Innovation and Invention in Medical Devices: Workshop Summary (2001)

Presenters in this session described and analyzed the sources and amount of resources available in the medical device innovation field; discussed the role of small, large, and multinational medical companies in medical device innovation; identified the issues and opportunities confronted by innovators in this field; described the role of standards and product applications; discussed the effects of venture capital on this field; and evaluated the role of the legal system in influencing innovation.

Clifford Goodman, Ph.D.

Senior Scientist

The Lewin Group

Sources and timing of support for medical device innovation can be viewed in the context of the medical device life cycle, which can be described in five main streams or pathways of activity: (1) regulation, (2) research and development, (3) manufacturing, (4) marketing, and (5) legal. Medical device innovators and manufacturers increasingly look downstream to the sequence and height of the hurdles along these interrelated pathways to inform their decisions to continue development of a device, modify it, divert resources to other products, or otherwise alter the process of innovation. Continued pursuit of innovation when these hurdles are not aligned or harmonized can require considerable corporate

resources, adding risk to the prospects for innovation. Venture capitalists and other investors consider where an emerging technology is in these streams of activity, and what its prospects are for overcoming these hurdles, when deciding whether to invest in the technology.

The regulatory pathway is generally well defined, with certain parallels between the United States and Europe. Some of the main benchmarks in the United States are design controls; investigational device exemptions (IDE) for devices requiring clinical (human) testing, which requires institutional review board (IRB) approval; premarketing approval (PMA); good manufacturing practices (GMPs); and various forms of medical device reporting once a device is on the market. There are analogs to these hurdles in Europe, for example, the CE mark, which is the market clearance hurdle in much the same manner as the PMA in the United States. Still, the requirements for a CE mark and a U.S. PMA, and the accompanying paperwork, are likely to differ, necessitating resource expenditures for companies seeking international markets for their products.

Although the R&D pathway is anchored by iterative design, preclinical testing, and development of device prototypes, there are other important benchmarks along the way. These include permission to conduct clinical testing (granted by, e.g., the IDE in the regulatory pathway); clinical evaluation (leading to market approval in the marketing pathway); outcomes and health economic research to persuade payers, technology assessment agencies, and others of the worthiness of the technology; and postmarket research and surveillance to gather data about the experience of the device in the field, which helps to fulfill regulatory requirements and provides information for further marketing efforts. One of the challenges in the medical technology industry in the more developed nations is that most new technologies do not result in obvious gains in mortality or morbidity, so that it is important to demonstrate improvements in quality of life and economic advantages. Further, more health care providers and payers want to see evidence of effectiveness in community settings rather than just efficacy in the carefully controlled settings that characterize data gathering for purposes of regulatory approval.

The marketing pathway often starts with market research on user needs, competition, and other factors that can influence device design as well as the regulatory and R&D pathways to be taken by the technology, including what types of health and economic evidence will be required to demonstrate the value of the technology. Education and promotion of the device can begin even before market approval, preparing the target markets and informing those who will be in a position to order and use the device. Sales, distribution, and customer support functions must be in place upon market approval. Third-party payment, which usually depends on market approval by regulators and can be further influenced by other evidence from outcomes research and health economics studies, can strongly mediate sales. Helping device users with third-party payment can be critical. Some companies have established 800 numbers for physicians to call if they are having problems with procedure coding or other aspects of reimbursement for use of the device.

In parallel to the other streams of activity, companies must gear up their manufacturing capacity. This includes facility requirements, maintaining the flow of materials and components, the production process, quality management, and ongoing modifications and retooling of the manufacturing process as needed to cope with device redesign, swings in demand for the device, and other changes.

Legal considerations must be managed throughout the device life cycle, including gaining patents, licensing, maintaining patient protection, and protecting against product liability.

