Originally Published MDDI November 2005
CDRH hopes the threat of publicly posting the status of postapproval and condition-of-approval studies will push manufacturers to complete such research efforts.
By James G. Dickinson
FDA's Guidant Investigation Becoming Media Driven | Defibrillator Regulation Moves to the Top at CDRH | Report Every Malfunction Publicly, FDA Advisor Urges | FDA Still Pondering Bar Codes for Devices | FDA Quality Inspections Are Getting Longer | FDA Emergency Use Regulation Raises Liability Concerns | FDA Cites Staar on 3 QSIT Issues | Third-Party Inspections Could Cut Others, FDA Says
FDA is threatening public exposure to force the long-festering issue of unfinished condition-of-approval studies by device makers.
CDRH intends to publicly post the status of premarket approval (PMA) postapproval and condition-of-approval studies. The plan was announced in an August draft guidance.
The problem has been plaguing both drug and device regulators for a long time. However, FDA's Center for Drug Evaluation and Research tackled it years ago under a congressional order. The center has achieved major reductions in the number of unfinished, and even unstarted, studies.
But CDRH has been dragging its feet. Device reviewers are no longer confident in their power to condition approvals on such studies. That, in turn, has stopped them from approving problematic devices. Last March, the center issued a report admitting this undesirable result. It also disclosed that of 45 devices conditionally approved between 1998 and 2000, 26 (58%) had no record of the required studies even being started.
August's draft guidance said CDRH is also considering putting into effect other conditions for new device approvals. For example, it might have FDA staff and sponsors make reports to FDA advisory committees on the status of postapproval studies suggested by those panels.
Both plans are part of awkward efforts by FDA to be more transparent in the wake of recent drug and device safety issues. Criticism from stakeholders about FDA's lack of oversight on postapproval matters also prompted the plans.
CDRH now intends to have PMA sponsors submit and gain approval for their follow-up studies prior to final device approval, the new draft guidance says. If a protocol is not agreed upon prior to approval, FDA can allow it to be submitted as a PMA supplement after approval based on an agreed-upon time frame.
Device sponsors also must submit a postapproval study status report every six months during the first two years after approval. Thereafter, they should submit one annually until a final study report is submitted.
But what if a sponsor and FDA cannot agree on an acceptable protocol, or a sponsor fails to comply with study conditions? In those cases, the agency may take further regulatory action, says the guidance. This could include a postmarket surveillance order or an order to withdraw the device from the market. “In addition, a significant or knowing failure to report information about a postapproval study, or where such failure constitutes a risk to public health, may result in civil money penalties,” it says.
To access the draft guidance, go to www.fda.gov/ohrms/dockets/98fr/05d-0348-gdl0001.pdf.
What happens when an FDA investigation into a firm's manufacturing quality problems and delays in reporting malfunctions is propelled by news media disclosures? In some cases, it may permit glimpses of a normally cloistered process.
Consider the recent case of Guidant Corp. Its long-running FDA quality and defibrillator recall issues have been detailed in The New York Times. This has spawned a novel response from Guidant. The firm has created an outside expert advisory panel on how it should deal with the issues.
Although FDA investigations are usually kept under wraps, this one seemed to be instigated in a highly public way: because of a Times report in May. The report stated the company had not told doctors about the potential for electrical short circuits in its Ventak Prizm 2 DR Model 1861 defibrillator. It also said that the firm kept selling older versions of the product after making an improved one in 2002.
The Times began exploring the issue after a college student who had one of the units implanted died of sudden cardiac arrest. This prompted Guidant to issue a letter conceding that 26 implanted Prizm 2s had failed. The firm described the problem as a rare failure that results in the device's inability to deliver therapy. Guidant suggested that doctors continue monitoring patients every three months as described in the labeling. But the letter said early replacement was not recommended because “overall device reliability remains high.”
Only after continued prodding by the Times and other media did FDA issue a safety warning about the Guidant products on June 5. Two weeks later, it followed this with a mandatory recall order to Guidant. The recall affected 38,700 implantable defibrillators and cardiac resynchronization therapy defibrillators. It included not only the original Ventak Prizm 2 DR Model 1861, but two other defibrillator models with short-circuit defects. Those products are the Contak Renewal Model H135 and Contak Renewal 2 Model H155.
