CDRH Moves to Strengthen Postmarket Surveillance

CDRH Moves to Strengthen Postmarket Surveillance It’s a cynical old joke that gets frequent play—in Washington: “I’m from the government and I’m here to help.” It got a re-run in September after CDRH posted a whitepaper entitled Strengthening Our National System for Medical Device Postmarket Surveillance. The document contains four specifics the Center said it is taking to achieve just that goal. Stakeholder reaction at a subsequent public meeting seemed cautious, if not skeptical.

The first specific was development of the long-sought “unique device identifier (UDI)” system, initiated in a proposed rule published in July. As CDRH sees it, the UDI system will be tailored to interact with electronic health records, clinical information systems, and claims data sources.

As a result, the whitepaper says, this information would potentially become available for use in assessing the benefits and risks of medical devices. UDIs will also allow FDA, the healthcare community and industry to more accurately report and analyze device-related adverse events by ensuring that critical device information is included in the reports. These device identifiers may also help reduce medical errors by enabling health care professionals and others to rapidly and precisely identify a device, obtain important information concerning the device’s characteristics (such as whether it contains latex or is magnetic resonance imaging compatible) and improve the clinicians’ ability to trace the device through the supply chain to the point of patient use.”

The second action, the whitepaper said, would encourage national and international device registry development. Creating individual registries to meet the postmarket surveillance needs for a specific manufacturer or a specific product is probably not efficient or economical, the paper said. So for targeted areas, CDRH believes it may be wiser to pursue nationwide medical device registries “focused on certain product areas of high importance as reflected by a large public health need, patient exposure, uncertain long-term or real-world device performance, or societal cost. For other device areas where the benefit-risk profiles are well-understood, registries may not be needed.”

Next, CDRH proposed to modernize its adverse event reporting and analysis to overcome limitations seen with the current spontaneous reporting system. The whitepaper said the Center it is working with partners to explore automated adverse event reporting systems that would minimize the effort involved in submitting device-related adverse events.

“For example,” it said, “we are working with 20 hospitals from our MedSun Network to develop software capabilities to export real-time adverse event data with device identifiers from hospital incident reporting systems.” Additionally, the Center is creating a new reporting system to replace its aging MAUDE system, and it aims to create a mobile application for electronically submitting adverse events over personal electronic devices.

And lastly, CDRH said it is planning to develop new methodological approaches for evidence generation, synthesis, and appraisal. “The development of new tools and methods to generate, synthesize, and interpret postmarket information will improve the efficiency and quality of decision-making by identifying new and better ways to leverage existing data sources by providing more timely information about the benefits and risks of marketed products, and by translating data into knowledge to help better inform regulatory and clinical decisions,” the whitepaper said.

CDRH also said it sees automatic signal detection as a promising opportunity. Real-time databases, such as EHRs or registries, can be mined with statistical software and techniques to identify early signals, the document says. “For example, the Data Extraction and Longitudinal Time Analysis (DELTA) system has been applied to cardiovascular device registries by FDA and academic investigators to establish proof-of-concept for detection of safety signals of approved cardiovascular devices. While automated signal detection software can facilitate the identification of potentially underperforming products, the review and validation of the findings remain critical aspects of their use.”

At a September 10 stakeholders meeting on the whitepaper, concern was expressed that any national medical device surveillance system that not only meets the needs of FDA and other government agencies but also those of various stakeholders would run the risk of being too burdensome for clinicians.

Several stakeholders expressed skepticism about pitfalls from a national, one-size-fits-all device registry. One stakeholder who is involved in a current national registry that was designed for various stakeholders said the registry has now become a database requiring 480 separate data fields for each registrant.

“Being all things to all people is easier said than done,” he told the meeting, adding that the challenge will be to create a user-friendly registry that is not unduly burdensome for clinicians.

But CDRH stuck to its guns. “Because there currently is no integrated and coordinated postmarket system in the country, perhaps today we need the registry that has 400 data inputs,” responded Division of Epidemiology director Danica Marinac-Dabic. “As we develop the postmarket surveillance strategy and as electronic health records are fully integrated and utilization of all this data by many stakeholders occurs... ultimately, I see in the years to come, we may not need to collect all the data in these current registries, which has become a burden on the clinical community.”

As the surveillance system progresses, she continued, “there is going to be a linking and it will much more seamless than it is working today... The path forward is that these data sources will be less burdensome as we work together.”

Various stakeholders were also concerned about what constitutes a safety signal and what to do once a signal is identified. “There is so much in terms of nuances of public health risk, the uncertainty of information, and the complexity of addressing it,” observed one stakeholder.

Addressing this, CDRH Surveillance and Biometrics director Thomas Gross said the agency’s goal under its signal management effort is to “develop a framework that presents systematic approach to identifying signals from various sources... and then to engage across the Center with subject matter experts on interpreting the signals and gathering additional information which could help refine our understanding of potential signals.”

Gross said the Center is cognizant of the great concern that when it deploys these systems for signal identification that it is done correctly to reduce chances for false positives. “That is the last thing you want to trigger,” he said, adding that more work has to be done to “enhance the sensitivity of the systems and decrease the tendency for false positives...

“There is some trade-off in terms of how much information (is needed) and how assured you have to be before you act on signals, and that is part of the dialog that we hope to engage stakeholders on, including methods development and the need to understand how confident we can be in the methods we deploy against the databases we use to generate potential signals.”