In contrast to the pharmaceutical industry, the medical device industry is characterized by a large number of small entrepreneurial companies and startups. While these companies are largely focused on gaining “proof of concept” and overcoming initial regulatory hurdles, they tend not to have the staff size, experience, and other resources needed to manage these different pathways in the device life cycle. Further, since they tend not to have broad product lines that can sustain their cash flow, their risk profile is more closely tied to the success of one or a few products. As such, the viability of these companies is highly sensitive to changes in regulatory hurdles, payment requirements, sales and distribution, manufacturing capacity, legal challenges, and other factors that can divert their limited resources. These factors tend to influence the points at which small companies are more amenable to being acquired by larger ones or to engaging in other partnerships that will provide the resources needed to manage these requirements.

Current patterns of public- and private-sector funding for R&D, particularly for basic research in the sciences and engineering underlying medical device development, affect the nature and flow of new technology. Aside from traditional funding sources for R&D, changes in payment criteria are providing some earlier revenue streams that can improve the risk outlook for innovation. These are exemplified by various “conditional coverage” arrangements for investigational technology, such as the 1995 Health Care Financing Administration/FDA interagency agreement on reimbursement of investigational medical devices, and greater collaboration of research agencies and health care payers to support clinical trials and other studies of investigational technologies.

John T.Watson, Ph.D.

Acting Deputy Director

National Heart, Lung and Blood Institute

The NIH Revitalization Act of 1993 included language (the so-called Durenburger Amendment) requiring the DHHS Secretary to report on “Support for Bioengineering Research.” The study included an inventory of federal bioengineering support, a non-federal consultant working group, an evaluation of patenting trends for implantable prostheses, an estimate of non-profit and for-profit support for bioengineering research, and an open workshop to access all

findings and make recommendations to NIH. The term “bioengineering” was used in its broadest context, in terms of medicine and biology.

This report led to congressional inquiries about its conclusions. In response, NIH, under the leadership of Director Harold Varmus, formed the NIH Bioengineering Consortium (BECON) in 1997. Simultaneously, many groups increasingly recognized the need for bioengineering support, including the Whitaker Foundation, the American Institute for Medical and Biological Engineering, the Small Business Innovation Research Program, and the NIST Advanced Technology Program. The interests of groups collectively demonstrated the growing importance and awareness of the central role of bioengineering to innovation and invention in medical and biological research and clinical procedures.

The NIH and the Consultant Report recommended some sort of central NIH focus for bioengineering. Support for basic bioengineering research, contrasted to applied and developmental research, was reported as 30% of the total, compared to an NIH average of 60% for all other fields. These reports also addressed the need for an evaluation of the NIH peer-review process for bioengineering research, membership on advisory committees, the movement of new device introduction overseas, the biomaterials availability problem, uncertainties in the innovation process, and using patent information to trace back to related federal research support. Finally, regulation must be meshed with innovation so that entrepreneurs can figure out a way to meet the regulatory guidance in a more cost-effective and shorter time frame.¹

Susan Alpert, M.D., Ph.D.

Director, Center for Food Safety and Applied Nutrition Food and Drug Administration

Regulatory oversight of medical products is an accepted part of the government’s role in providing protection of the public health. At the same time, governmental acceptance of a technology or product contributes to broader acceptance and reliance on the claims of the product’s provider. The threshold that innovations cross to reach the market sets in place an important foundation. This foundation must be established on the basis of good scientific principles and data to have its intended impact—benefit to the public health without undue delay.

The actual interface of the regulatory agencies in the federal government and the innovators in medical device technology and products is broad. There are far-reaching areas of impact, such as those resulting from the development and recognition of technical consensus standards that may be used in design, manufacturing, or regulatory activities. There are specific and more limited areas of impact, such as the individual developer’s meetings with the scientific and regulatory staff of an agency, which focus on details of the information to

be submitted and evaluated for market entry or reimbursement, for example. The impact on device innovation from activities at each of these levels and those in between must be acknowledged and evaluated.

Given the focus on government spending by regulatory agencies for programs seen to be beyond their specific mandate, and which might present conflict of interest, the type of financial support that can be provided to the industry is limited in this sector. There are programs, however, that may be used to broadly enhance the development of new technologies while being responsive to concerns regarding a level playing field among competing companies. In addition, the tasks of the regulatory agencies should, and frequently do, include (1) providing pathways to market that are responsive to the changing timeframes for technology and product development, (2) creating processes that are sufficiently flexible to facilitate novel product development, and (3) incentives for innovators whose product provides a significant contribution to the public health.

