Medical Device Link.

Originally Published MDDI December 2004

WASHINGTON WRAP-UP

Schultz continues his campaign to improve postmarketing study implementation and calls for a standardized reporting system.

James G. Dickinson

FDA Won’t Adopt ISO Standard on Device Trials | FDA Is Getting Fewer New Devices | FDA Guidance on Third-Party 510(k) Reviews | FDA Wrestles with Advisers’ Clinical Study Ideas | FDA Inspections Increased by 19% Last Year

CDRH director Daniel Schultz, MD, is pressing ahead with his number one priority: improving the quality and efficiency of medical device postmarketing studies. In October 2004, he repeated the call he made soon after becoming director in July.

Addressing the annual meeting of the Medical Device Manufacturers Association, Schultz repeated his view that problems with such postmarket studies are all too common. He also said that FDA would spend more time looking for ways to get better and more-complete information on consumers’ experiences with medical devices. He did not elaborate; presumably, details are still being developed.

“Many postmarket studies of medical devices are ill conceived,” Schultz told the conference. “All too often, they are not even initiated—or if they are, they are not completed.” But he noted, as he has before, that FDA also shares some of the blame.

“Too often, postmarket studies of medical devices that are undertaken by manufacturers are either not tracked by FDA, or requirements for postmarket studies specified by the agency are not enforced,” he said.

In addition, Schultz called for a standardized reporting system to be used with such postmarket studies. He also suggested that the status of postmarket studies of medical devices be made public.

“Steps are being taken to make information on postmarket studies of drugs and biologics more easily available to the public. There is no reason why similar information related to medical devices should not also be more accessible to the public,” he said.

Schultz did not, however, offer any specifics concerning the agency’s intention, if any, to develop new rules or guidances for the conduct of the postmarket studies.

According to Schultz, certain issues are of particular interest to FDA. These issues should be clearly addressed in designing postmarket studies. They include the long-term safety of devices and their performance in community practice. Study designs should note any changes in the user setting, rare or unexpected events associated with use of the product, or change in the rates of anticipated adverse events. Information about factors such as use errors, incorrect product use by consumers, and off-label use should also be features of the study design.

The need to strengthen and improve postmarket studies of medical devices assumes special significance, Schultz said. CDRH is now in the process of striking a new balance between the pre- and postmarket requirements that the agency will impose on device manufacturers.

“By moving some premarket requirements to the postmarket period, we can help speed products to market,” he said. Schultz said this move would free up Office of Device Evaluation staff to conduct premarket reviews. The change also is expected to help generate data for approving next-generation devices and enhanced labeling, Schultz noted.

FDA Won’t Adopt ISO Standard on Device Trials

FDA does not plan on adopting the ISO 14155 standard for medical device clinical trials as a “recognized standard.” However, it may consider doing so when planned revisions are complete in 2005. This is according to an address from CDRH deputy director for medical affairs Kimber Richter to the Regulatory Affairs Professional Society (RAPS) annual meeting in October.

The FDA Modernization Act of 1997 enables the agency to recognize certain standards to which device makers can declare conformance in lieu of submitting data, Richter explained. “But in the case of good clinical practices, I think we are going in the other direction,” she continued. “FDA is actually looking for more evidence and more submitted information on good clinical practice compliance.”

Richter said the agency sees ISO 14155 as a “common sense” worldwide baseline for good clinical practices. It provides good guidance to clinical investigators and clinical research organizations.

“We think it should increase acceptability of clinical data,” Richter said. It could speed up approvals because sponsors’ data would be more internationally acceptable. ISO 14155 includes updated terminology, emphasis on the need for a protocol, and clarification on how literature review is part of the clinical assessment. However, she said, it does not meet FDA’s expectations on informed consent and institutional review board (IRB) membership.

As for the revisions planned for ISO 14155, Richter would like to see closer harmonization between ISO and the International Conference on Harmonization’s good clinical practice expectations. This could enable similar terminology and minimize the confusion of IRBs and investigators who are involved in both drug and device testing, she said. Revisions are needed for adverse-event reporting, which is currently a bit confusing, and more detail is required on clinical protocols and final reports, she added.

Meanwhile, Richter said, the Global Harmonization Task Force is forming a new study group on clinical evaluation to address the processes.

“The concept of clinical evaluation really comes from Europe. It’s a sense that we should look at all existing information and determine then whether a product is safe and effective and whether clinical studies are needed,” she explained. The goal is to begin with a literature search, examine laboratory testing, and assess all information available for the clinical performance of the product. Then the group can determine whether a clinical study is necessary, she said. The new study group will also provide guidance on assessing international data and how it can prevent repeat studies by having different countries accept them. It will also discuss government oversight of clinical trials and informed consent.

FDA Is Getting Fewer New Devices

When the final numbers are tallied, CDRH expects to find that it received 7% fewer new medical device submissions in FY 2004 than in FY 2003. So does that mean review times should be improving, with more resources from user fees and fewer submissions to review? Not necessarily, deputy CDRH director Linda Kahan told the RAPS annual meeting in October.

