FDA Aims to Eliminate the "Pre-1976" Stigma

Originally Published MDDI January 2003

WASHINGTON WRAP-UP

A new CDRH office will be the showcase for director David Feigal's concept of the total product life cycle.

James G. Dickinson

Risk Management Comes to Devices | User Fee Implementation Delayed | Head of Bioresearch Monitoring Leaves FDA

One of the oddities of U.S. medical device law is that most new products are cleared for marketing by being deemed equivalent to devices marketed before 1976. Short of going to Congress to revise the law, how can FDA avoid the perception that it is stuck with a backward regulatory system? Device center director David Feigal answered this question in November by opening the Office of In Vitro Diagnostic Device Evaluation and Safety. The center's newest office will serve as a pilot of Feigal's total product life cycle (TPLC) concept, which he has been describing at every opportunity for the past two years. The director of the new office, Steven Gutman, hopes it will free U.S. devices from the pre-1976 perceptional straitjacket.

Gutman told a Regulatory Affairs Professionals Society in vitro diagnostics conference in San Francisco in November that the TPLC concept is the "entire fuel" of the new office. TPLC, he said, will be shown to "maximize efficiency" by providing "seamless oversight" and a "cradle-to-grave" approach to regulation.

While CDRH has long been familiar with TPLC, the new office will demonstrate it to the world. Inside the center, Gutman said, "the mantra is that this stuff is everybody's business. Everybody is supposed to be working off the same page."

Gutman noted that, to someone "who has spent 10 years in premarket science," the idea of TPLC "has tremendous intellectual appeal." The 1976 and subsequent laws governing premarket review, he said, are "starting to show their age." Though treating products as equivalent to those marketed in 1976 was tenable in 1986, he said, "in 2002, as a regulatory scientist, it's not a comfortable place to be."

The problem is most pronounced in the international scientific circles where device standards are discussed, Gutman said. Other countries' delegates, he said, look at U.S. representatives "as if we're aliens from another planet when we talk about seeing new products as equivalent to 1976."

Everyone recognizes, he said, that products "behave differently" after they have been introduced into real-world use. The TPLC approach not only gets away from the stigma of equivalence to 1976 standards, but recognizes this "different behavior."

It's no secret, Gutman said, that the "snapshot" of premarket data that reviewers see doesn't always reflect the actual performance of the product after it has been "stressed" by "dozens or thousands of labs and perhaps tens of thousands of users."

CDRH chose IVDs to pilot the pan-regulatory TPLC approach because these devices have "very stereotyped review issues," Gutman said. "Whether it's premarket, postmarket, compliance, or adverse reporting, the questions are always the same. How accurate is the device? How precise is the device? What is the sensitivity of the device? What is the specificity of the device? Always, these are the same four core questions. And across the center there is a cadre of like-minded scientists who are interested in diagnostic outcomes as opposed to therapeutic outcomes."

Gutman told his RAPS audience that "in spite of the flat economy, we have an amazing emerging technology positioned to enter the marketplace with really wonderful new tests to profoundly impact public health. And in our [in vitro diagnostics] area, we are uniquely multitasked. We already do both FDA and CLIA [Clinical Laboratory Improvement Amendments] programming, and I think that we have the technical savvy to globalize our perspective."

Risk Management Comes to Devices

As it has in other FDA-regulated areas, risk management for medical devices is becoming a major focus of the agency. In the world of devices, however, the term has a handicap: it isn't defined in FDA's 1997 quality system regulation. Instead, the agency must extrapolate it from the restrictive term that is found in the regulation: risk analysis.

In a November presentation to the Regulatory Affairs Professionals Society in San Francisco, FDA's Kimberly Trautman said, "Risk analysis is not a proper term. We adopted this term due to international harmonization . . . ." However, the medical device quality systems expert added, the term as used in Europe has much more to do with risk management than plain risk analysis. When it was adopted in the U.S. device design control requirements, she said, "we were not as sophisticated as we should have been."

Risk management in its broadest sense is critical to all aspects of regulatory compliance, including medical device design validation, Trautman said. "Under risk management, we do risk identification as well as risk analysis and risk reduction," she said. "If FDA finds that you have identified an unacceptable risk but you have taken no action to reduce it, we will cite that as a deficiency."

Trautman reminded her audience that the preamble to FDA's quality system regulation emphasizes that manufacturers must take scientifically documented actions that are commensurate to the risk of their products. In addition, she said, ISO Standard 14971 on risk management is becoming increasingly important to FDA.

