Intended for healthcare professionals


Postmarket surveillance for medical devices: America’s new strategy

BMJ 2012; 345 doi: (Published 11 October 2012) Cite this as: BMJ 2012;345:e6848
  1. Sharon-Lise T Normand, professor of healthcare policy (biostatistics)1,
  2. Laura Hatfield, assistant professor of healthcare policy (biostatistics)1,
  3. Joseph Drozda, director of outcomes research2,
  4. Frederic S Resnic, chairman, Department of Cardiovascular Medicine3
  1. 1Department of Health Care Policy, Harvard Medical School, Boston, MA 02115, USA
  2. 2Sisters of Mercy Health System, Chesterfield, MO, USA
  3. 3Lahey Clinic Medical Center, Burlington, MA, USA
  1. sharon{at}

The FDA’s plan is a big ask but will be well worth the effort

In September 2012, the US Food and Drug Administration (FDA) released its vision for a postmarket surveillance system for medical devices,1 describing four key strategies to protect public health. But why this initiative now? Firstly, several concerns about the safety of specific devices, such as the Medtronic Fidelis defibrillator lead and the DePuy ASR hip implant, have garnered substantial press coverage.2 3 4 Secondly, the Institute of Medicine recently called for a substantial overhaul of processes for the approval and postmarket safety surveillance of medical devices.5 Thirdly, the FDA has undertaken several premarket initiatives to improve the efficiency of the regulatory review for innovative devices.6 With the number and complexity of medical devices growing rapidly, it is timely to complement the new premarket initiatives with a strong postmarket surveillance program.

Regulation of the entry into use of medical devices in the United States was introduced after that for drugs (table), and it quickly became apparent that the public health assessment of devices would need to be different from that of drugs. Unlike drugs, medical devices are assembled from many components, may affect patients long after the initial exposure, and have rapid product development cycles, which lead to frequent incremental improvements. Medical devices also pose unique challenges to studying safety: mechanical failure, operator (user) learning, and even software errors are potential causes of failure—none of these issues is usually a problem with drugs. Traditional pharmacosurveillance approaches may therefore be inapplicable or insufficient for devices.

United States FDA premarket and postmarket regulatory history

View this table:

What are the four new FDA strategies for improving the postmarket safety of medical devices? The first establishes a unique device identification system—an innovation that is long overdue. The FDA recently released its proposed identification rule,7 calling for a standard numerical identifier to be placed on all medical devices. Designed to be readable by humans and machines, these identifiers can be used in the supply chain software of health systems, in electronic health records, and in registries, greatly facilitating device tracking, safety surveillance, and research.

The second FDA strategy takes advantage of existing device registries within and outside the US and promotes the development of such registries where they do not exist. Clinical registries have arisen around specific medical procedures, such as orthopedic implants, coronary bypass surgery, and percutaneous coronary interventions. The FDA has identified the tremendous potential for evaluating devices by combining information between and within registries. The FDA helped to establish the International Consortium of Orthopedic Registries,8 which currently includes 29 registries from 14 nations and captures data from more than three million orthopedic procedures. Separately, and together with the US Centers for Medicare and Medicaid Services, the FDA has engaged the American College of Cardiology and the Society of Thoracic Surgeons to jointly develop the transcatheter valve therapy registry. This registry is a vehicle for both assuring postmarket safety of these new technologies and forming the basis for payment for the procedures.9 The FDA’s efforts to date, although laudable, have shown how far we have to go to tap the full potential of registries. Challenges abound, including problems regarding data ownership and data quality; the lack of well studied statistical methods for combining observational data from multiple registries; the difficulty of reaching consensus among multiple registry stakeholders with disparate interests; and, of course, costs for achieving all of this.

The modernization of the reporting and analysis of adverse events forms the third component of the FDA strategy. Current passive strategies of reporting adverse events do not provide adequate detail on the frequency of device malfunctions or complications. For drugs, the US Sentinel Initiative has established a mechanism to explore potential safety concerns in a large representative population of more than 100 million patients. The complexity of safety surveillance for medical devices, including the need to understand details of medical implant procedures and the impact of the effect of the learning curve on outcomes, challenges the existing capabilities of the Sentinel projects, which rely on administrative data, rather than detailed clinical data. Optimal safety surveillance is likely to include a combination of enhanced tools for reporting adverse events and automated prospective postmarket surveillance of the safety and efficacy of medical devices.10 This will probably use detailed clinical registries initially, with future efforts targeted toward the capture of emerging clinical data from electronic health records.

The fourth and final strategy involves the development of new methods for the generation, synthesis, and appraisal of evidence. Surveillance, by its very nature, involves observation of outcomes of medical devices used by clinicians in routine practice. It is difficult to agree on what constitutes sufficient evidence to warrant public health concern.11 Information from diverse settings and statistical designs will need to be appropriately treated and enhanced, and practical experience of doing this will need to accumulate. In addition, the use of data from sources that provide and pay for medical care will represent a substantial change for all stakeholders. Regulators and device manufacturers will need to embrace mechanisms for using a wide range of globally available data on devices, rather than relying on traditional sources of information alone.

The FDA’s proposed strategies will provide an important opportunity to improve the safety and effectiveness of medical devices as they are deployed in routine care. At the same time, these strategies present substantial challenges to all participants in the production, deployment, and receipt of devices. The FDA and its partners—industry, health systems, professional societies, academic medical centers, individual physicians, and patients—can rise to these challenges but not without considerable effort. Over the past 50 years, medical devices have contributed greatly to increasing the length, and improving the quality, of life for patients around the world. The FDA’s plan will support and enhance the ongoing development and effective use of these technologies and appears to be well worth the effort.


Cite this as: BMJ 2012;345:e6848


  • Competing interests: All authors have completed the ICMJE uniform disclosure form at (available on request from the corresponding author) and declare: all authors receive support from the Center for Devices and Radiological Health, USA Food and Drug Administration (contract DHHS/FDA-223201110172C) for the submitted work; S-LTN receives funding from the Patient-Centered Outcomes Research Institute in her capacity as vice chair of its methodology committee and funding from Circulation Cardiovascular Quality and Outcomes Research where she serves as senior statistical consultant; FSR receives funding for consulting services provided to Medtronic and Afga Healthcare, and from St Jude Medical for a cardiology fellowship educational program directorship; JD’s non-dependent son is a sales representative for Boston Scientific Corporation, Cardiac Rhythm Management division; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract