Journal of Interventional Cardiac Electrophysiology

, Volume 36, Issue 2, pp 119–127

Postmarket surveillance of medical devices: current capabilities and future opportunities

Authors

    • Center for Medical Technology Policy
Article

DOI: 10.1007/s10840-013-9778-6

Cite this article as:
Blake, K. J Interv Card Electrophysiol (2013) 36: 119. doi:10.1007/s10840-013-9778-6

Abstract

Recalls of cardiac implantable electrical devices (CIEDs) currently impact hundreds of thousands of patients worldwide. Premarket evaluation of CIEDs cannot be expected to eliminate all performance defects. Robust postmarket surveillance systems are needed to promote patient safety and reduce harm. Challenges impacting existing surveillance mechanisms include underreporting of defects, low rates of return of explanted CIEDs, lack of integration of surveillance into normal workflow, underutilization of existing resources including registries, a lack of capacity of aging resources, multiple proprietary platforms that lack interoperability, and the unmet need for common data variables as well as newer methods to generate, synthesize, analyze, and interpret evidence in order to respond rapidly to safety signals. Long-term solutions include establishing a unique device identification system; promoting expanded use of registries for surveillance and post-approval studies; developing additional methods to combine evidence from diverse data sources; creating tools and implementing strategies for universal automatic, triggered electronic event reporting; and refining methods to rapidly identify and interpret safety signals. Protection from litigation and creation of financial and other incentives by legislators, regulators, payers, accreditation organizations, and licensing boards can be expanded to increase participation in device surveillance by clinicians and health care facilities. Research to evaluate the comparative effectiveness of surveillance strategies is needed. Interim solutions to improve CIED surveillance while new initiatives are launched and the system strengthened are also presented.

Keywords

SurveillanceMedical devicesCardiac implantable electrical devices

The public expects that medical devices will perform as intended to improve or protect health and will not cause harm. Unfortunately, this expectation has not always been met. In recent years, heart rhythm patients and doctors, the Food and Drug Administration (FDA), and manufacturers of cardiac implanted electrical devices (CIEDs) have had to address performance failures of implantable cardioverter defibrillator (ICD) and pacemaker pulse generators and leads [1]. CIEDs are typically listed as class III devices, those that “support or sustain human life, are of substantial importance in preventing impairment of human health, or which present a potential, unreasonable risk of illness or injury” [2, 3]. When a serious class III device performance failure is identified, the FDA and manufacturer issue a class I advisory or “recall.” Unlike the automotive industry, which has the option of addressing defects with repair or replacement, CIEDs are generally not removed or replaced in response to a class I advisory. More often, monitoring frequency is increased, detection algorithms are added to device software, and action is taken only when a safety signal consistent with either impending or early failure is detected [4]. Recalls remind patients and clinicians of the importance of having robust and reliable systems for premarket evaluation and postmarket surveillance of medical devices.

1 Premarket evaluation of medical devices

FDA approval of a new drug or device in the USA requires evidence that it is safe and effective [5]. Evaluation of drugs for use in humans has four phases. Phase I studies examine the basic safety and effects of a new agent in humans. Phase II studies in healthy individuals establish drug dose and identify additional safety problems. Phase III studies in the target population include at least one randomized clinical trial. Phase IV studies are initiated if collection of additional safety information during the late-phase premarket and postmarket periods is deemed necessary [6]. Device evaluation is based on a total product life cycle (TPLC) approach that includes consultation with industry during the conceptual, prototype, and preclinical phases; studies of biocompatibility and materials, environmental, and hazard analysis; and simulation, reliability, and human factors testing. Iterative device modification in response to the results of preclinical testing is an integral feature of the TPLC approach. Once an investigational device exemption is issued, clinical studies are performed in the target population, a premarket approval (PMA) or clearance decision is made, and postmarket studies are ordered if deemed necessary [7].

The vast majority of medical devices in use today in the USA have not undergone the PMA process. Instead, medical devices deemed equivalent to a previously approved predicate device are cleared, not approved, for marketing and commercial distribution using the 510(k) process, named after the section of authorizing legislation passed by Congress in 1976 [8]. A 510(k) clearance requires a manufacturer to notify the FDA 90 days in advance that it intends to market a new product it deems substantially equivalent to an already approved predicate device. Agency action at this phase is limited to confirmation that the device is indeed equivalent to its predicate. In 2011, in response to a number of high-profile device safety issues, the FDA asked the Institute of Medicine (IOM) to review the 510(k) process and address two questions: “Does the current 510(k) clearance process optimally protect patients and promote innovation in support of public health? And if not, what legislative, regulatory, or administrative changes are recommended to optimally achieve the goals of the 510(k) clearance?” [9]. The committee found that, because the 510(k) process was only intended to confirm equivalence of a device to its predicate, it could not be modified and used to also evaluate device safety and effectiveness. The committee also reported that the available evidence was insufficient to determine if the 510(k) process had contributed to or hindered innovation. The IOM committee recommended that the FDA issue a call for PMA applications for, and otherwise reclassify, class III devices that had previously received a 510(k) clearance. Although in practice most CIEDs have been subjected to the PMA process, adoption of the IOM committee recommendation would result in all CIEDs receiving the more intense scrutiny associated with that PMA review pathway.

2 CIED postmarket approval experience

At the time of market release, it is impossible to know how a CIED will perform over the long term in the real world. Biocompatibility, environmental, simulation, and reliability testing may not always replicate the patient experience. Pre-approval device studies may enroll only a few hundred individuals who comprise a relatively homogeneous study population. Rare adverse events may not be observed until a device is used in thousands of patients. PMA studies rarely span the TPLC of a new or comparator device. After approval, devices may be implanted in patients who differ significantly from the participants in premarket studies. As a result, meaningful evaluation of real-world device performance depends on a high-performing postmarket surveillance system.

