A Report to Congress by the Secretary of Health and Human Services
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration
Center for Devices and Radiological Health
Rockville, MD 20850
Table of Contents
- Feasibility of Including Device Investigations in the PHS Data Bank
- Public Health Need for Access to Information
- Impact of the Data Bank on Device Innovation and Research
- Other Issues and Concerns Raised
- DHHS Conclusions and Recommendations
- Appendix 1 - Federal Register Notices and Comments Received
- Appendix 2 - Summary Minutes Public Meeting July 8, 1999 (follows Appendix 1)
Background – Section 113 (b) of the Food and Drug Administration Modernization Act of 1997 (FDAMA) requires the Department of Health and Human Services (DHHS) to determine the feasibility of including device clinical trials within the scope of the PHS data bank established by §113 (a) of FDAMA. FDAMA mandated that the Secretary of Health and Human Services provide a report to Congress, by November 1999, on the feasibility of a device trials data bank.
The data bank that §113(a) established is limited to pharmaceutical clinical trials and is meant to be used by prospective patients, health care providers, researchers, and the general public. This provision requires sponsors of federally and privately funded studies of new drugs to treat serious or life-threatening diseases and conditions to submit information on the purpose of the experimental drug in the study, eligibility criteria, study sites, and a contact for patients wishing to enroll. FDA is collaborating with NIH to begin operation of the data bank in 2000.
To write this feasibility report for a device trials registry, FDA sought comments from consumers, investigators, and members of industry and institutional review boards (IRBs). The agency received 13 comments (see Appendix ), and about 24 people attended a public meeting FDA held on July 8, 1999. The comments are summarized below:
Public Health Need for Access to Information – Two patient advocacy groups, consumer representatives serving on FDA’s device advisory panels, and a health organization believed there is a need for access to valid, unbiased device trials information so patients and their physicians can make fully informed healthcare decisions. One advocacy group believed that access to device trials information would enhance patients’ and doctors’ expectations, speed recruitment, and bring companies investment capital.
A few manufacturers argued that any data bank should be voluntary rather than mandatory, but agreed that those with serious or life-threatening diseases or conditions that lack approved treatments need information on relevant device trials. However, some noted that patients in this category have made limited use of current FDA mechanisms to expand access to device trials, and acceptable alternative treatment exists for the diseases involved in most other device trials. Other responders noted that a significant amount of medical information is already publicly available (e.g., centerwatch.com).
Some commenters noted that many patients do not recognize the additional, often unpredictable risks associated with investigational devices. Patients may equate research with state-of-the-art treatment when standard treatment might be satisfactory.
Impact of Data Bank on Device Innovation and Research – Manufacturers stressed that, unlike the pharmaceutical industry, most device firms are small and resource poor and depend on investment capital. Many manufacturers believe that disclosing information about research could jeopardize the company itself as well as development of the device. Competitors could copy research ideas from the data bank, or shift resources to beat another company to market. Because the first firm to get a new device to market usually retains the largest market share, small firms with only a single product in development could be significantly disadvantaged.
The device industry was also concerned about the cost of preparing data bank information and of training investigators to respond to patients seeking information beyond what the data bank provided. Unlike drug trials, most device clinical trials involve a rather limited number of patients. The number of subjects is determined by the statistical rationale for the hypothesis being tested, and it usually represents the maximum the company can afford to enroll because of the cost of the device and of patient monitoring. A publicly accessible data bank might cause doctors and patients to pressure companies to increase the size of the trial. The device industry was also concerned about the legal implications of denying access as well as the increased costs of customizing plans for subjects ineligible for the protocol.
Other Concerns – Other commenters believed a data bank would result in:
- more work for investigators and already overburdened IRBs
- pressure to increase off-protocol treatment and third-party reimbursement
- the appearance that DHHS endorsed unapproved devices
- pressure from the public for more information than the data bank would contain
- direct access to information which patients should receive through their physicians.
Conclusions – The report concludes that manufacturers fear the potential negative impact of a mandatory device trials registry on the industry and on the conduct of quality clinical trials to support device approval – even if the registry is restricted to device trials involving serious or life-threatening conditions that lack alternative therapies. In addressing industry concerns, Congress and DHHS would benefit from assessing the performance of the drug trials registry. DHHS recommends that Congress should delay any decision about a device trials registry until the year 2001, after there is one year of experience with data from privately funded trials in the drug trials registry. The prototype being tested this fall contains primarily data from NIH-sponsored trials.
On balance, DHHS believes a data bank could offer those who may not otherwise have access to information about clinical trials a way to consider entering a study of an experimental device intended to treat their condition. The public health is likely to be best served if the data bank lists device trials only for life-threatening or serious diseases or conditions for which no other therapies are available. A device trials registry could make a wider audience aware of clinical trials that have had difficulty achieving full enrollment. In that situation, the public health may be served by allowing enough subjects to be identified to permit sponsors to do timely testing on promising products.
A Device Clinical Trials Data Bank - Public Health Need and Impact on Industry
Section 113(b) of FDAMA directs the Secretary, the Director of NIH, and the Commissioner of Food and Drugs to collaborate and determine the feasibility of listing clinical trials of medical devices in the data bank. In addition, FDAMA directs the Secretary to submit a report to the Senate Committee on Labor and Human Resources and the House Committee on Commerce addressing the following:
- the public health need, if any, to include device investigations within the scope of the data bank under § 113(a) of FDAMA;
- the adverse impact, if any, on device innovation and U.S. research if sponsors must publicly disclose information about device investigations; and,
- other issues relating to §113(a), as the Secretary determines appropriate.
Currently, the Food and Drug Administration (FDA) is collaborating with NIH to implement § 113(a) of FDAMA. The NIH National Library of Medicine (NLM) is developing the system in close and ongoing collaboration with all NIH institutes. NLM already houses clinical trials information through the AIDS Clinical Trials Information Service.
The PHS drugs data bank is meant to contain current information on clinical trials of experimental treatments. It is meant to be used by individuals with serious or life-threatening diseases, other members of the public, health care providers, and researchers. As the statute specifies, the data bank will contain a list of federally and privately funded trials for serious or life-threatening diseases. Each listing will describe the purpose of the experimental therapy, eligibility criteria, location of the trial sites, and a contact for patients wishing to enroll. In addition to this required information, the study sponsor can choose to add information about the results of the therapy or associated adverse events. Finally, FDAMA requires that all of the information in the data bank be in a form that the public can readily understand.
The PHS data bank will contain additional information that may help a patient decide whether to enroll in a particular trial. The data bank will include the phase of the trial, the disease or condition, and the particular therapy or intervention under study. An important feature of the data bank will be to provide access to other online health resources to help place clinical trials in the context of patients’ overall medical care.
A prototype of the clinical trials data bank is being made available for focused evaluation and testing in the summer/fall of 1999. The prototype and first version of the system will contain primarily NIH-sponsored trials. Beginning in the year 2000, NLM will collect clinical trials information from other Federal agencies and private industry to add to the system. FDA and NLM are developing a guidance document to inform sponsors of investigational drugs about their responsibilities regarding the data bank. A second document to notify the public and provide guidance on using the data bank use should be available in 2000.
To prepare this report, FDA obtained comment from interested parties, including consumers, the medical device industry, researchers, and institutional review boards. To help ensure that interested parties had the opportunity to share their views, FDA published a notice in the Federal Register (FR) soliciting public comment on the statutory provision. The FR notice summarized the provision and posed questions about the public health need for the data bank and its possible impact on device innovation and research. FDA received 13 comments. (See Appendix 1 for copies of the FR notice and comments received 1) FDA received 8 comments from industry or industry trade associations, 3 from health organizations, and 2 from the general public. FDA also invited comments from consumer groups and consumer representatives on FDA's device advisory panels. Five consumer representatives responded. A consumer and a health professional submitted 2 more comments directly to FDA rather than to the docket.
FDA and NIH believe that if Congress were to enact legislation authorizing the establishment of a device clinical trials registry, it would be feasible to include such information in a central data bank. At this time, however, additional information is necessary to determine the requirements for full implementation, and what, if any, additional resources may be necessary.
Commenters have suggested that the legislators who drafted FDAMA believed FDA would have considerable experience with the clinical trials data bank for drugs when this report issued. Although FDA and NIH have experience with other clinical trials data banks (e.g., cancer, AIDS) and are proceeding on schedule toward implementing FDAMA’s drug trials registry provision, DHHS has no direct experience with the actual operation of the drug clinical trials data bank at this time. Certainly, the early experience with this data bank will suggest ways to streamline the listing of trials of investigational devices if Congress decides that it is in the public’s best interest to include device trials in the data bank.
Implementation of the pharmaceutical trials data bank will provide important information to FDA and NIH about the development of a similar device data bank. However, there are basic differences in the make-up of the pharmaceutical and medical device industries and in the research and development and clinical use of drugs and devices. These differences may affect whether and how information about device trials should be made available in the data bank.
The following are comments from interested parties regarding the public health need for a device data bank, its potential impact on the device development process, and other related issues.