There are tensions in place between the need for and the speed of introduction of the new technologies and products. There is a need for better communication between the large companies and the small innovators. Regulation is a necessary obstacle because society demands some type of oversight and accountability. FDA is charged by the public to ensure that device market entry involves products that are both safe and effective. FDA publishes summaries of safety and effectiveness data for devices so that it can be made clear and transparent as to what this product is, and what can be expected from it.

John A.Parrish, M.D.

Center for Integration of Medicine and Innovative Technology

Massachusetts General Hospital

Academia has much to offer the field of medical device innovation, including “problem-rich” and “solution-rich” environments, “molecular” understanding of pathophysiology and mechanisms of therapy, expertise and skills, access to patients, and a culture of scientific methodology.

There are multiple barriers intrinsic to most academic institutions that limit the development of diagnostic and therapeutic devices. There is often a large cultural and psychological gap between the disciplines of biology and engineering that prevents effective dialogue. This results in a lack of clear understanding of clinical problems by the technologist, lack of awareness about technical options by clinicians, and difficulty finding appropriate collaborations. Within the academic medical community, specialization has been a powerful force in learning more about individual diseases and organ systems but has also resulted in turf wars in patient care, destructive competition, and poor communication.

One model for enhancing the role of academia is the Center for Integration of Medicine and Innovative Technology (CIMIT), a collaboration of academic physicians and engineers working with industry and government to solve im-

portant medical problems. The founding institutions can be considered as the problem-rich environments (the teaching hospitals of Harvard Medical School) and the solution-rich environments (Massachusetts Institute of Technology and Charles Draper Labs). The mission is to improve patient care by bringing together technologists, engineers, scientists, and clinicians to catalyze development of innovative technology, emphasizing minimally invasive diagnoses and therapy. CIMIT is led by senior academicians whose full-time commitment is to integrate technology into health care by systematic, non-random purposeful mixing and matching of appropriate clinical champions and engineering experts. The process is intended to:

• identify difficult health care problems amenable to technological solutions,

• encourage teams of clinicians and engineers to generate new solutions,

• provide resources to develop solutions for safer and more efficacious treatments, and

• facilitate the application, transfer and commercialization of CIMIT technology.

CIMIT provides funds to develop and demonstrate new ideas and expertise to guide the development of commercializable products. This expertise includes:

• business development,

• technology development,

• regulation affairs, reimbursement issues,

• patient safety, simulation, and

• industry liaison programs.

Longitudinal programs include development of selected technologies (e.g., devices, tissue engineering, imaging) and focus on selected clinical problems ripe for new technological solutions (e.g., acute stroke management, identification and treatment of vulnerable plaque).

Success is measured by scientific presentations, published papers, patents, and receipt of NIH grants. There is also evidence that CIMIT support for numerous multidisciplinary projects and programs resulted in outcomes that would not have occurred absent that support. Dedicated funding is essential: there is no better way to get people’s attention.

Robert W.Anderson, M.D.

David C.Sabiston Professor and Chair

Department of Surgery

Duke University

Health care incentives have been perverse for many years. There are high prices with overcapacity, common technology is extremely expensive, and we

support mediocrity in clinical care and technology development. Researchers have encouraged the use of health care services because of the fee-for-service mentality, and researchers have failed to promote health and wellness.

There is a tremendous competitiveness in the market. Purchasers have been more sophisticated with an emphasis on cost, often at the expense of technology and innovation. Unlike other industries that blossomed in recent years with high initial start-up costs with a lot of capitalization—and shrinking prices as volume grew—prices only recently shrank in health care.

Shifting sites of care toward more outpatient services has increased. This change has produced technology expenses due to improper use and overuse. Moreover, pharmaceutical spending is on the rise. The factors that are driving change are economics, outcomes and evidence-based medicine, new technology, preventive medicine, and new procedures. Evidence-based and outcomes-based medicine are going to lead to greater accountability and possibly risk-based reimbursement.

New technologies, for example, molecular biology, gene therapy, organ substitution, stem cell biology, and smart devices, will allow researchers to identify high-risk groups of patients and, in many instances, start treatment to prevent onset of disease. New minimally invasive procedures have tremendous potential.

Other factors that researchers need to consider in innovation and health are changing demographics, in particular the aging of the United States population. The number of elderly and disabled Medicare beneficiaries is growing rapidly. In addition, the growth of the uninsured—48.7 million people in the year 2000 in the United States—poses real challenges to the health of the nation.

What can the device industry learn from other industries? First, short-term solutions do not sustain survival. Second, competition creates value. Third, innovation drives continuous quality improvement, and fourth, incentives drive innovation. The problem with determining quality is that no one has adequately defined its parameters. The basic elements for health care change are going to be corrected incentives to improve efficiency, access to relevant information, and sophisticated information systems. As always, any player in the health care market had better be able to demonstrate improved clinical outcomes and cost-effectiveness.

Increasingly, teaching hospitals are the place where complex illness and procedures come together. Teaching hospitals provide care for more severely ill patients. More than half of all major teaching hospitals now have operating margins of less than 0%. As a result, there is a lot of cost cutting going on. The key success factors for academic health centers will be human and fiscal resources. Researchers have to continue to build on their human resources, getting the best people, and training and retaining them. Researchers have to shore up their fiscal resources.

Despite this, academic health centers retain great advantages for clinical trials. Researchers have access to patient populations, and have highly trained personnel. The disadvantages are cultural conflicts, limited capital, and, often, underdeveloped infrastructure. The hindrances to new product development have always been shortage of important new product ideas in certain areas. In addi-

tion, the new product development process is very expensive. Academic institutions must become better at working with industry while maintaining their freedom, and seeking non-traditional revenue sources.

Thomas M.Loarie

Chief Executive Officer

Kera Vision, Inc.

The product Loarie has been involved with for the last 13 years is a device for correcting common vision problems. It is two half rings made of biocompatible polymer that can be inserted into the periphery of the cornea. It stays outside the optical zone and reshapes the cornea so one can get the light rays to fall on the retina. The product can be removed, and in most cases the eye goes back to its original status. This device received FDA approval in 1999. Researchers raised $160 million to bring this to market, making this probably the largest up-front investment in history for medical technology. It is ultimately the public that will be the judge of whether or not researchers are doing their job.

The medical device industry includes 6,000 companies and 3,000 product lines covering 50 clinical specialties. There are only 64 product groups that have revenues over $150 million, and there are only 100 companies that have revenues over $100 million. Seventy-two percent of the medical device firms employ fewer than 50 people. This is really a cottage industry. In global terms, there is great competition. A sustainable advantage by any company can only be attained by leveraging knowledge.

The aging of the population will place demands on this health care system. Technology becomes an important player in helping to solve a dilemma that is before researchers in the very near future. It is the small companies that drive innovation, yet out of 60 ideas, only one product actually makes it to the marketplace, so this is a very fragile process, one that is challenged by the typical mentality of investors. In most areas of the United States, in venture capital there are only two things that are important, feasibility and market acceptance. The regulatory and health care payment environments introduce additional levels of uncertainty. Small companies are agile, with a tremendous tolerance for ambiguity, and are therefore well suited to be the source of innovation for medical devices.

John P.Wareham

Chief Executive Officer

Beckman Coulter, Inc.

Beckman Coulter, Inc. makes products that patients do not see and physicians seldom see—that is, genetic analysis systems, drug discovery enabling systems, and diagnostic systems used by laboratorians. One of the things that

researchers do is make available technologies that can be used to create value in the marketplace. Every medical device company, whether large or small, is focused on bringing real-life patient utility to technology. The unique contribution of a large company is that it can address the development imperative.

Once a technology is invented, bringing it to the point where it can be commercialized requires capital, along with certain competencies and management processes. Large companies are involved in discovery and commercialization processes. The special role of large companies is to contribute infrastructure, market knowledge, and financial resources to validate technologies and make them valuable to patients.

