Children and Medical Device Safety: A Shared Responsibility
“It is critically important that the safety and performance of medical devices are continually assessed when they are in use…. No amount of rigour in the pre-marketing review process can predict all possible device failures or incidents arising from device misuse. It is through actual use that unforeseen problems related to safety and performance can occur.”
World Health Organization (WHO), 2003, p. 13
As this report was being drafted, studies that identified serious adverse effects of several commonly prescribed medications were attracting widespread news coverage and focusing congressional and public attention on the U.S. Food and Drug Administration’s (FDA’s) evaluation and monitoring of drug safety (see, e.g., Grassley, 2004; Harris, 2004; Horton, 2004; Vedantam, 2004; Eisenberg, 2005; Okie, 2005). Particular questions were being raised about the timely availability of information to clinicians and the public from postmarket studies of drugs previously approved by the agency. Other questions focused on the independence of FDA assessments and decision making, the balance between premarket and postmarket regulatory programs, and the adequacy of resources for postmarket safety monitoring.
By and large, the public concerns about FDA responsibilities for drug safety have not extended to medical devices. This report has noted that drugs and devices differ in a number of significant ways and that Congress and FDA have reasonably recognized those differences. At the same time, previous chapters of this report have also raised concerns about the limitations of adverse event reporting and analysis and the monitoring of—and public availability of information from—postmarket studies of medical devices. Thus, with respect to certain critical aspects of FDA procedures and performance, differences between drugs and devices are not so acute, and many of the same questions that have been asked about FDA postmarket surveillance of pharmaceuticals can be asked about its surveillance of medi-
cal devices. Furthermore, to the extent that public and professional trust in the FDA is affected by controversy over the agency’s monitoring of drug safety, it is appropriate to be concerned about a loss of public confidence in the agency’s performance in other arenas.
Much of the discussion in this report pertains to both children and adults because adequate regulatory safeguards for children require an effectively performing basic program for medical device regulation. Upon that foundation, resources and expertise focused on the special needs and characteristics of infants, children, and adolescents should be added. As described in Chapters 1 and 2, Congress recognized this in the Medical Device User Fee and Modernization Act of 2002 (P.L. 107–250), which directed FDA to develop guidance and other resources to strengthen the premarket assessment of medical devices used with children. The Center for Devices and Radiological Health (CDRH) has recently created a pediatric webpage, a useful step to encourage attention to pediatric concerns related to medical devices. Responding to further legislative directives in the Medical Devices Technical Corrections Act of 2004 (P.L. 108–214), the agency solicited views and prepared a report on barriers to and possible incentives for the development of devices to meet children’s special needs and characteristics (FDA, 2004y). Box 7.1 summarizes key points in that report.
Other legislation has also directed FDA attention to pediatric issues, mostly related to pharmaceuticals. Among a number of provisions related to children, the Best Pharmaceuticals for Children Act of 2002 (P.L. 107–109) directed FDA to create an Office of Pediatric Therapeutics to be responsible for coordinating and facilitating FDA activities that affect children and the practice of pediatrics. The Office was created in October 2002. The legislation also directed the creation of a Pediatric Advisory Committee to advise the FDA Commissioner on pediatric issues (see, FDA, 2003k). Although pediatric therapeutics includes the use of medical devices, the website for the Office of Pediatric Therapeutics focuses nearly exclusively on drugs.
Beyond the existing CDRH pediatric webpage and guidance documents, FDA should act to organize a central point of knowledge and coordination for pediatric issues within CDRH specifically. That focal point or unit would be responsible for evaluating the adequacy of the Center’s access to pediatric expertise and its attention to pediatric issues in all phases of the Center’s work, including premarket reviews, postmarket surveillance, device labeling related to pediatric use, and outreach to other public and private organizations and individuals concerned with children’s health and well-being. For example, to support families caring for a device-reliant child in the community, the unit could work with the Center’s home health committee to develop easier and more accessible adverse event reporting opportunities and more resources on safe device use as recommended in
[N]o commonly accepted definition of “unmet pediatric device need” exists…. [W]ithout a clear definition of this term, identifying the needs of pediatric patients cannot be fully accomplished (p. i).
The majority of commenters [responding to the Department’s request for comments] representing all groups indicated that the cost of developing pediatric medical devices is the most significant barrier to the development of new pediatric medical devices. The economic challenges noted include not only the limited size of the pediatric device market, but also that the return on the investment required to develop and test pediatric devices usually falls below the profit goals of most medical device companies (p. 12).
[I]t is unknown what the relative contributions are of the small market for pediatric devices, insufficient [health plan] reimbursement, difficulty in obtaining clinical data for FDA approval, and perceived increased liability associated with pediatric devices to the creation of unmet needs in this patient population. In addition, the barriers actually contributing to the existence of unmet needs may not be the same for each device (p. i).
[R]ecommendations for fostering the development and availability of pediatric devices … generally fall into the following categories: legislative actions; regulatory actions; funding for research and development; financial incentives; and enhanced information gathering and exchange (p. ii).
[F]urther discussion and study are necessary to evaluate which proposals might be more, or less, effective in addressing the barriers to pediatric device development generally or which proposals might be more, or less, effective in facilitating the development of specific categories of pediatric devices (e.g., for pediatric cardiology) (p. ii).
