Stakeholder Meeting on MDUFA III Reauthorization
March 22, 2011, 1:30 – 3:30 PM
HHS Humphrey Building, Washington, DC
To discuss post-market signal detection.
|Malcolm Bertoni||Office of the Commissioner (OC)|
|Nathan Brown||Office of Chief Counsel (OCC)|
|Kate Cook||Center for Biologics Evaluation and Research (CBER)|
|Natalia Comella||Center for Devices and Radiological Health (CDRH|
|Synim Rivers||Office of Special Health Issues (OSHI)|
|Don St. Pierre||CDRH|
|Ruth Watson||Office of Legislation (OL)|
|Paul Brown||National Research Center for Women & Families|
|Eric Gascho||National Health Council|
|Catherine Jeakle Hill||American Association of Neurological Surgeons|
|Mellanie True Hills||StopAfib.org|
|Darby Hull||Consumer Federation of America|
|Jeanie Kennedy||American Academy of Orthopedic Surgeons|
|Jennifer Liljeberg||American Society of Cataract and Refractive Surgery|
|Andrew Morris||United Spinal Association|
|Martha Nolan||Society for Women’s Health Research|
|Kate Ryan||National Women’s Health Network|
|Rosalyne Schulman||American Hospital Association|
|Cindy Tomlinson||American Society for Radiation Oncology|
Additional Registered Stakeholders
|James Baumberger||American Academy of Pediatrics|
|Cynthia Bens||Alliance for Aging Research|
|Diane Edquist-Dorman||National Organization for Rare Disorders|
|Hilary Hansen||Spina Bifida Association|
|Andrew Sperling||National Alliance on Mental Illness|
Meeting Start Time: 1:40 pm
Post-market Signal Detection
Dr. Ron Yustein, Associate Director for the Office of Surveillance and Biometrics (OSB), delivered a presentation on the detection of post-market safety signals at CDRH.
The limitations of pre-market clinical data must first be appreciated to understand the focus of FDA’s post-market work. The clinicians involved in pre-market trials are typically experts with a great deal of experience, are closely monitored, and highly trained; they are not necessarily representative of physicians who will use the device when it is on the market. The patient population in a trial typically reflects a very select group meeting strict inclusion and exclusion criteria; these trials enroll patients most likely to benefit from the device and are not necessarily representative of all patients who will receive the device when it is on the market. The study designs for device trials also suffer limitations: limited numbers of patients are enrolled, which limits the statistical power to evaluate rare safety events or subgroups; some potential risks may not be anticipated and therefore not evaluated; there is a potential for product-product interactions; and, long term follow-up is costly and not always practical. Important safety issues may not be identified even by large trials. For example, in the human drug area, the post market identification of safety issues has led to labeling changes and market removals for drugs that were approved after large preapproval trials that did not identify the safety problem. . For all of the reasons described above, CDRH believes it is important to complement our pre-market processes with a strong post-market surveillance program.
CDRH defines a signal as “information about a product we regulate that suggests an unexpected risk to patients or users.” Unexpected risk could refer to “new events” (i.e., not observed/suggested by pre-market data, not addressed in labeling, not generally recognized as a potential clinical complication, or not generally recognized as a potential understand malfunction/failure) or “not new events” (i.e., unexpected frequency/rate of an expected event, unexpected severity of outcome for an expected event, or unexpected population experiencing the event). Not all signals are generated from safety signals; unexpected lower effectiveness can be a signal as well, especially when it impacts the risk/benefit profile.
There are many categories of signals, including signals relating to: device performance or use (e.g., malfunctions or incorrect use by physicians); recalls/uncleared modifications (e.g., as in the case of “silent recalls” when a manufacturer making changes to a device in response to a safety event they noted); primary device design/human factors (i.e., how the device and user interact with each other); labeling/instructions (e.g., inadequate/misleading); manufacturing and quality systems (e.g., problems at manufacturing plant where something is not being done or inspected correctly); marketing and promotion (e.g., off label use); and MDRs (Medical Device Reports) (e.g., sponsors may not meet requirements of reporting adverse events in a timely fashion).
Post-market signals can come from within the FDA or from outside the FDA. Within FDA, potential signal sources include: MDRs; MedSun reports; post-approval studies; device registries; research; review of annual reports or device modifications; inspectional findings; and/or other Centers. Outside FDA, potential signal sources include: other U.S. government agencies; foreign government agencies; medical journal articles, media stories or inquiries, list-serv/blog chatter, presentations at society or professional meetings, manufacturer website or promotional materials, and patient safety organizations.