This is a very complex time in history, and a large medical company has the internal resources to help deal with this complexity. For example, in the not too distant future researchers might expect to run a diagnostic test that identifies a gene protein or cell profile that enables a physician to prescribe a course of treatment. While this is the direction of the genomic revolution, policy and regulatory requirements will add layers of complexity to such a capability. These issues, combined with concerns about health care cost management, may drive some innovation into the realm of large companies. Cost management concerns are driving more automation, product standardization, systems integration, and information management, all of which can eventually drive costs up.

It is vitally important for large companies to capture and quantify patient outcomes so that researchers can fairly assess the economic impact of health care technologies. If researchers are lowering costs by preventing disease and more effective monitoring and treatment of disease, then investments in innovation are being made well.

All of these innovations have the potential to control health care costs and improve quality, but there is another factor at play, consumerism. Patients, as consumers of health care, are becoming more informed. They are demanding better information, choices of treatment options, and control. Individuals want to know why tests are ordered, what the results mean, and how they can monitor their own health and prevent disease. As individuals become more involved in their own health care, there will be even a greater demand for information and technology. This drives up costs.

Large medical device companies have the ability to bring scale to the challenge of globalization and successful product development. In a global business environment, the small incubators of technology are particularly challenged with respect to financial pressures. While small companies lack the infrastructure to take on worldwide development and marketing activity, large companies have the infrastructure to help small companies get their technology to patients around the globe. Large companies can supply capital and credibility. In fact, large companies are already playing a key role in the growth of the medical device industry through acquisition, joint ventures, strategic alliances, contract research, licensing, and royalty agreements.

Mergers and acquisitions in supplies, equipment, and devices grew dramatically over the last five years, from $5 billion in 1994 to $32 billion in 1998. De-

spite this, large companies will continue to rely on smaller startups for their ideas and inventions. At Beckman Coulter, this has led to externalizing technology innovation. The objective was to spend half of the budget for technology innovation in collaborative agreements. This may involve a number of parties, including small companies, institutes, academia, and contract research organizations. In this model, researchers use their capital and credibility to build teams that are necessary to assess and pursue these worthy causes, with a net result of bringing utility to patients faster. To sum up, large medical companies have the capital, competency, and management processes to fulfill the development imperative.

J.Casey McGlynn

Partner, Wilson, Sonsini, Goodrich and Rosati

The financial players in health care innovation have changed in recent years. The venture capital community, in particular, has changed over the past 20 years from diversified funds to organizations in which individuals have become more focused on particular technologies. In addition, a number of traditional investors in the health care field have left or disbanded their health care divisions. They turned away from health care because development time in the Internet area is so short and the returns are so staggering. In addition, the lack of returns from the biotechnology industry was discouraging. The percentage of venture capital dollars going into health care approximated a third to a half several years ago. Today it is about 10 to 12%. The medical device sector includes about 208 companies in the United States. In 1999 there were roughly 30 first rounds. These numbers are very small with little room for expansion.

Thus, diversified funds are not as sure a source of capital for entrepreneurs as they were in the past. Fortunately, corporations are more active in the venture world today than they were five years ago. They have become basic supporters of venture capital and one of the major stalwarts of getting companies funded and technology into the marketplace.

Incubators continue to be critically important. They are incredibly valuable to the doctor who might be the ultimate innovator but who lacks the needed resources to build infrastructure and get an idea transformed into reality. The number of incubators is on the rise, a positive development in the medical device industry.

Factors that challenge the device industry in terms of raising capital are long development time, regulation, and uncertainty about reimbursement. In addition, limited liquidity from the public sector limits venture capital’s ability to guess predictably when it will turn a profit on an investment in the medical device industry. One area where initial public offerings have been on the rise in recent years is in genomics, in large part leveraged by the enormous investment made by the federal government in basic research in this field. Nevertheless, there are more venture capitalists interested in the medical device field than in

the biotechnology field, probably because the science is not as complex and therefore the risks are clearer.