[I]t is premature to recommend any substantive policy changes, including administrative and legislative changes, at this time…. [T]he next step is to conduct a systematic needs assessment to determine the scope of unmet device needs in the pediatric population (p. ii).
SOURCE: FDA, 2004y.
Chapter 4. More generally, the unit could promote partnerships with industry, health care providers, patient advocates, and others to strengthen device surveillance in pediatrics and demonstrate its value in reducing risks to children.
Recommendation 7.1: FDA should establish a central point of responsibility for pediatric issues within the Center for Devices and Radiological Health to evaluate the adequacy of the Center’s use of pediatric expertise and its attention to pediatric issues in all aspects of its work.
The remainder of this chapter summarizes what the committee concluded about FDA postmarket surveillance of pediatric medical devices. (A complete list of recommendations is included at the end of the summary in the front of the report.) The discussion recognizes concerns about FDA resources and notes some of the tensions in public policy that complicate FDA’s work. The final section emphasizes that assuring the safe design and use of medical devices is a shared responsibility.
FDA PERFORMANCE OF POSTMARKET SURVEILLANCE OF MEDICAL DEVICES
The basic goal of FDA’s program of adverse event reporting and postmarket studies for medical devices is to protect patients from harm by identifying and evaluating potential or actual safety problems with legally marketed devices. Appropriate responses to problems in the form of safety alerts or public health notifications and monitoring of device recalls are also part of the postmarket framework to protect patients from unsafe devices as are inspections of device manufacturing sites. These activities should operate in tandem with the agency’s premarket activities, which include its guidance to developers and manufacturers on the design and testing of medical devices and its assessment of the safety and effectiveness of devices that require clearance or approval before marketing. The feedback of information from the postmarket unit to the premarket unit within CDRH is a critical step to assure device safety.
Postmarket surveillance, ideally, should promote the early detection of previously unsuspected safety problems as new devices are used with larger and more varied populations and in different circumstances than those evaluated before market approval or clearance. At the same time, it should take a long-term view, supporting investigations or monitoring that can identify problems as devices are used for longer periods than during their initial testing. Surveillance should cover the use of devices in all settings of patient care, including the home.
As undertaken by FDA, postmarket surveillance should be seen as objective, trustworthy, and effective. The program should seek to minimize avoidable constraints on beneficial innovation while also serving as a resource and stimulus for product improvement. The identification of problems with the design or use of medical devices should be understood as a shared responsibility for the clinicians, hospital staff, manufacturer person-
nel, researchers, and others who supply the basic data or undertake the studies necessary to achieve the goals of surveillance.
With respect to the questions posed for the Institute of Medicine (IOM), preceding chapters of this report have noted some shortfalls from adequate performance. As summarized below, these shortfalls, by and large, are not specific to pediatric questions, so responses must be general.
Monitoring of Postmarket Study Commitments
The most obvious deficits in the FDA’s program of postmarket surveillance are, first, the lack of effective procedures to monitor the fulfillment of postmarket study commitments and, second, the lack of public information about the status and results of studies ordered by the agency. In the absence of a study monitoring and reporting system, neither the agency nor the committee could reliably identify required postmarket studies (or negotiated voluntary studies) that included pediatric questions. To the extent that premarket approvals of devices identify important follow-up questions that are subsequently ignored, FDA’s failure to monitor the status of studies threatens confidence in both premarket and postmarket protections.
As described in Chapter 5, the agency has announced plans to institute systematic monitoring of postmarket study commitments for medical devices, but it has not released specific details. For example, how much information will be available to the public? Will studies involving children be readily identifiable? Will additional resources be available for new monitoring and reporting activities? If the monitoring of postmarket study commitments is simply an addition to the workload of those responsible for evaluating adverse event reports and other postmarket surveillance activities, then progress in monitoring may be offset by reduced performance in other areas.
The agency has until 2007 to submit a required report to Congress on the extent to which companies comply with postmarket study commitments. The agency should strive to have a monitoring and reporting system in place well before then. That system should allow searches to identify any studies directed at pediatric populations, and it should also cover voluntary studies negotiated between FDA and manufacturers.
Public Access to Information from Postmarket Studies
Monitoring of postmarket study commitments is important, but so is greater openness about study methods and findings. Although negative study findings may prompt a safety alert and positive findings may prompt publication and promotion of those results, no reliable avenue now exists to disseminate study findings—especially negative information—that may
be of value to clinicians, patients, and families as they make treatment decisions. Furthermore, when information is made public, it can be difficult to assess if details are not also provided about the methods employed to generate it. Such details are particularly important to help clinicians and other knowledgeable intermediaries evaluate reported results and advise patients and families.
Thus, beyond reports on the status of studies, Congress and the agency should provide for more open access to information about required studies, including summaries of findings that take into account the quality of the study’s methods and threats to the validity and generalizability of findings. The details of open access will require careful consideration and consultation with researchers, manufacturers, and other interested parties so that the agency does not publicize findings from studies that are badly designed, poorly executed, or inappropriately analyzed. The thinking now being applied to the design of a registry of clinical trials and the operation of an independent drug safety board may produce useful guidelines for such access.
