Originally published July, 1996
An Interview with Gordon Higson
Consulting Director of MTC-BRI International, Staines, Middlesex, UK,
and Chairman of ISO TC 210
In the decade ahead, Gordon Higson envisions a globally harmonized system for regulating medical devices, one based on a comprehensive, coherent set of standards. Standing in the way of such global harmony are the existing regulatory systems throughout the world. But increasingly there are signs that these systems are showing flexibility--even in the United States, which constitutes the largest medical device market in the world and whose regulatory system might be expected to have the greatest inertia. There is serious talk of reform on Capitol Hill and in the trenches at FDA, and proponents of the European model--which emphasizes safety while leaving clinical practitioners to decide on efficacy--are being heard.
For years, Higson has supported the development of global standards. As chairman of the International Organization for Standardization's (ISO) technical committee that is working to apply quality systems standards to the design and manufacture of medical devices--ISO TC 210--he is involved in decision making that will affect the development of worldwide standards pertaining to the medical device industry. In this interview with MD&DI, Higson discusses the differences between the European and U.S. regulatory systems, the potential for harmonization between the two, and the challenges that lie ahead for efforts such as those of ISO TC 210.
How does the European regulatory system differ from the one used by FDA?
When the Europeans were developing their regulatory system in the 1980s, they learned a great deal from the U.S. system and tried to overcome some of the difficulties they found. In particular, the European system is based on a very clear definition of the approval requirements for devices. Known as essential requirements, these 40-odd items are written in general terms, and they cover all the known sources of danger to patients. The European regulation (directive) then goes on to use the principle of reference to standards. It specifically states that technical elaboration of the essential requirements for individual types of devices is to be contained in voluntary standards.
Those standards are written by the voluntary standards organizations: the European Committee for Standardization (CEN) and the European Committee for Electrotechnical Standardization (CENELEC). They are empowered to write standards on behalf of the member states of the European Union (EU), but mostly draw on standards produced by the international standards bodies. A manufacturer who complies with standards that have been recognized by these organizations is deemed to have satisfied its legal requirements. I think this is quite an advanced approach to lawmaking, and by defining the requirements in general terms the law itself is likely to remain stable for many years. And that's a good point, because it's very difficult to change laws. If we were to put technical detail into laws, we would soon find ourselves stuck with obsolete requirements. But by allowing technical elaboration to take place in the standards field, this system frees the requirements from the bureaucratic burdens of the legal system and enables them to draw on expertise that's available anywhere in the world rather than just in an individual European state.
How do these documents work from the manufacturer's point of view?
The essential requirements are written into the medical device directives. Technical details are written into standards. And all
of these documents are available to everybody--manufacturer and regulator alike. So the manufacturer has every opportunity to determine that its device is safe and satisfactory before submitting it to the regulatory process for approval. This helps things enormously. As far as I'm aware, it's actually quite difficult for a manufacturer in the United States to know that its device is approvable before it's submitted to FDA. The manufacturer submits the application and responds to questions from FDA. And that seems to me to be an inherently slow process, because the manufacturer can't quite know what the questions are beforehand. In the European system, the requirements are clearly defined in advance, so manufacturers can do the work necessary to gather supporting evidence. Indeed, manufacturers are expected to get supporting evidence of the safety of their devices before moving them into the regulatory process.
In recent years, FDA has become increasingly interested in having device companies develop clinical data about the safety and effectiveness of their products. Is there such an emphasis in Europe?
No. The European approach requires clinical evidence only when the safety of a device cannot be established in the laboratory. But for many devices it is possible to do so, and therefore no clinical data are required.
Even when such data are needed, in Europe the requirement is to show clinical evidence only for safety and performance--not for effectiveness. There is no efficacy requirement in the European law. Our approach is to determine whether the technology performs in accordance with its labeling; we leave it to the medical profession to decide whether one diagnostic or therapeutic methodology is preferable to another for an individual patient.
How does the European system of device classification relate to the requirements for clinical testing?