The FDAMA mandate for a data bank for certain pharmaceutical clinical trials was based on the perceived need to have access to this information. Two patient advocacy groups stated that there is a similar need for access to information about device investigations for serious and life-threatening diseases or conditions that lack approved alternative treatments. Advisory panel consumer representatives and a health organization representative noted the trend for consumers to partner with their providers as informed participants in their healthcare decisions. They contend that patients need access to clinical trial and other information for proper decision-making. They also believe that consumers have the right to know what is on the clinical horizon, including investigational medical devices. They need valid, unbiased information, which they do not always find on the Internet. One commenter stated that mandatory public disclosure would promote improved science and benefit public health although disclosure may be painful for industry.
An advocacy group for rare diseases contended that the public is not well served when drug and device development is conducted in secret. Patients with serious diseases or conditions are left searching for clinical trials in which to participate. The group cited the National Commission on Orphan Diseases’ 1989 report, which stated that a significant proportion of researchers could not find enough patients for their rare disease trials and that patients could not find investigators studying their condition. The advocacy group also noted that disclosure of device trials would speed innovation because it will enhance expectations of patients and physicians, speed recruitment of patients for clinical trials, enable physicians to locate clinical sites appropriate for their patients, and allow investors to analyze the potential growth of the company. The group noted the greater difficulty in obtaining information about device trials than for drug trials, but also noted that there is little or no need for information about "me-too" investigational devices because acceptable alternative treatments exist.
A few device manufacturers agreed that a data bank could meet a public health need if it listed trials involving devices intended to treat serious or life-threatening diseases or conditions for which no approved treatments exist. However, they stipulated that the data bank must be voluntary, so they could balance the benefit of disclosure against a possible negative effect on their device development plan. Other commenters questioned how a voluntary data bank could address a public health need. One commenter noted that voluntary disclosure would not ensure that patients received adequate information on which to evaluate their options.
Other manufacturers and private citizens noted the limited use of expanded access provisions (such as the "treatment use" investigational device exemption). This limited use, they believed, indicated that a device trials registry would not be useful if restricted to conditions for which no approved therapies exist. However, these manufacturers also argued that broader criteria would not serve a public health need either because acceptable alternatives exist for most device trials. These comments highlight the need for the agency to define "serious or life-threatening diseases or conditions" to establish the scope of the data bank.
One manufacturer stated that creation of a data bank could serve a public health need because it would facilitate trial enrollment. Other manufacturers disagreed. They stated that trial enrollment was not a concern because the pace of enrollment is dictated by availability of the device and by the need to limit exposure to the risks of the investigational device by limiting the number of devices in distribution. Comments also cautioned that any benefit a patient may receive from participating in the trial could be offset by harm the patient could experience. Other manufacturers stated that inadequate enrollment in device trials is not sufficient reason to create this data bank because FDA already permits advertising to recruit subjects.
Several responders stated that, without a clear definition of "public health need," they could not comment. Other commenters noted that a significant amount of medical information is available to the public via various sources2 and questioned the need for the additional information a device trials registry would provide.
Many comments questioned whether a device registry would serve a "true public health need" or simply satisfy a "perceived need" for additional access to investigational devices. One member of an institutional review board stated that the American public has developed the perception that investigational devices offer a benefit over current technology. The IRB member referenced the DHHS Office of Inspector General’s 1998 report,3 which stated that "the line between research and therapy has become increasingly blurred." Some comments noted that many patients do not recognize that there are additional risks, many unknown, associated with investigational devices. Patients may incorrectly equate research with state-of-the-art treatment and seek to participate in clinical trials even though the currently accepted device may be satisfactory. For such patients, the commenters stated, it would be difficult to argue that there is a true public health need for access to clinical trial information.
One of the recurring comments from the device industry was that device investigations and the industry itself are different from their pharmaceutical counterparts. Some commenters argued that these differences would cause a device trials data bank to adversely affect device innovation and research in the U.S.
One of the major issues raised by the commenters was the importance of protecting trade secret information. The trade associations, several device manufacturers, and a private citizen commented on the statutory provisions and legislative history regarding the protection of proprietary information and the importance of these to this country’s device innovation and research. The commenters noted that when Congress drafted the Medical Device Amendments in 1976, it recognized the need to protect the trade secret material submitted to FDA if competitors were "not to gain unfairly from the time and financial commitment expended in developing a device."4 Many of the comments received on § 113(b) of FDAMA emphasized the importance of a manufacturer’s "competitor’s advantage" to survival and success in the medical device industry. The industry relies on the law and FDA regulations that protect trade secret and confidential commercial information, including information about the existence and content of a trial of an investigational device.