It's not clear when FDA staff actually reviewed the relevant Guidant report. CDRH says it would require too many resources to review the hundreds of MDR filings FDA receives each year. Also, it would take time to resolve which data could be released and which should be treated as confidential.
CDRH director Daniel Schultz also talked with the Times. He said he does not believe the effort of disclosing massive amounts of data submitted by manufacturers would be an effective use of FDA resources and time. “It does not at first blush look like an efficient way of getting information to the public in a timely fashion,” he said.
Meanwhile, Guidant is keeping an eye on its legal liability exposure. It is trying to get out ahead of the curve by broadening its FDA issues to a potential industry-wide problem. The firm says that the problem lies with the MDR regulation. It selected University of Miami professor of medicine and cardiologist Robert J. Myerburg, MD, to recruit members for an independent panel. The panel will advise Guidant on its medical device surveillance efforts and related communications.
Myerburg has recently notified Guidant that he has completed the recruitment process. The 12 people he identified include MDs, PhDs, and one attorney: Peter O. Safir, a senior partner at Covington & Burling. Safir is known as one of the leading food and drug regulatory lawyers in the country. Also appointed was former CDRH director David W. Feigal. He is currently a research professor at Arizona State University's Biodesign Center.
The panel is charged with many goals. One is to develop “more-transparent, understandable, and clinically useful communication processes to physicians and patients.” This includes triggers for communication, timing, and novel methods of transferring data.
Guidant also intends for this panel's results to be made public, “in conformance with FDA and regulatory guidelines.” It will use both internal data from Guidant and other sources. The panel will report on ways that Guidant can “enhance its capabilities in understanding, detecting, and disseminating important information.” The panel will also make public its nonproprietary remarks and suggestions about these issues that may be useful to others in the industry.
Guidant's novel approach to its problem with FDA may not be appropriate for everyone, though. First, can you afford it? Second, can you spin your MDR compliance problem into an industry-wide issue?
How plaintiffs' lawyers, judges, the rest of the industry, and FDA regulators will view this panel's output remains to be seen. But the step Guidant has taken is both bold and imaginative, with the clear potential to be far-reaching.
FDA in September announced it is taking a new approach to regulating defibrillators and pacemakers. The new approach is based on an internal study of 12 years' data on implantable cardioverter defibrillator (ICD) and pacemaker failures. The agency said it needs to improve the way it regulates these products and that changes are being made. The study included a review of 366 PMA yearly reports that covered hundreds of pacemaker and ICD models.
A team of FDA officials released results and discussed the new FDA changes at a Heart Rhythm Society meeting in Washington, DC.
Overall, the study found that the number of faulty pacemakers removed and replaced in patients had decreased. However, the numbers for ICDs have increased. Reasons for the increase have not been established, but FDA staff said it may relate to the devices' complexity, the manufacturing challenges they pose, or more-frequent physician reporting.
An FDA news release said the study also noted that careful monitoring of device performance is needed. Also, doctors need better ways to return explanted devices to firms for analysis and to report adverse events.
The study found that from 1990 to 2002, about 2.5 million pacemakers and 416,000 ICDs were implanted in the United States, and 8834 pacemakers and 8489 ICDs were removed from patients due to confirmed device malfunction. The yearly ICD failure replacement rate of 20.7 replacements per 1000 implants was much higher than the 4.6 pacemaker replacements per 1000 implants. The pacemaker failure replacement rate also decreased during the study. The ICD replacement rate trended down during the first half of the 1990s but increased in the study's latter half. Notably, more than 50% of the ICD failures occurred during the study's last three years. The bulk of reported failures did not lead to death or serious injury. However, an FDA statement said failures were to blame for 61 confirmed deaths out of the three million devices implanted.
CDRH director Daniel Schultz said the data don't call for any patient action. The data “alert FDA that there is a trend that needs to be addressed.” The data also point out the need for FDA to improve the way it regulates these products, he said. He assured industry that FDA was doing just that.
Schultz said the top priorities to improve monitoring and issue earlier safety warnings include the following:
• Improving CDRH's ability to obtain critical data about medical device failures.
• Communicating those data clearly and rapidly to doctors and the public so they can use the data to make sound, informed medical decisions.
• Coordinating company yearly report data within CDRH to allow for an integrated approach, leading to more-efficient and timely review.
• Forming a CDRH working group to improve communication within CDRH so data about ICD and pacemaker problems can be quickly reviewed and evaluated by staff and shared more rapidly with the public.
