Robert Hauser, MD published a perspective piece in the New England Journal of Medicine today entitled "Here We Go Again — Another Failure of Postmarketing Device Surveillance." In the piece, he offers these suggestions on how to detect these failures earlier:
Opportunities have emerged for automated tools to prospectively monitor multicenter device databases for early, low-frequency adverse events and to compare suspect devices with established products that have been shown to be reliable.(4) The goal is a postmarketing surveillance system that not only detects device problems early but also accumulates the data needed to guide patient care. Until such a system exists, St. Jude Medical should initiate a study with these attributes for recalled Riata and Riata ST leads and for the currently marketed Durata leads. Indeed, all manufacturers should conduct postmarketing studies of this type for marketed class III devices that sustain or support life.
It has been 3 years since the FDA launched the Sentinel Initiative, as Congress, in 2007, directed it to do.(5) The intent of this new system is to supplement the current passive adverse-event reporting with an active, real-time network capable of identifying any safety or efficacy issues soon after a new drug or device is marketed and then communicating the information in a timely manner to health care providers and the public. Creating such a system is an ambitious undertaking, and the initiative aims to gather electronic health data from 100 million people by the end of 2012. Formidable challenges lie ahead, such as setting priorities, developing analytic tools, and deciding when and how to alert the public if a safety signal is detected. Thus, we are years away from a fully operational Sentinel Network.
Jay Schloss, MD offers another thoughtful perspective on this issue in his guest post over on Cardiobrief:
It is worth noting that all high impact device failures in the cardiac rhythm device industry have occurred well after device approval. In most cases, failures have been rare events occurring many years after these devices became clinically available. The industry is littered with high profile device failures, some of which serve now as footnotes or memories from the more gray haired members of the EP community: Medtronic polyurethane pacing leads, Ventitex Cadence ICDs, and Telectronics Accufix pacing leads were all devices that failed well after they had been introduced to the market. It was through case reports from concerned doctors – not the FDA — that these problems were brought to light.
Trying to absolutely prevent the failure of medical devices through the FDA approval process would be challenging to say the least. This would require an anticipation of the potential failure mechanism before the fact to design a study that would detect said defect. Moreover, the number of patients and length of study for these trials would be daunting, impractical and quite likely prohibitively expensive.
Hauser’s call to enforce a robust active postmarket surveillance system would allow ongoing innovation and appropriately timed approval of new technology. With the ability to carefully monitor a large population of approved devices “in the field,” sentinel failures could be detected and acted upon as they occur. With public reporting of these events, doctors could draw their own conclusion of whether to keep implanting these devices and prospective studies could then be designed promptly to determine the scope of the problem. That would be a system that protects patients while still allowing innovation.
Neither author mentions how their approaches might be funded. Nor is there a clear consensus on how to determine a threshold for sounding a recall (or "advisory"). Should it be announced before a root cause of the problem is known or suspected, or after? Pulling a recall trigger too early before a root cause of the defect is identified leaves doctors and patients completely in the dark about how to manage these complicated situations. These issues are still debated today.
What is clear, though, is that the efforts to prevent such widespread recalls continue and doctors (as shown by these two perspectives) remain eager to participate in the process.