Medical Device & Diagnostic Industry Magazine | MDDI Article Index

An MD&DI February 1997 Column

For more than three years, Kim Trautman has been at the center of FDA's efforts to revise the medical device good manufacturing practices (GMP) regulation. Now issued as a quality system regulation that incorporates GMP requirements, the new regulation will take effect on June 1, 1997. But getting the regulation published was far from the final step for Trautman, who has also been intimately involved in producing educational materials to get industry ready for the transition.

Now the burning question is, "How is industry responding?" In this interview with Steve Halasey, editor of IVD Technology, Trautman evaluates industry's progress toward understanding the requirements of the new regulation and explains FDA's next steps toward implementing it. From Trautman's description, it appears there's still a lot of work to do on both sides.

Q. What concerns has the agency heard from device manufacturers since the quality system regulation was published in final form in the Federal Register last October?

A. In general, companies are asking the agency to be more prescriptive. That seems a little odd to me, since industry comments on the drafts issued over the past three years have urged us not to be prescriptive. The new regulation leaves it up to manufacturers to determine, develop, and document their procedures. They're responsible for telling the agency how they're going to implement the regulation, and for justifying the way they choose to do so.

Q. Are companies suggesting that they want additional or more-prescriptive guidance documents?

A. Well, they just want the answers. They want to know how they're supposed to implement the regulation. But there are so many different ways of doing it that the agency really didn't want to be prescriptive. Companies may not like it, but the answer is that they will have to decide how to implement the regulation. They have to think about it with respect to the size of their own organizations and the risk levels of their products, and then devise an appropriate program.

Q. How will the new regulation change the ways that companies think about quality?

A. It's going to force manufacturers to integrate quality into every part of their process, and to think of quality more as a system than as a discrete function. They'll have to understand the relationships among corrective and preventive actions and the design, R&D, manufacturing, and management functions of their companies.

No matter how a company chooses to implement the regulation, it's going to force them to establish better communications among functional areas. QA professionals aren't going to be isolated anymore; they will have to interface with management, with R&D professionals, and so on.

Q. Are managers and others using FDA's educational materials as much as you'd like?

A. I'm still seeing the same types of people at conferences. Managers know the regulation is something new, and they're sending their people to conferences, but I don't think managers have begun to go to the conferences themselves yet.

Hopefully, once the RA and QA people get educated they will convey to their bosses what management is responsible for, and then maybe the bosses will start wondering what they have to know. We have to let the RA and QA people have enough time to learn the new regulation and give feedback to their management.

All of this should help RA and QA personnel to implement the regulation. Just like anywhere else, when you have command emphasis from your boss it gives you a much stronger leg to stand on.

Q. Overall, is the phase-in period that lasts until June 1 going as smoothly as you had hoped?

A. It's going smoothly, but I'm still very disappointed at how few people have read the preamble to the regulation. The preamble is an important statement because it expresses the intent of the regulation and provides an administrative record of its development. The preamble can also be used in a court of law to establish the agency's intent, and that gives it legal standing higher than that of a guidance document. More practically, the preamble helps manufacturers understand the complications that have arisen over the years, and why the agency decided on the rules that are included in the regulation.

By reading the examples and guidance included in the preamble, companies can come to a much better understanding of the regulation as a whole. But I can tell from the questions I receive at conferences that some companies haven't done this homework yet. For example, attendees ask about the use of prototypes, which is covered by several paragraphs in the preamble [comment 80], or about the distinction between critical and noncritical devices, which was eliminated in the first proposed revision of November 1993 [comment 4]. These questioners clearly haven't read the preamble or the previous proposed versions.

To me, this is the single greatest source of frustration. The agency has made a lot of educational materials available, but the single most important source of information is the preamble. If I could get across one point to manufacturers, it would be to read that document.

Q. Are there particular sectors of industry--IVD manufacturing, for instance--where additional guidance from the agency might be necessary?