• Changing operations based on an ongoing internal review of CDRH's postmarketing program.
FDA says that the changes being considered include the following:
• Designing an electronic system for adverse event reporting to make data available to CDRH analysts more quickly.
• Targeting resources to inspect firms making high-risk devices.
• Developing a guidance for firms submitting annual reports.
• Developing a guidance that more-clearly defines when changes to devices need prior FDA review and approval before being implemented.
To see the FDA news release, visit www.fda.gov/bbs/topics/news/2005/new01231.html.
William H. Maisel thinks device makers should be required to report every confirmed device malfunction to the public. Maisel is chairman of the FDA Circulatory System Medical Devices Advisory Panel. His comments appeared in the August 25 issue of the Journal of the American Medical Association in response to reporting delays by Guidant involving an ICD.
Maisel also leads the cardiovascular division at Harvard's Beth Israel Deaconess Medical Center. He says many believe that notification about every failure is too burdensome and results in physician liability. That argument, he says, ignores a doctor's duty to act in the patient's best interest. “Public reporting does not necessarily need to be in the form of a direct physician or patient letter. Updated malfunction information, for example, could be provided on the manufacturer's Web site. Direct physician notification, FDA advisories, and manufacturer press releases would be warranted for particularly important safety issues.” Such safety issues include observations of recurrent failures or identification of a product that does not meet expectations for reliability.
Maisel also called for mandatory patient notification about device safety issues and recalls. These should be similar to how consumers are notified when an automobile is recalled, he said. Currently, doctors are the ones notified when a device problem arises.
“The outdated device notification system is rooted in the concept that medical information contained in a recall notice is too complex for a patient to understand,” he wrote. “However, modern patients are knowledgeable, informed, and proactive, and they have the right to know about malfunctions or potential defects affecting their implantable device. Physician notification should remain in place, and patients should be advised to seek the counsel of their physician when deciding on the clinical management of their recalled device. Ideally, a staged notification system would be initiated, whereby physicians are notified first and are given the opportunity to contact their patients. Patients would then be contacted directly to ensure that all those affected are notified.”
Bar coding or similar device identification systems are reviewed in a just-released joint FDA/ECRI white paper. In 2004, FDA rolled out its bar coding plan intended to reduce medication errors involving pharmaceutical and biologic products. At that time, FDA said it was not including devices under the rule because they raised a number of concerns. Such concerns included product diversity and a lack of a standard identification system. The white paper is intended to serve as a resource as FDA moves forward and discusses whether such a device ID system is feasible.
The white paper is based on a literature search. It explores a number of technology solutions for applying a device ID system, and seems to favor radio-frequency identification technologies because they offer more capabilities than traditional bar coding systems.
To view the white paper, visit www. fda.gov/cdrh/ocd/ecritask4.html.
Medical device inspections under FDA's quality system inspection technique (QSIT) are averaging about 56 hours in 2005. This represents a steadily lengthening “creep,” according to New Jersey District consumer safety officer Robert Ruff.
Speaking at a University of Rhode Island School of Pharmacy GMP conference in Cambridge, MD, Ruff said he recently completed an analysis of QSIT's first five years of operation. He found the inspections have consistently been 50 hours until the past three years, when they began to creep up.
Similar inspections took 45 hours in 2002 and 48 hours in 2003, according to Ruff's data. Before QSIT was first rolled out in 2000, device inspection time was averaging 96 hours, he said.
QSIT was originally developed to decrease inspection time, increase inspection focus, and harmonize inspections with international regulatory agencies. It was also to provide adequate coverage of FDA's quality system regulation (QSR).
In FY 2004, QSIT establishment inspection reports found 17% of inspections were classified as official action indicated, compared with about 13% before QSIT was implemented, Ruff said. Inspections classified as no action indicated accounted for 43% of all inspections in 2004, compared with 48% before QSIT began. Inspections classified as voluntary action indicated stood at 40%, the same rate as before QSIT was used.
Ruff said his data show the agency is making more FDA-483 observations. However, he said, it is not necessarily finding significantly more violative companies.
Ruff said 38% of all 2005 FDA-483 citations were related to corrective and preventive action (CAPA) issues. That was followed by management controls (15%); records, documents, and change control (14%); production and process controls (12%); design controls (9%); material controls (7%); and equipment and facility controls (5%).