She noted that CDRH’s workload is changing and that “the numbers can be deceiving.” Kahan said that in FY 2004, the center logged in a record-high of nine expedited PMA applications (only three were submitted in 2003). The expedited PMAs and combination-product submissions require more resources, she said. Real-time PMA supplements also have been rising, with 193 submitted in FY 2003 compared with 139 the previous year. Modular PMAs also have been increasing; 73 were filed last year—more than twice as many as were received in FY 2001, when 32 came in. Modular reviews require reassembling the review team many times and use more resources than traditional PMAs, she said. And 510(k)s being submitted with clinical data have also increased.

Reduced submissions will be felt in original PMAs, PMA supplements, and 510(k)s, said Kahan. Original PMAs are down from 54 in FY 2003 to 50 in FY 2004. PMA supplements are down from 669 in FY 2003 to 619 in FY 2004, and 510(k)s are down from 4247 in FY 2003 to 3533 in FY 2004.

With CDRH’s user fee program, the center is struggling to meet its cycle goals, Kahan told RAPS. For example, only 19% of the eligible PMA supplements submitted in FY 2003 met the center’s “not approvable” decision timeline. And 58% of 510(k)s met an “additional information” cycle goal, she said. “Cycle goals will be much harder to meet than decision goals, especially at the higher performance percentiles.”

Kahan said the center is having trouble reducing GMP preapproval inspection times. “New review timelines will require a 28% improvement in timing, conduct, and conclusion of preapproval inspections and reports,” she said, adding that it will require increased coordination with the Office of Regulatory Affairs and district offices. She also said the center may develop criteria for industry to determine inspection “readiness.”

One area needing major improvements, Kahan said, is the center’s information technology (IT) infrastructure. There needs to be better tracking, reviewing capabilities with IT tools, collaboration, correspondence management, and archiving and imaging. With imaging, Kahan said, paper submissions are taking up to two weeks to scan electronically. With improved IT functions, it will be possible in the future to begin accepting electronic submissions.

Kahan said the center has been reviewing supplemental validation submissions on cleaning, sterility, and functionality of device reprocessors’ critical and semicritical devices. Most of these submissions are currently under review. FDA began posting information about its decisions on CDRH’s Web site on November 1, 2004, so purchasers and hospitals can find out which products will be available.

FDA Guidance on Third-Party 510(k) Reviews

A new CDRH guidance provides assistance to third-party reviewers on their 510(k) substantial-equivalence determinations under FDA’s third-party review program. The guidance, titled Guidance for Third Parties and FDA Staff—Third Party Review of Premarket Notifications, supersedes a 1996 document. It discusses key elements to consider when evaluating a 510(k) submission and documenting the reviewers’ recommendations. It applies to third-party reviews of traditional, abbreviated, and special 510(k) submissions regulated by the Office of Device Evaluation. It also applies to in vitro diagnostic devices regulated by the Office of In Vitro Diagnostic Device Evaluation and Safety.

The guidance may be seen at FDA’s Web site, www.fda.gov/cdrh/ode/guidance/2237.html.

FDA Wrestles with Advisers’ Clinical Study Ideas

FDA faces many challenges when requiring postapproval studies for PMA applications. One challenge comes from recommendations advisory committees, says CDRH Division of General, Restorative, and Neurological Devices director Celia Witten. Such panels often make suggestions for follow-up studies, Witten told a RAPS audience in October.

She said it could be difficult refining a panel’s question or concern into a form that can be answered by a postapproval study. “Some of the questions a panel may have or the suggestions they may have are very general, and if we think it’s important to get that information, then it’s important for us to work on refining that question.”

Sponsors are included in discussions following a panel. That way, when the PMA is complete, the basics of the study can be outlined in the approval order. Such basics might include study duration, size, and assessment tools. Then sponsors are expected to formally submit the clinical protocol in a PMA supplement for the agency to review. However, Witten said, often all the details of the study have not been agreed upon.

Witten outlined for RAPS the types of questions CDRH expects to answer from postapproval studies rather than from other forms of postmarket surveillance. They include:

• Long-term information on failure rates and modes.
• Instructions for a device modified during trial.
• Focused information on specific safety events noted in a premarket study.
• Evaluation of a population that was not well represented in a trial.
• Confirmed performance characteristics of a modified device.

Often, a device that is the subject of a postmarketing study may undergo a postmarket change. The change presents a new challenge, Witten said. In such cases, CDRH will need to decide whether the study should enroll additional patients for the modified device. Comparative studies are also a challenge in the postmarket phase because the device is already available, and patients want the newer technology.

Additionally, if a sponsor is required to follow a cohort that was enrolled in the premarket trial, there could be concerns about patient retention and informed consent. Witten suggested that sponsors talk to FDA when designing premarket trials to determine whether there will likely be postapproval concerns or follow-up data collection.

FDA Inspections Increased by 19% Last Year

Medical device inspections jumped 19% last year, from 2490 in FY 2002, to 2970, according to statistics in FDA’s annual Enforcement Story report. At the same time, Class 1 recalls, FDA’s most serious classification, nearly doubled, leaping 91% from 44 in FY 2002 to 84. Domestic and import device sampling rose 11%, from 1664 samples in FY 2002 to 1847. Additionally, medical device recalls dropped 10%, falling from 1454 recalls in 2002 to 1300 in 2003.

To see the data, visit FDA’s Web site at www.fda.gov/ora/about/enf_story/ch2/cdrh_charts.htm.

Copyright ©2004 Medical Device & Diagnostic Industry