Author + information
- Received April 15, 2005
- Revision received August 9, 2005
- Accepted September 8, 2005
- Published online April 18, 2006.
- Richard L. Popp, MD, MACC⁎,1,
- Beverly H. Lorell, MD, FACC†,2,
- Gregg W. Stone, MD, FACC‡,3,
- Warren Laskey, MD, FACC§,4,
- John J. Smith, MD, JD∥,5 and
- Aaron V. Kaplan, MD, FACC¶,6,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Aaron V. Kaplan, Cardiology Section, Dartmouth-Hitchcock Medical Center, One Medical Center Drive, Lebanon, New Hampshire 03756-0001.
Public registration of clinical trials is fundamentally important to the integrity of the medical device development process. In addition to fulfilling obligations to those study volunteers, a complete record of trial results provides the general public, clinical community, and medical device manufacturers with a more accurate understanding as to how a specific therapeutic should be used. Although the issues associated with public disclosure of clinical trials are similar to the pharmaceutical industries, the iterative nature of device development introduces differences in what type of information needs to be disclosed during development and commercialization. The Second Dartmouth Device Development Symposium (3D2) held in October 2004 brought together thought leaders representing many of the stakeholders associated with medical device development. This consensus document arising from the proceedings of the 3D2 is offered to provide background to these issues and recommend pathways to implementation of device trial registration.
Recent reports regarding the suppression of data from clinical trials that identified a relationship between serotonin-reuptake inhibitor antidepressants and suicide in children as well as the connection between cardiovascular events in patients taking cyclooxygenase-2 inhibitors has led to broad public debate regarding the importance of public registration of clinical trials (1–3). This has led to congressional hearings, editorials in leading medical journals, and position papers from leading medical societies (4–6). The focus of these discussions has been primarily on pharmaceuticals, with little attention on the role of public registration of clinical trials involving medical devices. The Second Dartmouth Device Development Symposium (3D2), held in October 2004, brought together thought leaders representing many of the stakeholders associated with medical device development (DD). This consensus document arising from the proceedings of the 3D2 is offered to provide background to these issues and to recommend pathways to the implementation of device trial registration.
All those participating in clinical trials seek to advance medical knowledge to allow us to treat patients more effectively. Clinical trials are indispensable for the development of new device-based therapies because animal models usually do not adequately replicate human anatomy, physiology, and pathology. Although human research is necessary, subjects participating in such research often derive no direct personal benefit, and in fact, negative effects of experimental therapies are known to occur. It is assumed that the benefit to society balances the risk for an individual participating in a clinical trial. The making public of data from clinical trials is fundamental to this assumption.
The results of these studies are of deep interest to the general public (including subjects and patients), the clinical community (including clinicians, investigators, professional organizations, journal editors, and conference organizers), industry sponsors (including their shareholders), private payers and government (including the U.S. Food and Drug Administration and the Centers for Medicare and Medicaid Services), and the financial community (including venture capitalists and public equity markets). Therefore, it is important that the results of clinical studies be made public.
The results of clinical studies may remain unpublished for many reasons, which include intentional withholding of data by sponsors to shield proprietary interests. Furthermore, publication bias seems to impede the ability to publish results of a “negative” trial in a quality peer-reviewed journal.
Failure to make public the results of clinical trials has many implications, which include:
• Violates the compact with study volunteers and undermines the public trust that is crucial to the ability to perform clinical trials going forward
• May lead to repeat studies needlessly subjecting new patients to risk and expending scarce resources
• Distorts the knowledge base on which medical care is determined
When devising the proper mechanism for disclosing clinical trials, one needs to acknowledge the difficulties associated with presentation of raw data that are frequently complex, contradictory, and difficult to understand. Furthermore, such a system needs to be implemented without usurping the fundamental role of the peer review process before formal publication.
Discussion: Reporting of medical device studies
The issues surrounding the role of public registration of clinical trials in DD were the focus of presentations followed by roundtable discussion at the 3D2. There was complete consensus that both the safety of human subjects and informed participation (consent) must be the overriding ethical concerns of all stages of human device experimentation. The participants also debated the distinctive features of early exploratory studies versus later clinical trials, as well as the similarities and differences between the clinical trial processes of pharmaceutical and device therapies. Discussion focused on which clinical studies should be subject to public registration and what type of data should be disclosed.
There was also consensus among the participants that the type of information disclosed should be specific to the stage of DD being studied (Table 1).However, there was not complete agreement regarding the extent of data that should be disclosed or the appropriate timeframe for its disclosure. In this context, medical device innovation and development are frequently characterized by four distinctive stages of investigation:
1. Early exploratory human studies examining the feasibility and safety of a device, typically with secondary efficacy end points (“first in man” investigation and limited multi-center studies)
2. Hypothesis-driven clinical trials (usually randomized) testing the final device prototype before regulatory approval, with primary efficacy end points (“pivotal” studies)
3. Post-approval studies and trials that examine new indications or patient populations differing from the original device labeling
4. Post-approval studies within the device labeling that examine marketing preferences and behaviors
Early “first-in-man” feasibility studies are initiated after extensive bench and animal studies, and are often carried out in patients for whom the device may ultimately be intended. Such studies typically examine safety using device prototypes, the results of which often stimulate iterations in device design as well as procedural technique. This early stage of medical device investigation is distinct from early-stage pharmaceutical studies (phase I studies), in which safety and pharmacokinetic features of novel drugs are initially tested in normal human volunteers. There was consensus among the symposium participants that reporting of safety observations is the overriding issue of public disclosure during early phases of DD. There was disagreement regarding the extent to which proprietary design features and preliminary measures of efficacy must be disclosed and introduced into the public domain during phase I studies.