"Without a formal risk management program, you cannot comply with the intent of the regulations," Trautman warned. "You will be hearing a lot more about this," she said, adding that the agency will increasingly look at how a risk management program is integrated into a company's quality management system.

Trautman's presentation ranged over FDA's experience with design controls, and she repeatedly emphasized the importance of correct terminology. "Too many times," she said, "I hear people talk about 'product validation.' There is no such thing as product validation. You're either validating the design, which is the complete device—labeling, any accessories, etc.—or you are doing process validation to ensure that you are consistently meeting a predetermined set of specifications or requirements."

Trautman also stressed the importance of controlling design changes "through the life of the product." The longer you wait to do so, she said, the more expensive it becomes, especially when it results in product recalls.

Asked whether FDA has hard evidence that effective design controls reduce recalls, Trautman answered tentatively. Although inadequacies in FDA's databases have prevented analysis, she said, the agency does have anecdotal evidence from individual companies: "They have been able to show . . . an almost linear decrease in the amount of recalls."

When FDA analyzes recall data, Trautman said, "the really sad part of it is that there is still a high percentage of recalls that are due to errors that are simply things that you would think would be common sense. I can't tell you how many recalls we still get that are due to mislabeling—the wrong product in the wrong box." It's "mind-boggling," she said, that many recalls still involve "silly, sloppy mistakes."

All design changes, including labeling changes and manufacturing changes, must go through design validation. Only rarely will verification alone be sufficient. Design changes are the focus of 21% of FDA-483 citations during inspections of design control systems. One of the top design control deficiencies is having an inadequate design plan.

"You do not have to have a design plan for every single design change," Trautman advised. It is sufficient, she said, to establish "an appropriate design control procedure that lays out how you're going to evaluate that change and what type of different steps it may kick into. The design change control procedure should allow for multiple different ways, depending on the complexity. . . . The procedure itself, if it's written properly, can act as the design plan for design changes."

The requirement for design transfer in 21 CFR 820.30(h) is poorly understood, Trautman said. "Design transfer does not happen at one particular time—it happens throughout the design process."

There is also some confusion about the difference between design verification (making the product right) and design validation (making the right product), Trautman said. Acceptance criteria must be established up front, before starting any verification or validation activity. Too often, she said, manufacturers perform validation studies without first establishing their goals and criteria. In effect, they launch elaborate studies, then try to make the results fit what they want to hear. FDA investigators "are always looking for traceability and documentation that the firm's verification confirms that design output meets the design input requirements, to see in the validation studies how the firm has gone beyond the verification testing," she said.

Trautman's presentation kept harking back to the role of risk management. "It's very important that you are able to show us how you have integrated the risk management system," she said. "We're going to be looking at that not only in the design control section but throughout, with respect to decisions on purchasing controls, acceptance activities, and any type of failure investigations or [corrective and preventive action] activities."

User Fee Implementation Delayed

Although FDA device user fees for PMAs, PMA supplements, and 510(k)s came into legal effect on October 1, FDA is not collecting payments until separate legislation authorizes it to do so. CDRH said last November that companies should not send in any payments until a Federal Register notice is published explaining the process.

Businesses with sales under $30 million seeking a reduced fee must submit their federal income tax return showing that their sales and receipts do not exceed the threshold. Firms must qualify as a small business at least 60 days before their first submission in any fiscal year, CDRH says.

For more information on device user fees, visit http://www.fda.gov/cdrh/mdufma/index.html.

Head of Bioresearch Monitoring Leaves FDA

Well-known CDRH director of bioresearch monitoring Charma Konnor retired from the agency on January 3 after 25 years of service. She will join the consulting firm of Phoenix Regulatory Associates (Sterling, VA) as senior manager/consultant for devices and drugs. Former FDAers Adam Trujillo (from the Office of Regional Operations) and Frank Fazzari (from Case Management/New Drug Stategy) already work for the consultancy.

In addition to her device work, Konnor served in FDA's Center for Drug Evaluation and Research, the Commissioner's Office, and in the field (as Florida District acting director). She has been recognized for her leadership in improving FDA premarket and postmarket programs, and in representing FDA in regulatory, legislative, and educational matters.

At CDRH Konnor was most recently responsible for the conduct of preclinical and human trials for investigational medical devices, implementation of CDRH's Application Integrity Policy, and enforcement of promotion and advertising regulations for investigational medical devices. Under her direction, CDRH's Division of Bioresearch Monitoring was transformed into a performance-based division known for its skillful management of compliance issues, its improved interactions with the Office of Device Evaluation, and its educational outreach to the clinical research community.

Copyright ©2003 Medical Device & Diagnostic Industry