One commenter questioned the impact that the PHS data bank could have on the Freedom of Information Act (FOIA) procedures. That commenter stated that DHHS and the device industry should consider how the disclosure of clinical trial information in the PHS data bank would affect the disclosure of more detailed information in an IDE. This comment noted that recent changes to FOIA and Office of Management and Budget procedures elicited thousands of comments that emphasized the importance of protecting the confidentiality of scientific research, research subjects, and research parties.
Manufacturers stressed that, unlike the pharmaceutical industry, most device firms are small with limited resources. They depend on investment capital to support their research endeavors. In other words, the typical device manufacturer is extremely vulnerable to competitive pressures. Many manufacturers believe that disclosing information, such as a description of the investigational device and indications for use, could jeopardize the development of a new device as well as the company itself. Comments suggested that competitors could copy research ideas garnered from the data bank, or shift resources in an attempt to beat another company to market. Comments also raised the possibility that venture capitalists would inappropriately use the information in the data bank in a way that would disproportionately disadvantage small manufacturers. Because the first firm to get a new device to market usually retains the largest market share, some manufacturers argued that small firms with only a single product in development could be significantly disadvantaged.
Finally, a trade association raised the possibility that the release of information about a device through the data bank might affect an inventor’s patent rights. Because much device innovation begins with private inventors, the association argued that the possible loss of patent rights could have a "chilling effect" on advances in medical technology in this country.
Because of concerns about the release of previously protected information, device manufacturers generally argued that any data bank that DHHS creates should be voluntary rather than mandatory. Some comments stated that mandatory disclosure could ultimately push manufacturers to conduct these important investigations offshore to protect the confidentiality of their research and financial investments.
Manufacturers pointed out that, unlike drug clinical trials, most device clinical trials involve a fairly limited number of patients. They stated that the number of subjects is determined by the statistical rationale for the hypothesis being tested. The number also usually represents the maximum the company can afford to enroll due to the costs associated with the device, patient follow-up, and trial monitoring. Commenters expressed concern that a publicly available data bank could pressure them to increase the size of their clinical trials because potential subjects would seek admittance to the trial even if enrollment were complete. Manufacturers and IRB members also expressed concern over the legal implications of denying access to these patients and the increased costs associated with customized plans for subjects who do not meet eligibility requirements.
The device industry also expressed concern that a public registry would increase pressure on sponsors to alter device trial design or sample size to address potential reimbursement issues. Device manufacturers stated that such protocol modification would significantly raise the cost of conducting clinical trials in the U.S.
Several commenters were concerned that creation of a device trials data bank would significantly affect the workload of IRBs. For example, several commenters stated that the data bank may be viewed as material to recruit study subjects under FDA’s current policies.5 Therefore, they assumed IRBs would have to review the information to ensure that it is balanced and not misleading. Further, it was stated that any information for the PHS data bank would require the initial and continuing review by FDA, the study sponsor, and IRBs. Commenters were concerned about the resources that these reviews would consume, and the loss of patient protection that could result without IRB and FDA oversight.
The Office of Inspector General’s report on IRBs of June 1998 stated that IRB workloads have increased significantly over the past 5 years, but staffing levels and budgets have remained the same. Commenters expressed concern that, for the reasons discussed above, the PHS data bank is likely to add to the workload of the IRBs and adversely affect their ability to review new research proposals and monitor current studies efficiently and adequately. This task is already challenging at current staffing levels.
The device industry was concerned that the PHS data bank could subject clinical investigators to added pressures. First, clinical investigators will need to address inquiries from potential subjects who seek to clarify information in the data bank or request information not publicly provided.
Second, as discussed above, some industry commenters believe the disclosure of device trials will lead to requests for treatment by patients who are not eligible for the study. Such requests often involve more frequent and intensive communications than for protocol patients. Special approval from the study sponsor and IRB must be obtained, and the informed consent document, treatment plan, and follow-up schedule usually need to be tailored to meet individual patient needs. These extra efforts require considerable resources – resources that, according to the device industry, may not be available to most healthcare professionals.
Finally, as noted above, it is important to researchers to be able to conduct research without the concerns associated with public or premature disclosure of research activities. This concern was expressed in response to the Department’s FR notice of February 4, 1999, regarding Public Law 105-277. That law requires Federal agencies to ensure that all data produced under a Federal grant or contract are made available to the public through procedures established under the Freedom of Information Act. Approximately 9,000 comments – many from private researchers – emphasized the importance of being able to conduct research without premature disclosure and outside pressures.