Adequacy of Postmarket Studies
With respect to whether FDA-ordered postmarket surveillance studies last long enough or are adequate to assess the impact on device performance and longevity of children’s active lifestyles or their growth and development, the committee lacked sufficient information to identify such studies and answer the questions posed to it. From the information it did locate, the committee expects that many important questions about long-term outcomes remain unstudied. By continuing to direct more attention to pediatric issues related to both medical devices and drugs, Congress will encourage more systematic consideration of long-term safety and effectiveness within and beyond FDA.
FDA’s authority to order postmarket studies of medical devices is limited in some important respects. One limitation is that for devices already cleared or approved, the agency cannot order studies to last more than 3 years. For children, some important developmental consequences may not be evident within that period. Another limitation is that the agency cannot order postmarket studies as a condition of its clearance of a device under 510(k) procedures, although it can order such studies after clearance. Just as FDA can and in some instances does ask for clinical data before clearing a medical device, it is reasonable for the agency to have the authority to ask at the time of clearance for additional information to be collected after a device is marketed. Thus, as recommended in Chapter 6, Congress should provide FDA with the authority to order postmarket studies as a condition of clearance for those devices for which it can now order Section 522
Postmarket Surveillance. It should likewise authorize studies lasting longer than 3 years if appropriate and reasonable to assess safety and effectiveness with children.
The committee recognizes that requiring more postmarket pediatric studies could, in some cases, prompt a device manufacturer to label a device as not indicated for children rather than risk the extra costs and complications of postmarket studies. Negotiation with manufacturers and consideration of alternatives to required studies may be more constructive in some situations.
The committee also recognizes that most postmarket research does not result from FDA requirements but is undertaken voluntarily by industry, academic, and other researchers. As FDA increases its consideration of pediatric use of medical devices, it should not focus solely on manufacturers to develop more systematic postmarket information on the safety of medical devices used with children. Some studies may be more appropriately sponsored or undertaken by the National Institutes of Health (NIH), the Centers for Education and Research on Therapeutics (CERTs), pediatric professional societies, academic medical centers, and other health care or research organizations. Agency advisory committees may want to consider directing some suggestions for further study to these groups. As noted in Chapter 6, the Agency for Healthcare Research and Quality (AHRQ) recently announced that it planned to fund a new CERT that would focus on medical devices.
CDRH should also be provided with resources for its own program of extramural research. In addition, as recommended in Chapter 6, FDA should collaborate with NIH and others to develop a research agenda and priorities for postmarket studies—especially long-term studies—of devices with substantial pediatric use. Given that there is no shortage of important questions, Chapter 6 listed some criteria that could be used by FDA and others to set priorities for postmarket studies of medical devices.
Adverse Event Reporting
Judgments about the adequacy of the FDA program of adverse event reporting must take into account the generally recognized problems with such reporting described earlier in the report. Underreporting and incomplete or inaccurate reporting are not confined to this program. In some important respects, substantial progress in detecting, reporting, understanding, and preventing adverse device events and medical errors involving devices will depend less on FDA regulations than on the results of individual and collaborative efforts by hospitals, health plans, professional societies, and others to increase professional and provider accountability for patient safety. Thus, FDA should continue to encourage, monitor, and
participate in adverse event reporting and other patient safety initiatives beyond its own boundaries.
It is essential, however, that FDA continue efforts to strengthen its own spontaneous reporting program for medical devices and its pilot MedSun program. Chapter 4 offered several recommendations for program improvements. It also expressed concern about the adequacy of resources for postmarket device surveillance.
In the MedSun program, the agency has an opportunity to use the participating children’s hospitals as connecting points for broader efforts to improve device-related surveillance and problem reporting at the 200 or so children’s hospitals and related institutions nationwide. Many of these institutions already cooperate in various safety and quality improvement efforts. Moreover, improved surveillance in these institutions can generate information that may help improve care even in facilities that do not emphasize pediatric services.
FDA and CDRH specifically are uniquely situated to promote attention to events, errors, and design problems related to medical devices. The value-added potential of FDA’s device-specific expertise and programs is worth mentioning in several specific areas.
One area is the promotion of more attention to the safety of complex medical devices used in the home and to resources to support patients, families, and other home caregivers in safely operating and maintaining devices and getting timely assistance in resolving problems. Although FDA has created a home care committee, more needs to be done. With responsibility so dispersed compared to hospital care, it will take considerable creativity and commitment to weave a better supporting structure of device safety and surveillance for adults and children cared for outside of medical facilities.
Another important value-added area for FDA, as recommended in Chapter 6, is further promotion of common standards and approaches to capturing data about medical device safety and effectiveness from large automated patient care databases. One priority here is continued work on a commonly accepted coding system that will allow more precise identification of specific types and models of devices than is possible with existing coding. Improved coding will strengthen multiple facets of postmarket surveillance.
Although its work focuses on device safety in the United States, FDA also participates in international efforts to harmonize regulatory policies, share reports on serious device problems in a timely fashion, and promote effective manufacturing facility inspection and other programs to assure device safety in global markets. One area of concern is confidentiality policies or practices in other countries that may impede the timely sharing of critical public health information.