The emergence of E-health companies, for example, medical records, has drawn a lot of interest and capital but there are many companies all pursuing the same products. Big companies that are innovative are adopting Internet tools to make themselves more efficient. The small companies will be adopting those same kinds of products to speed up the collection of data, to make that data more precise, and to make regulatory filings quicker and easier.

Tom Fogarty from Stanford began the discussion with a question contained in a story about a clinical trial involving carotid pathology. He was approached about participating in this multisite trial, he said, which upon close inspection turned out to be a brilliant effort to do something real stupid. When Dr. Fogarty called the individual responsible to ask why he was doing this and tell him why the trial really would not work, the individual readily admitted that he was using a new procedure and a new instrument, both of which were probably only 10 percent developed, but insisted that since randomization was what NIH would fund, he was determined to use randomization. Dr. Fogarty declined to participate, but he soon heard from numerous colleagues at other institutions who had not declined, simply because they were going to get paid for it. Here, he concluded, was an early, early stage evolving technology that everybody wanted to document and develop, but they were doing it by the wrong clinical trial method. His question then, for Dr. Watson and NIH, was whether it was not possible to take a parallel path to address the issue.

Dr. Watson agreed that researchers need to think more about those things up front, in some collective way, and get back to what he called guidance sections that work on trial design for this class of devices and give guidance up front. Although he said he is a very strong advocate of randomization from the very first patient, for a variety of reasons, there is an example similar to Fogarty’s with ventricular assist systems, where there have now been several studies conducted. A major meeting has been organized by the American College of Cardiology to look at what has happened and see whether researchers ought to employ different clinical designs.

Susan Alpert from the FDA volunteered that something else very important is involved, and that is that the clinical community is at the table for a lot of these discussions pushing the idea that the technology actually is ready. They are using the technology in ways that may or may not be appropriate and are actually forcing the initiation of the trials. That is, the clinical community is pulling the technology forward rather than waiting for it or working with an individual company. Dr. Alpert opined that it is the duty of everyone in the medi-

cal device community and the clinical community, not just the government, to protect patients. One of the suggestions that has been brought up today, she said, is to do uncontrolled trials to start, to do registry trials as the only trials, but that presents many difficulties. Researchers then never can ethically do the trial that actually answers the question as to how one technology compares to another because by the time the device has been used with 2,000 or 3,000 people under a registry, the same investigators state that it is unethical to do a randomized trial because they already have the answer. We have to clarify how technologies are to be developed, who is at the table as they are being developed, and how to more quickly obtain the small amounts of information that are needed in a focused setting rather than with thousands of people in a registry. Do a focused first cut if that is needed and then get those randomized trials started because they are the ones that are crucial. They are going to answer the question, not just for marketing, but for reimbursement as well.

Kshitij Mohan from Baxter International suggested that the issue with respect to clinical trials is a broader one of what is appropriate science for the validation of technology. For example, if the concern is durability, should that be evaluated on an engineering bench or in clinical trials? Obviously one is more appropriate, depending on the question. If a chip has been already validated to failure rates of 1 in 10¹², little additional information about that chip will be gained through a clinical trial. With that in mind, Dr. Mohan suggested that some outcomes research on the regulatory process itself should be done. He cited the tremendous amount of data available in the 5,000 or so premarket notifications (510(k)s) submitted to FDA each year over the last two decades, the 40 or 50 premarket approval applications (PMAs) submitted each year, and the tens of thousands of medical device reports (MDRs) on failures and malfunctions. Shouldn’t there be some work done, he asked, some systematic research into what validation tools yield the greatest value in terms of demonstration of safety, effectiveness, or economic value with respect to reimbursement?

Dr. Alpert pointed out that the PMA process and the clinical trials process are for unproven technologies, for real innovation where no answers are available, while 510(k)s are for incremental changes to proven technology. It is nevertheless very important to think about which aspects should be measured. Which aspects belong in a clinical trial to establish safety and effectiveness and impact on patients and which aspects of a technology actually are better tested at the bench?