Since December 1984, the FDA MDR regulations have required firms who have received complaints of device malfunctions, serious injuries, or deaths associated with medical devices to notify FDA of the incident if the device may have caused or contributed to a death or serious injury or if a malfunction is likely to cause or contribute to a death or serious injury, if it were to recur. User Facilities must report to FDA and the manufacturer if they become aware of information that reasonably suggests that a device has or may have caused or contributed to the death of a patient. They must also report to the manufacturer if there is evidence that reasonably suggests that a device has or may have caused or contributed to a serious injury to a patient of their facility. Voluntary reports are accepted from anyone for any reason. CDRH receives 150,000 – 300,000 MDRs/year and the trend has been increasing. Most MDRs represent malfunctions (50-60%) and most are submitted by manufacturers (95%). These reports are accessible through the public MAUDE database. When FDA receives an MDR, analysts consider multiple factors in interpreting the event (natural history of the disease/disorder, underlying comorbidities, concomitant medications or therapies, the surgical procedure during use of the device) and carefully interpret trends in numbers. If necessary, analysts may collect more information from the device manufacturer, review premarket information, and/or evaluate relevant compliance activities. There are 15-20 MDR analysts in CDRH, which makes it difficult to find signals in the 150,000 – 300,000 MDRs/year. Therefore, CDRH is piloting data mining software now and is looking for additional tools to help with signal detection.
There are limitations of MDRs as a signal source: they come from a passive surveillance system which relies on others to report;, under-reporting by end users who are often not aware of the need to report, the large volume of MDRs makes it difficult to identify problems; there are no reliable numerators and denominators upon which to determine rates; it is generally not possible to determine causality or root causes; currently, MDRs are paper based; and the MAUDE database is old and sometimes unreliable. Additionally, there are several reporter-dependent limitations to MDRs as a signal source: poor quality of reports (i.e., lack of sufficient details); poor follow-up on the clinical event; poor root cause analysis; inability of manufacturers to retrieve devices for evaluation due to hospital liability policies; and manufacturer misinterpretation of what is reportable.
Through the Medical Product Safety Network (MedSun) Program, CDRH maintains interactive relationships with 350 hospitals across the U.S. The risk managers at these hospitals are specifically trained to look for and report device-related adverse events. Approximately 3,000-4,000 adverse events are reported by MedSun hospitals as MDRs. Additionally, MedSun hospitals share information with CDRH on close calls/near misses, provide collaborative networks, are a resource to CDRH for quick access to physician expertise, and participate in CDRH rapid response surveys.
Post-approval studies are often ordered by CDRH as conditions of approval to Premarket Approval Applications (PMAs) and Humanitarian Device Exemptions (HDEs). They are intended to gather information related to device performance/problems for any or all of the following reasons: longer term performance (especially for implants); real world experience outside the select patient population and physicians from the premarket trial(s); evaluation of physician training program effectiveness; and/or sub-group performance. Post-approval studies are not used to evaluate unresolved issues from pre-market needed to establish a reasonable assurance of safety and effectiveness. General information on completed and ongoing post-approval studies is available at: http://www.accessdata.fda.gov/ scripts/cdrh/cfdocs/cfpma/pma_pas.cfm.
CDRH is working to identify additional signal sources. For example, CDRH works with third parties (i.e., academia, professional societies) to develop device registries from which rates of events can actually be determined. CDRH has also developed unique collaborations/partnerships with the National Veterans Affairs (VA) catheterization laboratory data repository (monthly discussions on adverse events), Kaiser Permanente (monthly calls with the national recall director and small scale surveys), and the National Children’s Medical Center (documentation of problems with ventilators). The Sentinel Initiative is an Agency-wide effort to build the national infrastructure of electronic healthcare data systems for safety surveillance.