A. Not really. The quality system regulation is very broad and has been designed to cover a wide range of products. There are many kinds of medical device technologies, each with its own risks. Manufacturers know their own products and the risks they carry. Each manufacturer has to develop a quality system program commensurate with the products it's developing, keeping in mind the size of the company.

The questions that a manufacturer has about the regulation will tell it what needs to be justified, and that will lead the manufacturer to the point of considering what type of quality system to establish and how to justify that system. Companies have to adjust their procedures to fit their own cultures, but for most companies simpler is probably better.

But companies have to find a middle ground. Their procedures shouldn't be

so strict that they offer the company only one way to satisfy the requirements, nor should they be so loose that the company doesn't derive any benefit. If a company's procedures don't make good business sense, then something's wrong with the way it's approaching the regulation.

Q. What other pitfalls should companies avoid in devising their quality systems?

A. The regulation gives companies a lot of flexibility and frequently uses the phrase where appropriate. The worst thing a company can do is to adopt that phrase into its own procedures. The agency expects that the company's procedures will define the circumstances under which it will perform a certain test or task.

Using the phrase where appropriate in a procedure is probably one of the major red flags that I will be talking to FDA investigators about, because it means that company management has not provided its personnel with guidance. What does this phrase mean to manufacturing personnel? Will they know what to do when they encounter it? By giving manufacturers flexibility, the agency intends that they should use it, not just pass the buck.

Q. How can device manufacturers avoid this trap?

A. When the regulation uses the phrase where appropriate, the manufacturer might create a procedure that explains when the requirement is appropriate and presents its justification for that decision. But the situation may not always be that clear-cut. So the manufacturer might choose instead to include a decision tree or rationale that can be used to determine when the procedure is appropriate. Both approaches are acceptable.

Q. What about the use of electronic systems and automated tools for implementing the regulation? Have you seen an increase in their use by manufacturers?

A. Over the past four years there has definitely been an increase of activity in this area. A lot of companies are developing software for calibration, for documentation, for compiling the device master record or device history file, for complaint tracking, and so on. Each company has to decide which tools are appropriate for it, for its products, and for its environment.

The agency will avoid trying to determine which of these programs is better than another. All we're concerned about is if a manufacturer does use an off-the-shelf software package that it meets the requirements under 21 CFR 820.70(i), which requires that software be validated for its intended use.

Q. How will the agency be carrying out inspections for the new design control requirements?

A. A draft of the strategy was placed on the device center's web site in late December. We have tried to get as much comment as possible so we can be sure what we're adopting will meet with industry expectations. The tricky thing was that we needed to have industry comments by the end of January so we could compile them for the March 12 teleconference on design control inspections.

We're now working on finalizing the document so we can go ahead and teach both industry and agency personnel. The teleconference and regional meetings that will occur in April and May will address those topics--design controls and how we're going to inspect design controls. So when June 1 arrives, companies should not be surprised about what questions investigators will be asking.

Q. Is there a danger that the strategy document will make investigators fall into a checklist mentality?

A. That danger was on our minds, and that's why we didn't go to a pure checklist approach. Instead the strategy uses more of a leading-question approach. In fact, we didn't allow questions that would have simple yes or no answers. When it's all finished, the document will include the text of the requirements, the questions that investigators are supposed to ask, and blank lines for the investigator to fill in. The final report will act like a combined FDA-483 and establishment inspection report for design control.

The strategy is going to structure what investigators ask, but not so much that all they have to do is check yes or no. Both industry and investigators are going to have to do some work to complete the questionnaire. That way both groups will learn, and that's what we want to happen during the transition year--learning and understanding.

Q. What is your sense of how well prepared companies are for meeting design control requirements?

A. We're not sure. A lot of companies have been working toward ISO 9000 certification, and now they're adapting their systems in order to be ready by June 1. On the other hand, some start-up companies seem mesmerized and awestruck by the whole thing. That scares me.

At this point, since we're not yet inspecting, I have no idea which of these extremes is most typical. Unfortunately, most conference attendees are from companies that can afford to send their people, and they're probably a little bit better prepared than the companies we're not hearing from.

Copyright © 1997 Medical Device & Diagnostic Industry