“A lot of the issues found in CAPA come back to preproduction design,” he said. “I'm a firm believer that spending time in design is well worth the effort. I think that a lot of problems and potential problems that find their way to the shop floor or to your customers take their first breath somewhere in design.”
Product liability tops AdvaMed's concerns about an FDA emergency use authorization (EUA) regulation. The regulation allows the use of unapproved medical products in emergency situations.
In comments to FDA, the group says that it is crucial that FDA understand that liability concerns may dissuade manufacturers from requesting EUAs. AdvaMed also stressed that firms may be unwilling to provide their products. This may be true even if a third party provides added labeling information for emergency use.
AdvaMed also mentioned a recent FDA draft guidance, Emergency Use Authorization of Medical Products Availability. The guidance says that a third party could obtain a labeling change without approval from the device maker. This interpretation, AdvaMed says, suggests another reason to immediately address liability concerns. The group worries that allowing such changes subjects the manufacturer to “liability that it would not foresee or otherwise face.”
The group suggested one possible solution, citing a legal analysis prepared by Mark Heller of Wilmer Cutler Pickering Hale and Dorr. Heller wrote that the Support Antiterrorism by Fostering Effective Technologies Act of 2002 (SAFETY Act) may provide liability protection for emergency-use devices. According to Heller, “a product authorized for emergency use under the Project BioShield Act should automatically be designated a qualified antiterrorism technology [QATT].” The SAFETY Act offers liability protections. And, since products designated as QATT fall under the SAFETY Act, they would be covered by such protections.
AdvaMed urged FDA to create a process allowing for immediate SAFETY Act designation for all EUAs. That will, it says, remove the barrier of legal liability from the use of unapproved products in emergencies.
A recent FDA inspection of Staar Surgical has led to three FDA-483 observations, according to the firm.
The inspection of Staar's Monrovia, CA, manufacturing facility began on August 29 and was completed on Sep-tember 14.
In July, FDA wrote Staar and said it faced an enforcement action. Staar was told to provide documentation proving it had corrected or was correcting GMP deficiencies. An inspection of the same plant a year ago found those deficiencies. The letter said Staar “failed to adequately correct numerous violations” noted on the FDA-483, according to the firm. It said that FDA is “gravely concerned about Staar's serious, continuing violations.” The company also said FDA is prepared to seek the appropriate remedies. At the time, FDA gave the company 10 days to show that the earlier deficiencies were corrected.
In the latest FDA-483, FDA noted one of the observations was “corrected and verified.” The firm promised to correct the remaining two, its statement says. “Based on the company's corrections to the quality system issues identified by FDA in previous inspections and the findings of FDA in this inspection, the company does not believe that enforcement action by FDA is likely at this time,” it says.
One portion of the inspection relates to a July 28 approvable letter from CDRH on the firm's PMA for the Visian implantable contact lens. The firm's statement says final approval remains subject to a review of the inspection's results by CDRH staff.
Staar has been in hot water with FDA for several years. An April 23, 2004, warning letter to the firm cited it on objectionable clinical study conditions found during an inspection the previous November. Prior to that, FDA sent the firm a December 22, 2003, warning letter for not reporting failures related to the firm's intraocular and implantable contact lenses. “In one case, a physician had to perform a vitrectomy on a patient after the lens delivery system failed,” the letter charged. “In another incident, a physician had to suture a patient after the cartridge failed,” it said.
Device firms requesting an inspection under FDA's Accredited Persons Program could reduce their inspections by other regulatory agencies, says a recent guidance. This may be possible if they select a third-party inspector recognized by the governments of the foreign countries in which they market their products.
The agency has been unable to attract industry volunteers for these inspections, which must be paid for by the inspected company.
The guidance is titled Requests for Inspection by an Accredited Person under the Inspection by Accredited Persons Program Authorized by Section 201 of the Medical Device User Fee and Modernization Act of 2002. It says firms can also maintain greater control over inspection timing by having third-party accredited inspectors audit their GMPs to satisfy FDA's biennial inspection requirement.
The guidance outlines eligibility criteria that companies must meet for such inspections to occur. For example, it says manufacturers are eligible for a third-party inspection if their previous inspection was classified voluntary action indicated or no action indicated. Additionally, eligible firms must market a device in the United States and in at least one foreign country.
To view the guidance online, visit www.fda.gov/cdrh/comp/guidance/1532.html.
Copyright ©2005 Medical Device & Diagnostic Industry