Another concern for the device industry was the possible effect the data bank could have on third-party payers and on reimbursement policies. In 1995, FDA and the Health Care Financing Administration entered into an interagency agreement to allow for the reimbursement of certain investigational devices. FDA, the device industry, and Medicare beneficiaries believe this has resulted in numerous benefits (e.g., improving the quality of clinical trials, assuring that investigational devices are tested on the appropriate patient population, and helping to keep device research from moving overseas). One of the conditions for Medicare reimbursement is that the patient must be enrolled and treated according to the clinical protocol. Manufacturers believed that increased access to investigational devices, both through expansion of the clinical trial and requests for treatment off-protocol, could compromise third-party cost-containment activities.
Some commenters expressed concern that DHHS involvement in a device trials data bank would further blur the distinction between research and treatment. The public could interpret DHHS involvement as endorsement of unapproved devices. Several commenters stated that patients would tend to believe a trial listed in the data bank indicates that the investigational device is promising.
Some commenters argued that DHHS resources should be used in other areas because it has not been demonstrated that there is a public health need for access to device trials information. One suggestion considered the potential benefit of an adverse event database shared by FDA, IRBs, and the industry to far outweigh the marginal benefit the commenter believes would be provided by a device clinical trials data bank. Other comments suggested that DHHS resources would be better used to fund patients who meet the enrollment criteria for a clinical trial but who cannot afford to participate. It was argued that this proposal would not only provide access to needy patients but would also foster the progress of device investigations.
Several commenters noted that if, in fact, there is a need for access to information on device investigations, the level of detail required under §113(a) may not be sufficient to address the real needs of consumers. Commenters noted that it would be shortsighted not to anticipate that the potential subjects would demand additional information, such as the progress of the trial.
One commenter suggested that the data bank should be made available only to qualified health professionals rather than the general public. The rationale was that this would allow professionals who understand the medical implications of the study and the therapeutic needs of a particular patient to determine if the investigational device might offer some benefit to the patient.
Manufacturers had concerns about the potential negative impact of a data bank on the industry and on the conduct of quality clinical trials to support device approval. These concerns apply even to a registry restricted to device trials involving serious or life-threatening diseases or conditions for which no alternative therapies are available.
DHHS believes that Congress needs more information before deciding whether to mandate listings of trials of new devices in a PHS data bank. Although the device industry’s concerns are largely theoretical, they raise issues that may be unique to that industry. Recognizing the differences between the development of new drugs and devices, Congress clearly intended to decide the issue of a device trials registry after evaluating the experience of creating and operating the drugs trials registry. DHHS agrees that considering the experience of the larger pharmaceutical industry with the data bank will help DHHS and Congress evaluate the impact of a requirement to list device clinical trials. Congress should delay any decision about a device trials registry until the year 2001, after there is one year of experience with data from privately funded trials in the drug trials registry.
An evaluation of the device trials registry should consider the concerns raised in this report, including:
- unrealistic expectations about access,
- risks to patients,
- impact on conduct of clinical trials,
- vulnerability of innovative companies if they are required to reveal confidential commercial information, and
- resource implications.
On balance, DHHS believes a data bank could offer access to information to those who might not otherwise have access, giving them a way to consider entering a trial of an experimental device intended to treat their condition. Moreover, the public health is likely to be best served if the data bank is restricted to device trials for serious or life-threatening diseases or conditions for which no other alternative therapies are available. DHHS also believes that a device trials data bank could make a wider audience aware of the existence of clinical trials that have had difficulty achieving full enrollment (e.g., trials for treatments of rare diseases). In these cases, the public health may be served by allowing enough subjects to be identified to permit sponsors to do timely testing on promising products.
1Appendix 1 also contains two requests to extend the comment period. FDA denied both requests in order to meet the statutory timeframe for submission of this report.
Also, FDA announced and held an open public meeting on July 8, 1999, at its Corporate Boulevard offices in Rockville, Maryland. FDA organized this meeting to allow interested parties to present their views and discuss data bank issues with FDA staff and other stakeholders. The meeting attracted 24 consumer, institutional review board, and device industry representatives. A statement from a professional society that represents institutional review boards was also read into the record. (See Appendix 2 for the minutes of the July 8 meeting.)
2 For example: "centerwatch.com"; "oncolink.upenn.edu"; and "drkoop.com."
3 June Gibbs Brown, Institutional Review Boards: A Time for Reform, Department of Health and Human Services, Office of Inspector General; Washington, D.C., June 1998, Page A-1.
4Medical Device Amendments: Report by the Committee on Interstate and Foreign Commerce. February 19, 1976, page 48.
5 Guidance for Industry and FDA Staff on Preparing Notices of Availability of Investigational Medical Devices and for Recruiting Study Subjects, March 19, 1999. FDA's Information Sheets: Guidance for Institutional Review Boards and Clinical Investigators. 1998 Update.