. The other major subsystems involve materials controls, facility and equipment controls, and records, documents, and change controls. To ease the inspection burden on FDA, the 2002 Medical Device User Fee and Modernization Act gave manufacturers with a good history of regulatory compliance the option, under certain circumstances, of choosing an FDA-accredited, nongovernmental entity to perform quality systems inspections (21 USC 374(g)). (This is described by FDA as its “Accredited Persons” or AP program.) FDA staff would focus on firms with a record of compliance problems and manufacturers of high-risk products, including implants and life-supporting or sustaining devices. International Efforts to Harmonize Policies Medical device development, research, and sales are international in scope. The ISO standards cited above are one example of a number of cross-national initiatives—some longstanding, some relatively recent—to promote consensus and consistency in regulatory and voluntary standards for medical and other products and industries, measurement and testing methods, management systems, and other areas. ISO is a nongovernmental network whose membership consists of the national standards institutes of

OCR for page 73
Safe Medical Devices for Children nearly 150 countries (ISO, 2004). Using committees that include producers, consumers, regulators, and other relevant parties, the organization develops consensus standards on topics such as terminology, testing methods, product characteristics, and manufacturing processes. Some standards are generic, such as ISO 9000, which concerns quality management. Other standards are industry or product specific, such as ISO 13485, which concerns medical devices. The agency currently has more than 14,000 standards and related documents. Individual countries may choose to adopt the standards by regulation. In the 1990s, the Global Harmonization Task Force (GHTF) was created specifically as a voluntary process to pursue harmonization of national policies on the regulation of medical devices. It includes participants from national regulatory agencies and industry. Of four GHTF study groups, one has focused on postmarket surveillance, including adverse event reporting programs. FDA supports this activity, but the study group findings and recommendations are advisory, not binding. One issue for the GHTF task force on adverse event reporting is promoting the exchange of event reports among “national competent authorities” (e.g., FDA in the United States) (GHTF, 2002, p. 4). In 2003, authorities in 16 countries exchanged more than 140 “international vigilance reports,” most related to recalled devices (FDA, 2004v). Other harmonization activities include those of the International Conference on Harmonization (ICH). ICH has, for example, provided guidance for clinical investigators, primarily those involved in drug studies (ICH, 1996). An ISO document has focused on clinical investigators studying medical devices (ISO, 2003a,b; see Giroud, 2004). FDA has not adopted that ISO standard but has said that it might do so after the next revisions (Dickinson, 2004b). This chapter has focused on description rather than assessment. Thus, it includes no conclusions or recommendations related to the adequacy of existing laws and regulations or their implementation as they relate to children. The next chapter examines FDA’s programs of adverse event reporting and offers recommendations for improvement.