Originally Published December 2000
In an effort to persuade medical device companies to convert to its new paperless registration system, the agency is considering offering firms free promotional space on its Web site.
James G. Dickinson
- Kudos for Design Controls
- Sponsors Fail at Monitoring Trials
- Too Slow on "Least Burdensome"
- Remembering Lillian Yin
FDA is trying hard to find new and powerful ways of persuading medical device companies to register their facilities and list their devices more promptly than they do now—and to update those filings every six months as required by law. At the same time, the agency wants them to convert to an all-electronic, paperless filing system.
Need an incentive? How about a free plug for your product on FDA's Web site, which is being converted into a product-specific, health information portal for patients and health professionals alike?
In this hypothetical scenario, simply file the way you're supposed to and you'll get the chance to tell all about yourself and your products, on a site where FDA hopes millions will go to find objective, official information about products and safety.
It's a revolutionary idea under consideration at CDRH, according to Center Office of Compliance consumer safety officer Bryan H. Benesch, who discussed it at a grassroots informational meeting in Dallas in September. The idea—which may yet be shot down by agency higher-ups— is intended to provide an incentive for firms to convert to the real-time paperless system, which will combine separate registration and listing submissions and replace the MDR baseline form 3417.
Exactly what kind of voluntary company and product information companies could add to the new site has not been decided, Benesch said. The Dallas meeting, and another like it the next day in Irvine, CA, was intended to gather informal industry input before CDRH develops a formal regulatory proposal. Benesch cited two ideas for presenting information on this site: one, a space-limited panel where firms could post "important public health information, certain information concerning the firm, establishment information, and product information," and two, URLs linking to company-maintained sites.
When asked what FDA would do if companies wanted to put off-label-use information on the site or to link to it by URL, Benesch said CDRH recognizes that courts have ruled that some off-label information may be disseminated under certain conditions, and industry would be responsible for providing only legal information.
CDRH's main objective in converting its paper-based registration and listing (R&L) and MDR baseline filing requirements to the Internet is to improve the accuracy and accessibility of the information. This would allow the agency to combine the separate R&L databases and achieve efficiencies for both FDA and industry, Benesch said. After a formal proposal is published and final rulemaking takes place, the agency is considering waiting through a 12-month transition period, after which it would refuse to accept paper submissions.
The Dallas meeting produced some probing questions and a surprise for Benesch and two other CDRH presenters, policy analyst James G. Norman and supervisory consumer safety officer Howard Press from Surveillance and Biometrics.
The surprise was the industry audience's unexpected willingness to provide more, rather than less, information to FDA—industry members volunteered to post on the new site all MDR baseline (Form 3417) data for their products, at their discretion, rather than having to do it piecemeal if and when a report came in for that product. MDR baseline data are far more extensive than FDA requires for universal R&L data— which by law must be updated semiannually—collecting among other things product model numbers, catalog/serial numbers, shelf-life, manufacture start and end dates, site of manufacture, brand and generic names, and basis for marketing (PMA/510(k)).
Benesch contrasted the universal R&L requirements on some 13,000 to 14,000 regulated medical device companies with the MDR baseline data burdens that affect only about 600 firms, remarking that the Dallas audience would like "the option to front-end dump it all in. Then if you subsequently have to do MDR baseline, that data would already be there for us to pull into the pseudo-3417. This would be an option until MDR cut in."
The discussion arose out of skeptical audience reactions to a presentation by Press on the requirements of MDR baseline reports. Press said the data help FDA identify 510(k) information for any particular report that comes in: "We use the MDR contact information of the baseline, not the [form] 3500A."
When industry questioners demanded to know why this was the case, however, Press admitted that he didn't know the answer.
Benesch came to his rescue. "The baseline report is not necessarily being used the way those of us that invented it intended to use it," he said. "It was supposed to be an aid to quality control. You were supposed to tell us what the information was supposed to be, and if it showed up differently on the 3500A, we were going to ignore what was on the 3500A because you had told us, 'This is the way it is supposed to appear on all of our records.'"