Additional or Independent Review of Safety Issues for Medical Devices
The IOM committee observed with interest FDA’s announcement in February 2005 of the creation of an independent Drug Safety Oversight Board within the agency but outside the Center for Drug Evaluation and Research (CDER). Noting that “[t]he public has spoken and they want more oversight and openness,” the Secretary of the Department of Health and Human Services (DHHS) said the new unit would provide “emerging information” to clinicians and patients about the risks and benefits of medicines (FDA, 2005i, unpaged). That is, discussion of potential safety problems would not wait until FDA reached conclusions firm enough to prompt a labeling change, safety alert, or other action. The board would, in addition, oversee CDER’s management of high-profile drug safety issues. Members of the board would include experts from FDA and elsewhere in DHHS and other departments, and they would consult with outside experts and patient and consumer representatives. FDA has said it will seek ideas from the public about how it might responsibly disseminate information prior to determinations about regulatory action. Many questions about details remain. One important issue is whether resources for the Board will be diverted from other activities important to public safety.
Notwithstanding certain differences between drugs and devices, the basic criteria for evaluating information and responsibly making emerging drug safety information public should—if sound in content and application—apply in broad outline to similar information involving medical devices. As the drug safety board begins to function, the relevance to devices of lessons learned with the new program should also be considered.
At this point, it is not clear that the independent board approach—that is, a unit separated from the relevant FDA center—makes sense for medical devices, although it may. More than is the case for drugs, administrative separation could limit the extent to which information gained in postmarket surveillance of devices feeds back into premarket evaluations and other activities and vice versa. It may also be considerably more difficult for devices than for drugs to draw adequate clinical and technical expertise from various government agencies to help support a separate unit. Finally, if postmarket safety oversight for both devices and drugs were to be combined in a single separate board, the committee is concerned about the potential for device safety to be subordinated to drug safety and for differences between devices and drugs to be ignored.
The committee also notes that the charter of an existing FDA group, the Drug Safety and Risk Management (DSaRM) Advisory Committee, mentions the possibility of the group acting as a medical device panel and, in that event, adding nonvoting representatives of consumer and industry
interests (FDA, 2004d).1 The DSaRM committee now includes members with expertise in risk perception, risk management, clinical pharmacology, clinical research, pharmacoepidemiology, and medication errors. Differences between drugs and devices and the lack of device-specific expertise on the DSaRM group suggest that it is not properly constituted to consider specific issues related to medical device safety and that simply adding representatives of consumer and industry interests would not rectify that deficit. (Through early 2005, all meetings of the group appear to have involved drugs only.)
RESOURCES TO SUPPORT POSTMARKET DEVICE SURVEILLANCE
The IOM was not charged with evaluating the funding and staffing of CDRH. Still, in examining the adequacy of FDA postmarket surveillance, it is difficult to avoid questions about whether the pressure for speedier approval and clearance of devices has led to an imbalance in resources for premarket and postmarket activities. This pressure is evident in the user fee legislation described in Chapter 3 and in the CDRH 2000–2005 strategic plan, which sets specific goals for reducing the time for device approvals (see, e.g., FDA, 2003f).
In its performance plan for FY 2004, CDRH stated that “recent improvements in review time have been achieved by diverting significant resources from other FDA programs” (FDA, 2003n, unpaged). The performance plan also explicitly characterized several areas within CDRH as “not working well.” These areas included postmarket surveillance of medical devices, inspection of device manufacturers, and radiation safety. (The plan characterized the mammography quality standards program as “working well” and the device review and regulatory science programs as “working but facing challenges.”)
Because clear numerical benchmarks exist, the most obvious shortfall in performance involves FDA’s statutory requirement to inspect facilities that manufacture Class II and III devices every 2 years (21 USC 360(h)).
Instead of inspecting half of such facilities each year, the percentage—and even the expressed goal of the agency—has been less than 25 percent (excluding inspections under the new program allowing accredited outside entities or persons to conduct certain inspections) (FDA, 2003n, 2004t). (The inspection requirements apply to an increasing number of device firms, approximately 13,700 in 2001 compared to 9,000 in 1997 [FDA, 2003n]). The FY 2004 performance plan noted that the 1997 legislation (FDA Modernization Act) had reduced the role of premarket review for many low-and medium-risk devices and, in theory, increased reliance on postmarket quality systems, a reliance not supported by resources for monitoring manufacturer performance. The plan also reported high rates of violations in past years.
Others have expressed concern about FDA resources. In 2001, an external review of the role of science at CDRH suggested that the Center’s scientific infrastructure may be “eroding in areas such as information systems, laboratory equipment, and training” (FDA, 2001l, unpaged). It particularly stressed concerns that the work of the FDA across the life cycle of medical devices was least attentive to “… the feedback loop from postmarket review of one device to pre-market design of its successor. This segment of the cycle is most heavily dependent on recording present experience and passing that on to the next generation of designers (and reviewers/evaluators). In this regard, it was noted by staff that there is no general catalogue or electronic database of decisions reached, or the basis for them, so that undue weight is placed on individual people remembering what happened in some earlier, related situation.”