Upon detection of a signal by CDRH staff by any means described above, CDRH evaluates all other potential sources of related information including: MDRs/MedSun reports, post-approval studies, registries, literature, compliance information, pre-market reports and history, labeling, and data on similar products. FDA has the authority to order a Section 522 surveillance study for class II and class III devices when certain conditions are met. This study could consist of a new clinical trial, a literature analysis, or an animal study. 522 studies are limited to three years of data (except for devices used in pediatric patients). OSB may call on other Offices to assist in evaluating and acting on a safety signal. For example, the Office of Compliance may inspect a manufacturing facility or clinical research site, issue a warning letter, issue an untitled letter, mandate a recall, or pursue a seizure or injunction. The Office of Science and Engineer Labs may conduct bench or animal research to better understand the signal. The Office of Communications may issue a public communication regarding the signal. Additionally, OSB ensures that lessons learned from safety signals are fed back into the pre-market through communications with the Office of Device Evaluation and the Office of In Vitro Diagnostics. Based on post-market safety signals, these offices may modify their approach to pre-market review for a device type. For example, infusion pumps now require assurance cases; radiation treatment planning systems are no longer eligible for review by a 3 rd party or via the Special 510(k) paradigm; CT scanners, hemodialysis tubing, and ceramic on ceramic hip systems have updated information in their labeling; guidance documents have been drafted/revised for infusion pumps and hemodialysis tubing; recognized standards for enteral feeding tubes have been reassessed; and letters have been sent to manufacturers requesting design changes to dental hand pieces and denture adhesives which contain zinc.
In conclusion, FDA believes that a strong post-market leads to a stronger pre-market, which leads to safer devices entering the market.
Questions, Answers, and Discussion
Stakeholders asked when HDEs are used. FDA explained that these are marketing applications for devices where the population is less than 4000/yr, similar to the orphan drug concept. FDA can order post-approval studies when approving HDEs and two such studies are currently ongoing.
Stakeholders asked about the differences in CDRH’s authority for post-approval studies and 522 orders. FDA explained that 522 orders relate to specific types of issues, such as evaluation of failures that are reasonably likely to have serious adverse health consequences. 522 studies are typically ordered after an issue is detected. Post-approval studies are ordered (and often developed) at the time of marketing approval.
Stakeholders asked about interpretation of reports in the publicly available MAUDE database. FDA explained that a single MAUDE report could be for one or many patients just as a single patient procedure may be represented in multiple reports. For this reason it is difficult to interpret numbers and rates. FDA cautioned those analyzing the MAUDE database to consider this important limitation.
Stakeholders inquired what CDRH’s main issues are for MDUFA reauthorization. FDA responded that no proposals have been presented to Industry yet. The Agency expects that discussion of specific proposals will begin in the next several weeks, and may be able to report on discussion of specific proposals with stakeholders beginning with the next meeting at the end of April.
Multiple stakeholders highlighted their interest in special populations, especially women, in clinical trials. FDA indicated that upon approval of a new device, FDA publishes a Summary of Safety and Effectiveness of the Device (SSED) on the web, which contains information regarding the population in the clinical trial(s) so that meaningful differences of patients enrolled may be understood. For devices cleared through the 510(k) process, less information is currently available; however, FDA is working on a labeling repository (i.e., centralized site for final device labeling) as an action item following the FDA’s 510(k) evaluation and report. Additionally, FDA is working on a guidance document to encourage increased enrollment of special populations in device clinical trials.
Stakeholders asked about CDRH’s authority to inspect manufacturing facilities. FDA responded that all PMAs include pre-approval inspections. For 510(k)s, the statute provides that FDA cannot withhold a finding of substantial equivalence for issues related to manufacturing, unless FDA finds that there is a substantial likelihood that the failure to comply with good manufacturing requirements will potentially present a serious risk to human health. In such cases FDA can conduct pre-clearance inspections. For example, FDA inspects heparin manufacturing facilities before clearing a device with heparin from that site as well as all infusion pump manufacturing facilities.
Stakeholders also asked when it is necessary to convene an Advisory Panel prior to determination of approval. FDA explained that there is a statutory provision which requires the Agency to take all “first of a kind” devices to the Advisory Panel for review. For second generation devices, FDA uses discretion in determining when such review is necessary. FDA may also choose to send specific information and specific questions to Special Government Employees who sit on Advisory Panels as “homework assignments” in lieu of or in addition to holding a meeting.
FDA acknowledged receipt of a letter from the Medical Device Patient and Consumer Public Health Coalition and thanked the Coalition members for their feedback and suggestions. One suggested topic (data registries for implantable devices) was discussed. FDA noted that some of the other topics suggested by the Coalition might fall outside the scope of MDUFA reauthorization.
Schedule of Future Meetings
Future meetings will be held on April 27, 2011, May 19, 2011, and June 16, 2011, all in the afternoon per stakeholder preference.
Meeting End Time: 3:30 PM