According to Benesch, the baseline should be the correct information, which is why the agency pulls the contact information from it. "That way, we don't have problems with our data when we do a trend analysis," Benesch explained. "Baseline was supposed to be a gold standard. Unfortunately, it hasn't worked out the way we had hoped it to be fully implemented."
At this point, an audience member suggested that baseline reports might no longer be needed. The industry questioner explained that at his company, "one of the identifiers that we put on the MedWatch form report is the 510(k) number, which is the key to the R&L database. So once you have the 510(k) number, you know who did it, you know who the contact is, you know where it was made. You've got everything—you don't need a baseline report."
Benesch explained that the agency doesn't have model-level listing, to which the questioner replied, "But you don't need it."
That's when an another industry audience member suggested entering the baseline data one time, up front, as part of product listing. Benesch expressed surprise that industry would prefer a procedure he considered "grossly burdensome."
Voicing what was apparently the consensus in the room, the industry member replied, "I'd rather do it one time and be done with it."
Kudos for Design Controls
Industry managers, regulatory affairs consultants, and several FDA officials expressed general satisfaction—if not enthusiasm—in Washington in October for their experience with the implementation of design controls, mandated by FDA to help ensure the development of quality medical devices.
Speaking before attendees at the annual conference of the Regulatory Affairs Professionals Society (RAPS), Medtronic director of regulatory affairs Michael Holgers described the experience of his firm. He had a total of seven FDA inspections at four different facilities, all focused on the implementation of design controls.
Inspections were carried out for products, such as defibrillators, in both pre- and post-PMA periods—in some cases using the newly implemented quality systems inspection technique (QSIT). Holgers said inspectors examined each element of his firm's design control system— for example, the design development plan, risk analyses, conduct of design reviews, verification and validation, change management, and transfer of products to production.
A matrix, or "road map," to the complex verification process for the company's products was developed by Medtronic to help facilitate the inspections. Holgers said that this matrix, which details the requirements for each product, has also proven to be a useful management tool.
Medical Device Development Corp. president and CEO Larry Stevens cited QSIT, in tandem with a good system of design control, as invaluable to managers for identifying problems and making changes or corrections on a timely basis. He described design reviews as "the backbone of design control."
Stevens said the goals of a design review should be to confirm design intents, evaluate design verification, monitor project progress, and assume management oversight of the process. He said the first design review for a product under development should be held at the time it is "blessed"—when management makes a formal commitment and assigns a budget for the project.
Stevens recommended that design review teams be made up primarily of middle managers—technical experts capable of bringing independent critical faculties to the review process. He counseled emphatically against having CEOs attend design reviews, as their presence may inhibit criticism, and because their opinions often carry undue weight.
CDRH Office of Compliance consumer safety officer Jan Welch told the RAPS audience that since 1998, FDA has noted more than 1200 design control deficiencies in medical device inspections and has issued some 75 warning letters on these deficiencies. No legal actions have been taken.
She said the lack of a design and development plan is the most frequently cited deficiency found by inspectors. Other frequently cited deficiencies include inadequate design history files, inadequate design change procedures, and the lack of design control systems and of procedures for design changes.
In a review of the center's procedure since January of implementing the quality system regulation (QSR) and the use of QSIT in medical device inspections, Welch said that FDA design control inspectors begin their task by looking for procedures within a firm that provide the structure for an overall design control system. These procedures must be clearly defined and documented. At a minimum, each firm must have a documented design change procedure.
Once a design is developed, there must be a design plan. The plan's complexity will vary with the project, but it must define responsibility for implementation of the design and development activities, and must identify and describe interfaces with different groups or activities.
In response to a question, Welch said it is too early to tell whether the agency's implementation of the QSR will result in fewer product recalls, but she expressed optimism on this point.
Sponsors Fail at Monitoring Trials
During FY 1999, more than half the device sponsors inspected by CDRH's Division of Bioresearch Monitoring failed to ensure proper monitoring of their clinical investigation sites, division director Charma Konnor told RAPS in October.