Improvements in information technology and information systems in use at FDA should improve the efficiency of agency operations and make information—for example, the status of postmarket studies—more widely and easily available to those who need it. Such improvements cannot, however, replace the application of substantive knowledge and judgment by a sufficient complement of staff with appropriate clinical, engineering, statistical, and other expertise. Likewise, advance guidance on FDA expectations of device developers and manufacturers can reduce the burden on FDA staff to catch and require correction of deficiencies in these processes, but they do not substitute for on-site inspections that include investigation of complaint handling and adverse event analysis and reporting.
In the Medical Device User Fee and Modernization Act of 2002, Congress authorized additional appropriations for postmarket surveillance activities—$3 million for FY 2003, $6 million for FY 2004, and “such sums as may be necessary” in subsequent years. After this endorsement of the agency’s need for additional resources, Congress did not, however, appropriate the additional funds (personal communication, Thomas P. Gross, M.D., Director, Division of Postmarket Surveillance, CDRH, March 28,
2005). The legislation also provided for an FDA report to Congress by January 2007 on the impact of user fees on postmarket device surveillance and the improvements and funding that may be needed for an adequate program.
TENSIONS IN PUBLIC POLICY AND DEVICE INNOVATION
Through statutes and regulations, FDA is charged with simultaneously safeguarding the public and encouraging timely access to safe and effective products. Tensions sometimes exist between these two broad roles of FDA, as suggested at several points in this report. Tension may also exist between the public’s desire for governments to protect them from an array of threats to their health and safety and the public’s willingness to pay for such protection.
Trade secret and confidentiality provisions related to studies of medical devices and other FDA-regulated products may create tensions between the desire to encourage product innovation and economic growth by allowing innovators and manufacturers to hold certain information secret and the desire to provide clinicians, patients, and others with information that may help them make knowledgeable choices. Given the scarcity of information about the safety and effectiveness of medical devices used with children, this report has argued for greater openness and availability of findings from device studies required by FDA.
Efforts to apply or strengthen safety regulations have the potential to discourage certain innovations. Regulations may add to administrative, testing, or production costs to the point that developers or manufacturers may consider development or marketing of a device economically unattractive. As noted in Chapter 1, the device industry is quite diverse compared to the pharmaceutical industry and includes a larger proportion of small firms. Some of these small companies may struggle with the many facets of medical product regulation, including special requirements for testing medical products with children. Regulations may also create uncertainty among investors about the future of a device, including whether it will be approved for marketing. For example, required reports by publicly held companies often include statements about risks or uncertainties related to clinical trials that are being undertaken to obtain FDA clearance or approval of a device.
For products that are developed primarily for adults and that have a relatively small pediatric market, the regulatory, ethical, practical, financial, and other added costs or challenges of testing a product with children and securing market approval may deter manufacturers from proceeding—especially if existing approval of the device has not restricted use with children. (As described in Chapter 6, special regulatory protections for children involved in research can impede certain kinds of research [IOM,
2004a].) Without systematic testing either before or after marketing, however, problems with pediatric uses of devices may be detected slowly and unsystematically. Then, if questions arise about unlabeled use of a device with children and the absence of pediatric studies, a manufacturer may choose to label a device as not indicated for use with children rather than undertake pediatric testing. This may limit children’s access to beneficial devices and increase the potential liability of health care professionals and providers.
The potential for regulations to create unanticipated and unwanted outcomes—such as discouraging beneficial innovation or limiting the information available to guide clinical decision making—makes it important for policymakers to consider the effects of policies on different relevant parties, with attention to the incentives that shape their behavior. Consistent with the protection of patient safety, policymakers should encourage an alignment of incentives and interests among manufacturers, professionals, providers, patients, and regulators that favors cooperation and shared responsibility over blame or conflict.
SHARED RESPONSIBILITIES FOR MEDICAL DEVICE SAFETY
Context: Shared Responsibilities for Patient Safety
FDA programs to protect public health operate within a broader system of shared responsibilities for patient safety and health care quality. In the past two decades, individual and collaborative initiatives to improve the quality of health care and protect patients from harm have grown remarkably and been the subject of many reports (see, e.g., IOM, 1990, 1992, 2000c, 2003, 2004b; AHRQ, 2001, 2003, 2004). A number of these initiatives are sponsored or supported by government agencies, including the various units of DHHS, the Department of Defense (which manages a large military health care system), and the Veterans Health Administration. Other patient safety activities have been sponsored by a growing number of state governments, private-sector coalitions of many kinds and sponsorship, health care accrediting organizations, consumer groups, and health professional societies.
In addition, hospitals and other health care institutions have acted individually as well as in collaboration with others to improve their own systems and promote change in larger systems within which they operate. Health professions educators also contribute to patient safety when they teach—and demonstrate by example—the importance of professional vigilance, including the prevention, recognition, and investigation of device adverse events. This diversity of involvement reflects not only the broad
concern about health care quality and patient safety but also the range of parties whose participation is necessary to improve health care outcomes.
Patient safety initiatives often emphasize drug safety or otherwise do not highlight quality or safety issues that primarily involve medical devices. The focus on medications reflects analyses of medical errors in which medication mishaps figure prominently, although some of these mishaps also involve errors in device use or shortcomings in device design as described in Chapter 4. Like most patient safety initiatives, those focusing on children also tend not to feature medical devices.