Thomas Loarie, CEO of Kera Vision, offered the view from a smaller company with limited resources. Their cornea-shaping technology is apparently being considered by some surgeons in Europe for use in treating keratoconus, a bulging of the cornea that is estimated to afflict 1 out of 2,000 people and often necessitates a corneal transplant. When a doctor in Europe approached him a few years ago, Loarie told him not to do it, for fear that it might affect their PMA in the queue at FDA. Anything that happens with the product must be re-

ported to the FDA, so some doctor going off and doing something that the product was never designed for could jeopardize the company’s entire investment. Doctors do not usually listen to people who run companies, Loarie continued, and six months to a year later he saw the same doctor in Paris, who reported that he had been treating six patients for keratoconus and had completely stopped the disease progression. He and some colleagues have now organized a physician-type investigational device exemption (IDE) trial to get more data, and Kera Vision has started getting more pressure to run formal trials both in Europe and in the United States. The problem is that they cannot afford it. It may be a great breakthrough for keratoconus, but the company just cannot afford to run a trial on a such a rare disease as keratoconus.

Dr. Alpert pointed out that there is nothing wrong with European data. If the studies are studies conducted appropriately, the data are perfectly acceptable to the FDA. However, she agreed that the cost of clinical trials in this country is an extremely important topic, especially the cost of overhead from the major academic institutions. There are some new tools, alternatives to clinical trials to support these things, she asserted, not perfect tools, but there are more tools. Researchers need to allow development of technology, she continued, but under the right controls that protect patients, and if they do not have all the tools, then they ought to be developing them.

Jim Benson pointed out that there are indeed such tools, humanitarian exemptions, treatment with investigational new drugs (INDs), IDEs, and postmarket coverage as opposed to premarket data, but what is often missing is knowledge on the part of a company that has an issue like this that they can, in fact, have those discussions

Robert Califf of Duke University promised to address academic overhead and costs of clinical trials in his afternoon talk, because, as he put it, there is no shortage of innovative ideas. It is the funding to actually get the necessary data that is in short supply. As a clinical trialist responding to Dr. Fogarty, Dr. Califf claimed that a lot of devices have gone down the tubes because of well-intentioned inventors who did not know the basic fundamentals of clinical research. He pointed to gene therapy, where once again very intelligent people just did not know the fundamentals of what one must do when doing a human experiment.

Dr. Califf also raised an ethical question regarding overseas clinical trials. What one calls clinical trials or device development, he said, is a human experiment. The idea of saying it is too hard to do in the United States, so researchers will do the experiments on human beings in Europe, raises a lot of issues that really need thought by many people. Europeans are no more expendable in terms of experimentation than Americans.

Califf went on to say that anyone who has worked with devices or done research with devices knows that the fundamental questions are who should be allowed to tinker, and how far can one go with tinkering before it is a new experiment that demands informed consent and reporting of what one is doing.

Researchers have evolved from the 1970s, when any doctor who wanted to fiddle with something could pretty much do it, to a year 2000 in which if a doctor changes the size of a urological catheter in a standard procedure for more than three or four people and writes a case series in a journal, he risks censure for doing human subjects research without the approval of an Institutional Review Board.

Mr. Loarie responded by noting that his company had started its investigational work in Brazil and done further research in Mexico and in Europe, and he had yet to meet a doctor in any of those countries who was unconcerned about harm to a patient. Integrity of clinical research is not unique to the United States. These doctors have practices. They have reputations and they demand a lot from Loarie’s company before they will do anything.

Dr. Alpert brought the discussion to a close by noting that FDA actually addressed the issue of tinkering during clinical trials. Devices are allowed to evolve during clinical trials, despite the implication yesterday that nothing can be changed. During an IDE, a product can, in fact, evolve appropriately as long as impact of the changes on the data is taken into consideration, that is, the ability to pool data to understand what that technology or that technological change or the tinkering or the modification of manufacturing has accomplished.

Obviously, communication is very important, but there is opportunity for products to evolve during clinical trials and not be absolutely frozen. There are formulation issues in dealing with drugs, and there are changes in device design that can, in fact, render the data non-poolable. That must be considered as one goes forward but one can, in fact, evolve.