Konnor emphasized the key role that clinical trial sponsors must play in ensuring the integrity of clinical data—primarily through their selection of only qualified investigators and by monitoring the clinical studies and exercising appropriate oversight. She cited the most common deficiencies among clinical investigators as failure to follow study protocols, failure to secure adequate informed consent, failure to keep adequate records, failure to maintain accountability of test articles, and failure to report adverse events.
In other presentations at the same RAPS session, speakers warned that investments in clinical studies in support of new product submissions can be rendered worthless unless careful attention is paid to ensuring the quality and integrity of clinical data.
Too Slow on "Least Burdensome"
Many in the medical device industry feel FDA hasn't moved fast enough in producing its long-awaited guidance document on the FDA Modernization Act (FDAMA) provision directing the agency to use only the "least burdensome" means of getting devices to market. This was Becton Dickinson regulatory affairs vice president Pat Shrader's reaction in October after hearing CDRH director David Feigal tell a RAPS briefing that the "least burdensome" guidance is a "work-in-progress." Feigal said his center has been working with industry on a joint task force to meld proposals from both sides, but that it will take a "cultural shift" at FDA to bring the concept to fruition.
Shrader, a member of the industry/FDA task force, did not seem to accept this. Although she acknowledged that it has become easier to communicate with FDA since the passage of FDAMA, she said more effort is needed on the "least burdensome" provision.
"I think industry would like to see FDA ask us to do what is needed to demonstrate safety and effectiveness, but not more than that," Shrader said. Issues still need to be addressed at the review level, which will take increased rigor and will require industry to stick its neck out. "If you don't get the response that you are hoping for, it will be very important to bring that to someone's attention." That may require an appeal through the center's ombudsman, Les Weinstein, Shrader hinted.
Feigal told the conference that FDA's historical approach to reviewing submissions has been to ensure a level playing field. "If your competitor had to do something a certain way last year, then it was likely you would have to take the same approach," he explained.
"Least burdensome" says that if FDA learned something from reviewing a competitor's product, however, then that knowledge could be used in reviewing the next product to make the process flow more smoothly. "It's a different mind set," Feigal continued, and one that will take a while to have an effect. "It's one of the things that if it is not right, then we need to hear from you."
CDRH Office of Device Evaluation deputy director Phil Phillips told the conference that a draft guidance on the principles of "least burdensome" is beginning to be circulated within the agency. He expects the guidance document to be published in the Federal Register in early 2001.
Also expected out in January is a draft guidance on 510(k) claims, Phillips said. "This is something FDA has struggled with for years," he added. The guidance will provide recommendations on when new claims will require a 510(k) submission and when they will require FDA to review the underlying substantiations, Phillips explained.
Remembering Lillian Yin
Lillian Yin, one of FDA's best-known reviewers and a 32-year veteran of the agency, died in September of an acute aortic aneurysm, at the age of 69.
According to an Office of Regulatory Affairs (ORA) note on her career, Yin was born in 1931 in Shanghai, China, where her father was an international banker. She traveled widely with him before moving to the United States with her sister just before the Chinese Revolution in 1949.
Yin and her sister went to Dominican College, a small Roman Catholic girls' college in San Rafael, CA, where Yin received her BS in chemistry in 1951. She received her MS and then PhD in organic chemistry from the University of Pennsylvania. She did postdoctoral research at the University of Pennsylvania, Hahnemann Medical College, the Medical College of the University of Pennsylvania, and the State University of New York before joining FDA's Bureau of Foods in 1968 as a laboratory research chemist.
In 1972, she joined the Office of Medical Devices, which later became the Bureau of Medical Devices and later still CDRH. In 1977, she was appointed the acting director of both the division of obstetrics-gynecology and radiology and the division of gastrourology and general hospital medical devices in the Office of Device Evaluation (ODE). From 1979 on, she served as the director of the division of reproductive, abdominal, ear, nose and throat, and radiological devices until she left ODE in April 1999 to join the Center for Biologics Evaluation and Research. Recently, she worked in ORA's Office of Enforcement coordinating the reinvention of its efforts on human subject protection and clinical investigators. "Lillian's devotion to her work will be missed by ORA and the entire agency," an ORA spokesperson said.
James G. Dickinson is a veteran reporter on regulatory affairs in the medical device industry.