Nonetheless, even initiatives that focus narrowly on medication safety or other topics may still encourage practices, procedures, and ways of thinking that can—indirectly or directly—help create an environment that promotes medical device safety. For example, organizational steps to improve team functioning, adherence to medication protocols or guidelines, accuracy in the recording of medication information, and patient and family education can have broader potential benefits. Likewise, efforts to improve the analysis of medication errors through root cause analysis should create skills and concepts that can be applied to other products or processes. In addition, certain specific actions to improve medication safety may at the same time improve medical device safety. Examples include training in the proper use of infusion pumps and redesign of these pumps to reduce opportunities for use errors.
Beyond appreciating these spillover effects, those concerned about the safe design and use of medical devices should examine quality of care and patient safety initiatives and consider how they might be adjusted or expanded to include devices. For example, as procedures to improve tracking and responses to drug safety alerts and recalls are instituted, the need for corresponding procedures tailored to the institutional intake, distribution, and day-to-day use of medical devices should also be evaluated.
Recommendation 7.2: All those engaged in improving the quality of health care and protecting patients from harm should evaluate and sharpen as appropriate their attention to medical device safety, including safety issues that particularly affect children.
As an example of how organizations might expand their focus on medical devices, the committee looked at a set of 20 tips for families to prevent medical errors with children that were developed by the Agency for Healthcare Research and Quality (AHRQ, 2002). Eight of the 20 tips involved medications, for example, “If you have any questions about the directions on your child’s medicine labels, ask.” Only one question (which recommended asking about the best device for measuring a child’s liquid medicine) dealt with a device. Some questions could have been modified to include medical devices, and device-specific questions could have been added. Examples of
such questions (some of which draw on items in FDA’s home health checklist (FDA, 2003l) described in Chapter 4) include the following.
Ask for written information, videotape, DVD, or Internet resources on how to safely operate your child’s medical device, what to avoid doing, how to recognize and troubleshoot problems, and when and where to seek assistance.
If you have any questions about your child’s medical device, ask your doctor.
Make sure that you have the name of the device, the model and serial numbers, the manufacturer’s name, and contact information.
Be sure your child’s doctors know that the child has an implant and relevant information is clearly recorded in your child’s medical record in the event that neither you nor your child’s physician is available to provide this information in an emergency.
Keep in touch with the surgeons or other clinicians involved in implanting a device or otherwise treating your child with a device, so they can find you if they become aware of safety concerns with the device or associated treatment.
If the manufacturer of your child’s implant has asked you to do so, inform them of your new address when you move.
Medical Device Safety
The sharing of responsibilities for making safe medical devices available for use with children extends throughout the medical device product cycle depicted in Chapter 1. That is, it covers both premarket and postmarket activities and the interaction between the two. Rather than identify roles and responsibilities by categories of individuals and organizations, the discussion below is organized by broad stage in this cycle.
Shared Responsibilities: Medical Device Innovation
As discussed in Chapter 6, a first line of safeguards for children begins with those who conceive of and develop high-risk medical devices. From the earliest stage of device development, including the design of prototype devices, engineers, clinicians, and others involved should systematically consider the potential harms as well as potential benefits of a device. Such consideration should include the possible use of a device with children.
FDA contributes at this stage by providing general or specialized guidance to help innovators understand from the outset what is necessary to secure FDA approval or clearance of certain kinds of medical devices. FDA’s recent guidance on premarket assessment of pediatric medical de-
vices is a helpful step, but further work is needed to identify and clarify characteristics of children that may affect particular aspects of performance and safety for different types of devices.
When possible and especially for devices used at home, designing devices to avoid hazards and harms is preferable to relying on professional and consumer instructions, education, and other steps that may be less effective at preventing problems (Arcarese, 2002b). The FDA human factors initiative has provided guidance to device developers on ways to identify and avoid hazards related to medical device use as part of a larger program of risk management (FDA, 2000b).
When problems do arise with existing devices, feedback from users can be an important source of inspiration or direction for the modification of such devices and the development of new devices. Users include health care professionals and patients or families, but risk managers and others in health care facilities may also provide important insights into a device’s strengths and limitations. Many companies have active surveillance programs that support and encourage this feedback of information through multiple means, including field representatives. The yield from these programs—both to manage problems with existing devices and to get direction for refinements and innovations—depends to a considerable extent on whether health professionals and, sometimes, consumers, are aware of their reporting options and are able and willing to use them.
Shared Responsibilities: Systematic Clinical Evaluation of Medical Devices
Beyond the initial stages of conceptualization and development, the systematic evaluation of medical devices, which includes preclinical and clinical testing of higher-risk devices, provides another important set of safeguards for children and another area of shared responsibilities. In addition to tracking hypothesized outcomes with innovative devices, clinical studies identify and evaluate unanticipated adverse events and thereby build the evidence base for device labeling and, subsequently, patient care. FDA has promoted early meetings with device developers to discuss testing and evaluation strategies that will satisfy its standards of evidence when approval is sought to market the device or add new indications for an already approved device. When manufacturers undertake postmarket studies directed by the FDA, they also seek approval of the study plan.