Most of what the Europeans learned from the United States was, of course, related to the classification system. The Europeans have followed many of the U.S. approaches to classification and the use of quality systems. In the European system, clinical evidence of safety is required for devices in Class III (the highest class), for implants in Class IIb, and, of course, for any device representing a new kind of technology. Such clinical evidence must examine all possible dangers, including the potential for side effects resulting from use of the device. The manufacturer must also carry out a risk-benefit analysis to make sure that any risks or side effects presented by the device are actually worth accepting because of its benefits. So an efficacy element can be introduced in cases such as this.
Are there Class IIa devices that don't present much danger and, therefore, would not necessarily have to go through much testing?
That's correct. They have to go through
an approval process, but it probably would not involve clinical studies.
Class I devices are the simplest?
They are very simple, and can be put onto the market with only the manufac-turer's declaration that they conform to the essential requirements of the directive and that there is some documentary proof of compliance, in case the authorities wish to examine it. But normally the manufacturer doesn't have to go to any approval authority to put a Class I device on the market.
So European regulatory agencies are mostly concerned with protecting the public health and prefer to let the market decide what devices should be used?
I think we would use the term professions instead of market. We think it's for the clinical professions to decide what devices are used, and when. Every patient is different, so generalizations about the effectiveness of a particular device or particular condition are quite hard to make.
You said earlier that European regulators looked at the U.S. system as a model upon which to base their own. What can FDA learn from the Europeans?
There are quite good relations between the authorities in Europe and at FDA. They meet quite regularly, and I'm sure that kind of learning process must go on all the time. For example, the United States has had its GMP regulation in place for a long time, and for many of those years the equivalent European programs operated on a much smaller scale. The wholesale use of quality systems standards as part of the regulatory process was adopted by the Europeans from the United States--but with a difference. In the United States, the GMP regulation has always been considered a follow-up to device approval, but the Europeans brought quality systems in as part of the premarket process. Now, we're beginning to see the United States leaning toward the European practice. I believe it's now the case that
premarket approvals won't be processed unless the manufacturer demonstrates it has an adequate quality system. I think this use of quality systems as a premarket requirement rather than as a postmarket check-up system is significant.
The quality systems now in place are built mostly around the ISO 9000 family of quality system standards. Can we expect any changes in these standards in the future?
The international standards community is already talking about a new version, which will come out perhaps around the year 2000 and will be considerably different from the current version of ISO 9001. The revision is expected to consider total quality management and the efficiency of the manufacturing company. Members of TC 210 have noted that such revisions might take the standard beyond its current use in medical device regulations. And our committee has sent a message to TC 176, the committee that is working on this revised standard, to request that it bear in mind that such changes could seriously compromise the usefulness of the standard as a regulatory tool, not only in the case of medical devices but also in many other product areas. We asked TC 176 to consider whether the present versions of the ISO 9000 standards could remain active in some way. To accomplish this, we may have to preserve the present texts of ISO 9001 and 9002 in some other form, perhaps as medical devicespecific quality standards.
Many U.S. device companies have been certified to an ISO 9000 standard, with the idea that this would also bring them close to conformity with FDA's forthcoming revised GMP regulation.
Yes, of course. It's my understanding that the forthcoming GMP revision is modeled very closely on ISO 9001--as are the European, Japanese, Australian, and forthcoming Canadian systems. For those reasons, I think it will be essential for this standard to be preserved in one form or another. It will be interesting to see how TC 176 responds to our request, because we in TC 210 work very closely with them. But it may end up as a task for TC 210 to preserve the present requirements of ISO 9001 in some form or another.
What would happen if the emphasis of the ISO 9000 standards was changed
I think what we would do then would be to provide a mechanism so that the regulatory systems that work by reference to standards would still have standards available to refer to. They have the 1994 version now, and TC 176 could agree to preserve that version in some way even after the revision of 2000 is completed. But if TC 176 decides to supersede it with a new version, I think the current version of ISO 9001 will still have
to be preserved in some way by TC 210--perhaps as a specific quality standard for medical devices.