In contrast to early first-in-man studies, later-stage randomized clinical trials are typically performed to establish the body of evidence required to support regulatory approval, reimbursement decisions, and widespread adoption, all of which are essential for successful device commercialization. Such pivotal clinical trials are usually characterized by use of refined devices, testing of predefined hypotheses regarding primary and secondary end points of safety and efficacy, pre-specified plans for statistical analysis, and comparison of the novel device intervention with a comparator therapy (placebo or active control). Widespread agreement was voiced that the results of all pivotal clinical trials should be available in a public registry to facilitate widespread dissemination and scrutiny.
After approval, when devices are in the marketplace, registry studies and/or randomized trials are often performed to formally examine new indications for use, modes of application, or safety and efficacy in populations that differ from the initial regulatory labeling of the device. The results of such studies should be disclosed through national trial registries in a manner similar to that for pivotal trials. However, the symposium participants agreed that such studies should be distinguished from post-approval observational studies, in which the device is used within approved labeling for the exploration of market preferences relative to competitors.
There was consensus that disclosure of all safety findings is essential for all clinical device human experimentation, including those within the four major categories described earlier. The goal of disclosing the safety results of such studies is critical not only for the study subjects, but also for subsequent decisions by institutional review boards, physicians, and patients regarding participation in ongoing and future studies as well as for the general use of a specific device.
We propose that when the first patient is enrolled in any clinical study or trial, there should be an announcement of the existence and intention of the study, as well as subsequent reporting of clinical experiences with emphasis on disclosure of major life-threatening or life-altering events. The suggested details of such announcements are outlined in Table 1.
Later-stage or pivotal clinical trials performed before regulatory approval, or after approval outside of device labeling, are done to assess both the safety and the efficacy of devices, and those results will facilitate choice of devices and decisions by physicians and patients regarding subsequent therapy. The critical feature here is publication in a peer-reviewed form of the results of the study within a reasonable period of time. We suggest that such disclosure be made in every instance and in a timely manner to best protect subjects and appropriately impact the activity of other clinical trials. The development of reasonable guidelines for appropriate timing of the reporting of safety and efficacy results is complex because accurate reporting depends on both complete ascertainment of outcomes and appropriate statistical analysis. Post-approval studies done within the context of labeling of the device for the purpose of assessing market preferences for a specific device or design feature were also debated by the group. Distinct from public reporting of safety outcomes, there was no consensus regarding the reporting of other results that relate to positioning of the device therapy within the competitive market environment. Although such data are very interesting to most stakeholders, their primary use in competitive marketing was seen as a potential problem. The participants did not fully discuss issues regarding the acquisition and public reporting of actuarial device outcome data acquired in humans (including claims data), which could potentially relate to both safety and efficacy.
Additional important issues regarding public disclosure of medical device trial results, as well as registry of active trial enrollment, include the appropriate site(s) for repository of the data, the sources of fair and reasonable funding and management of such databases, and harmonization of public reporting of clinical trial results acquired in and outside of the U.S. The mechanisms being implemented for registration of pharmaceutical trials should be an appropriate venue for disclosure regarding studies involving medical devices.
Public registration of clinical trials is fundamentally important to the integrity of the medical device process. In addition to fulfilling obligations to study volunteers, a complete record of trial results provides the general public, clinical community, and medical device manufacturers with a more accurate understanding regarding how a specific therapeutic device should be used. Although the issues associated with public disclosure of clinical trials evaluating medical devices are similar to pharmaceuticals, the iterative nature of DD introduces differences in what type of information needs to be disclosed during development and commercialization. It is important to emphasize that public disclosure should not be viewed as a substitute for peer-review publications, which provide a well-developed means of ensuring that data are presented in a rigorous and thorough fashion.
The Participants in the Public Registration Trial Discussion Group at the Second Dartmouth Device Development Symposium, Woodstock, Vermont, on October 14 and 15, 2004, included: William Braun (Magenta Medical Inc.), Warren Laskey, MD (University of New Mexico), Beverly Lorell, MD (Guidant Corp.), Richard L. Popp, MD (Stanford University), Steven Slaughter (UBS Global Asset Management), John Somberg, MD (Rush University), John J. Smith, MD, JD (Hogan and Hartson, LLP), Gregg Stone, MD (Columbia University/Cardiology Research Foundation), and Mirjam van Werven (Cordis Corp.). The authors wish to acknowledge and thank Denise Colety for help in preparing this manuscript.
↵1 Dr. Popp is a venture partner at Advance Technology Ventures, a leading venture capital firm.
↵2 Dr. Lorell is a corporate officer of a large medical device manufacturer.
↵3 Dr. Stone is a consultant to many venture-backed start-up companies.
↵4 Dr. Laskey consults to the United States Food and Drug Administration.
↵5 Dr. Smith is a partner at Hogan and Hartson, which has many medical device manufacturers as clients.
↵6 Dr. Kaplan is a director and holds significant equity positions in several medical device start-ups.
This paper is a consensus document that reflects the opinions of the working group and does not necessarily reflect the official policy of the institutions where the authors are employed.
This paper is the result in part of sessions at the Second Dartmouth Device Development Symposium (3D2) held at the Woodstock Inn, Woodstock, Vermont, October 14 and 15, 2004.
- Abbreviations and Acronyms
- Second Dartmouth Device Development Symposium
- device development
- Received April 15, 2005.
- Revision received August 9, 2005.
- Accepted September 8, 2005.
- American College of Cardiology Foundation