As it regulates and guides the evaluation of medical devices, FDA operates as part of a substantial research infrastructure in the United States that assigns varying degrees of responsibility for overseeing the scientific and ethical soundness of clinical research to a range of parties. Besides FDA, these parties include NIH and other research sponsors, research institutions
such as academic medical centers, and the Institutional Review Boards (IRBs) that review proposed human research for conformity with federal and institutional research protection policies, including special protections for children. FDA has provided guidance for IRBs and others about certain special aspects of device research (e.g., different requirements for “significant risk” devices versus “nonsignificant risk” devices).
This report has outlined some of the challenges involved both in conducting research with children and in conducting clinical studies of medical devices. The latter may involve testing practices that are different from those normally demanded for investigational drugs, for example, the routine use of randomized clinical trials. As a result of such differences, threats to valid research conclusions that may be controllable in drug trials may not be controllable in device studies. This potentially makes the process for evaluating devices before market approval less powerful and increases the burden on postmarket surveillance.
The world of medical device development and evaluation is diverse. It includes companies with focused teams of scientists, engineers, and clinical experts who are backed by significant organizational, financial, management, and other resources to support the translation of new ideas into safe, effective, and profitable products. It also includes smaller firms with more limited resources and, perhaps, more need for guidance from FDA. Less obvious is the role of clinical innovators who are seeking better ways to treat their patients’ medical problems and who may operate independently without appreciating that they share responsibility for systematic evaluation of medical interventions.
Although established medical device companies may be able to improve their clinical evaluation procedures, the individual clinical innovator is a weaker link in the process of testing new interventions for safety and effectiveness. New surgical procedures involving medical devices and modifications of medical devices can escape systematic assessment, despite calls for such innovations to be brought into the system of clinical evaluation and human research protections (see, e.g., Solomon and McLeod, 1995; Moss et al., 2001; Ehrlich et al., 2002; McKneally and Daar, 2003; Dent et al., 2004). Manufacturers may be concerned about certain kinds of device modifications that occur outside their infrastructure for testing, evaluation, and follow-up.
Trying to distinguish between progressive, acceptable clinical practice and unscientific and possibly unethical research on humans has proved difficult for the clinical community, research ethics organizations, FDA, and others concerned about the evidence base for clinical care and the ethical conduct of human research. Much of the difficulty in making distinctions stems from the complexities of patient care, especially care for high-risk children who are not being helped by standard practices and for whom specifically appropriate medical devices are not available. Despite
the frustrations and complexity of the task, efforts should continue to evaluate the safety and effectiveness of surgical and device innovation in clinical practice.
Shared Responsibilities: Acting on and Disseminating Findings of Device Studies
For Class III and certain Class II devices, FDA clearly has responsibility for evaluating clinical data and deciding whether it is sufficient to warrant the approval or clearance of a medical device so that a device can be legally marketed. For devices that have gone through the approval process, the agency makes public summaries of the evidence of safety and effectiveness (or safety and probable benefit). It generally does not make public any information about postmarket studies unless their findings prompt a recall, change in labeling, or public health notification. Companies may, however, promote positive study findings from such studies through press releases and other means, including journal publications.
FDA regulates the label and labeling of devices, which may include information based on study findings. Labeling is broadly defined to include promotional materials. (FDA can also regulate the advertising of “restricted devices,” i.e., devices whose sale, distribution, or use is restricted by an approval order or by regulation.) The FDA has provided extensive guidance to manufacturers on various aspects of labeling, including the development of information for lay users of home medical equipment (FDA, 1993b).
Within these boundaries, manufacturers have the opportunity and responsibility to provide understandable information and education on the safe and effective use of devices available to clinicians, patients and families, and others who need it. Clearly, the extent and type of information and educational support will vary depending on the complexity of a device and its risks. Professional societies also share responsibility for providing information and evidence-based guidance for clinicians and, in some cases, patients and families.
As described in Chapter 3, FDA operates within a set of statutory restrictions and traditions that have limited the sharing of information about findings of device studies, especially studies conducted following initial approval or clearance of a product for marketing. This report has argued for less secrecy and more openness in the sharing of information related to the potential benefits and harms of medical devices.
Shared Responsibilities: Safety in Device Manufacturing and Supply
Manufacturers must devise and sustain production processes that reliably produce safe products consistent with design specifications and appli-
cable public policies. The vast majority of medical devices are manufactured without defects, but each year sees some device recalls stemming from serious aberrations in manufacturing or packaging processes. These aberrations affect both complex devices and simple devices that normally present low risk to patients. Making devices safely available also requires the participation of distributors, vendors, user facilities, and others who have custody of the devices prior to their use with patients.
FDA provides extensive guidance on manufacturing processes and quality systems for medical products. It also has a system for inspections of manufacturing sites. As noted above, that inspection system has fallen substantially below statutory requirements for frequency of inspections in recent years. Although the reliable and consistent production of safe devices ultimately depends on the integrity and competence of the manufacturer, shortfalls in the FDA inspection program mean that some problems and opportunities for improvement are almost certainly not being identified.
Shared Responsibilities: Safe Use of Safe Medical Devices
Whereas manufacturers have the lead responsibility for the safe manufacturing of effective medical devices that are designed to minimize opportunities for unsafe use, health care professionals and providers have lead responsibility for the safe and appropriate use of devices. The emphasis in discussions throughout this report has been on institutional environments that help health care professionals and others do what they are supposed to do correctly and consistently. Still, individuals within these systems must understand and welcome their personal responsibilities to acquire and apply the knowledge and skills—including communication and other interpersonal skills—needed for safe patient care.
Initial and continuing education is an essential element in preparing physicians, nurses, and other health care professionals to use conventional and innovative medical devices safely and appropriately and to recognize device problems that should be reported. Beyond the education provided by health professions schools, health care facilities, and professional societies, manufacturers of complex devices must also provide or support proper training and assessment for clinicians. This is especially important for innovative products that require new procedures or skills. Simulation strategies may prove particularly helpful for complex devices and procedures that require unusual or particularly difficult skills and techniques.
Familiar devices and familiar safety problems also warrant public and private attention. An example of such attention involves the well-recognized threat of fire when certain devices such as defibrillators are used in oxygen-rich environments. In a 1994 publication, ECRI (a private technology assessment and research organization that has produced a number of fire
safety statements) observed that the danger of fire “during defibrillation has been forgotten or ignored—or was never learned—and must be periodically reemphasized” (ECRI, 1994, p. 307). In 2002, the American Heart Association amended its pediatric life support recommendations. Instead of a 3-step chant to insure people are clear of a patient before an electric shock, hospital personnel are now advised to use a 4-step chant to emphasize the danger of oxygen. Specifically: “I’m going to shock on the count of four. One—I’m clear. Two—you’re clear. Three—oxygen’s clear. Four—everybody’s clear” (Hazinski et al., 2002; cited in Theodorou et al., 2003). Other fire safety initiatives include a safety alert issued by the Joint Commission on the Accreditation of Healthcare Organizations as part of its sentinel event program (JCAHO, 2003).
The responsibility of patients and families for the safe use of medical devices is a more sensitive topic. Families whose children are reliant on complex medical equipment in the home are often under exceptional stress—emotionally, physically, cognitively, and financially. Unless they are fortunate enough to have excellent health insurance or substantial financial resources, they typically lack adequate 24-hour support from trained personnel to operate, maintain, and troubleshoot complicated, life-sustaining equipment. Some children’s hospitals and other institutions devote considerable resources to training parents in the use of equipment, evaluating home health agencies and vendors, checking the initial use of equipment in the home, and following up periodically on adherence to care instructions (Mast, 2004; Simpser, 2004). Likewise, some manufacturers have good consumer support programs. In general, however, much more can be done by primary care and specialist physicians, hospitals, home care providers, manufacturers, and FDA to help patients and families meet their responsibilities for using medical devices safely.
Shared Responsibilities: Postmarket Surveillance and Evaluation of Long-Term Safety and Effectiveness
Reflecting its charge, this report has focused primarily on FDA’s responsibilities for postmarket surveillance of medical devices, including devices used with children. The discussion of adverse event reporting made clear, however, that health care professionals and health care providers (user facilities) have critical responsibilities for initially detecting, documenting, investigating, and reporting adverse device events—and for putting in place procedures and practices to prevent such events. Manufacturers are crucial in investigating possible device failures and malfunctions to uncover the nature of the problem—which might be intrinsic to the device or could relate to the way it was used—and respond as appropriate. In addition, as described in Chapter 4, various other public and private orga-
nizations and collaborations include some form of adverse event reporting as part of programs to improve the quality of health care and protect patient safety. Some of these efforts involve children’s hospitals and pediatric professionals.
Just as reporting and responding to adverse events is a distributed responsibility, so is the systematic study of the safety and effectiveness of marketed devices used with children. In addition to postmarket studies ordered by FDA or undertaken voluntarily by manufacturers, NIH, the CERTs overseen by AHRQ, private foundations, pediatric professional societies, and other groups interested in child health should promote and support safety and effectiveness studies and the databases (e.g., registries) needed to support such studies. The involvement of patient, family, and consumer groups is particularly important for devising strategies to adapt device surveillance to the difficult realities of identifying, understanding, and preventing adverse events with devices used at home and in the community.
To support both adverse event reporting and systematic studies, FDA has been exploring how to use developing statistical tools and exploit large automated patient care databases to identify, understand, and prevent problems with the use or design of medical devices. As discussed in Chapters 4 and 6, initial results have been less promising than hoped, partly because patient records often do not contain information that identifies individual medical devices. Creating feasible coding systems and procedures for including such information has, however, proved daunting. Recommendations in Chapter 6 focus on steps to improve the usefulness of electronic patient records for device surveillance and strengthen FDA research capacity. In any case, university-based and other biostatisticians and epidemiologists can contribute by helping, first, to improve and apply the methods for formal clinical studies of medical devices and, second, to refine the tools for detecting and analyzing adverse events in large databases and developing the field of medical device surveillance.
Notwithstanding the responsibilities and essential contributions of other parties, FDA has an unparalleled combination of authority and expertise to promote more comprehensive, efficient, and effective medical device surveillance. It must use its position both to strengthen its own programs and to encourage others to recognize and meet their shared responsibilities for identifying, understanding, and preventing problems with the design, evaluation, manufacture, maintenance, and use of medical devices. As stressed throughout this report, successful programs of postmarket surveillance to protect all those whose health depends on safe and effective medical devices are an essential foundation for efforts to protect children in particular.