In 2005, in response to recall notices from three CIED manufacturers in the preceding year, the Heart Rhythm Society (HRS) and the FDA convened a multi-stakeholder National Policy Conference on Pacemaker and Implantable Cardioverter Defibrillator Performance [10]. In October 2006, a task force issued recommendations that included a call for greater transparency in device surveillance, analysis, and reporting; enhanced systems to increase the rate of return of devices to manufacturers; improvements in recall communications to physicians and patients; and more cooperation among industry, the FDA, and the physician community [11]. Recommendations to improve CIED lead surveillance followed in 2009 [12]. Many of the task forces’ recommendations, such as regular product performance reporting by industry, have been adopted.

More recently, clinicians have become aware of performance defects associated with the St. Jude Medical Riata transvenous defibrillation lead. A 2009 report by Epstein and colleagues, based on their analysis of data from four registries with median follow-up of 22 months, had concluded that the percentage of Riata leads with conductor fractures (0.09 %) and insulation damage (0.13 %) was very low [13]. However, the lead was removed from the market in November 2010 because of reports that conductor cables were wearing through the insulation that covered them [14]. A class I advisory was issued 1 year later [15]. The manufacturer estimates that 227,000 Riata leads have been distributed worldwide. In a 2012 New England Journal of Medicine perspective soon after the Riata advisory, Hauser expressed concern that, “our current passive postmarket surveillance system [still] fails to detect significant device defects before large patient populations have been exposed” [16]. In addition to advocating for postmarket studies of Riata leads (ordered by FDA in August 2012), Hauser urged regulators to quickly adopt strategies such as expanding their use of clinical registries and remote device monitoring to improve surveillance. Writing at the same time as Hauser, Resnic and Normand highlighted some of the challenges associated with monitoring medical devices that are made using multiple components and frequently modified complex designs [17]. Complexity is more than a device hardware issue. In a 2006 review of ICD troubleshooting, Swerdlow and Friedman estimated that ICDs have approximately 500 programming options and expressed concern that the time pressure of clinical practice and low reimbursement for device troubleshooting might increase the risk of lethal and nonlethal operator programming errors [18].

Concerns about CIED complexity are not confined to ICDs. A 2012 analysis of data from 2.9 million pacemaker implants from 1993 to 2009 showed a rise in dual-chamber implants from 62 to 82 % of all pacemakers, with a proportionate decline in single-chamber procedures. The total number of implants increased 55.6 % over the same time period. Recipients were older and sicker [19]. As more devices of greater complexity are implanted in sicker patients, the CIED surveillance burden is greater. Along with estimates by FDA officials that only 0.5 % of medical device failures are reported, these data suggest that more patients than ever would benefit from improvements in CIED surveillance and earlier (preclinical) detection of device failure [20].

3 Current postmarket CIED surveillance mechanisms in the USA

The most widely used mechanisms for postmarket surveillance of CIEDs are presented in Table 1.
Table 1

Current mechanisms for postmarket CIED surveillance

Strategy

Issues

Solutions

Routine clinician office visits

Not “automatic”

Use only for identified issues

Very infrequent

Requires scheduling

Consumes resources (time, money)

Low yield

Transtelephonic monitoring

Not “automatic”

Financial incentives and recognition programs for clinicians and facilities

 ●Scheduling

 ●PQRS

 ●Enrollment

 ●MOC

Infrequent

 ●VBP

Requires equipment mastery

 ●Accreditation by JC

Low yield

Transition to remote monitoring technology

Remote monitoring

Enrollment not “automatic” and may not be offered to everyone

Facility/clinician incentives to enroll patients

Multiple proprietary platforms

Develop interoperable software

Establish a common platform

Information overload

Develop decision support tools

Develop reliable and valid alerts

Analysis of defective product

Reprogramming resolves issue

Universal identification system

Features are abandoned

 

Not a part of normal workflow

Build into EHRs

 

Part of normal workflow

Facilities don’t allow return

Incentives to return product (JC)

Potential for litigation

Liability protections

Postmortem exam uncommon

Incentives to obtain prior consent to postmortem CIED evaluation (PQRS)

Medical Device Reporting (MDR)

Many explanations for a defect

Track all devices electronically

Not part of normal workflow

UPI

Threshold for action not precisely defined

MDR templates in EHRs

 

Interoperable software

 

MDEpiNet

 

NCDR-ICD for f/u

 

Bring DELTA to scale

 

Modify BAA; include FDA

 

Add Pacemaker f/u module

Post-approval and 522 studies

Costly

Use registries, EHRs

Inefficient

Expand MDEpiNet

Low yield

PQRS physician quality reporting system, MOC maintenance of certification, VBP value-based purchasing, JC Joint Commission, EHRs electronic health records, BAA business associate agreements

3.1 Transtelephonic, in-office, and remote monitoring

Transtelephonic monitoring and in-office interrogation of CIEDs have been used for more than two decades to track device function. Unfortunately, not all CIED recipients are routinely monitored. Patients must “opt into” monitoring, make and keep appointments, and master the use of equipment. In-office device evaluation is performed less frequently than phone monitoring and requires appointments, transportation, and expenditure of other resources. The yield in terms of actionable data per session from transtelephonic and in-office ICD interrogation has been reported to be as low as 6.6 % [21]. Because of the intermittent nature of conventional monitoring, signals that herald impending device failure may be missed between sessions.

Remote monitoring technology is a newer approach that can be used for automatic CIED surveillance. In-home transmitters download device information, sending it to a server that receives and forwards information to clinicians. Although CIED recipients must still opt into remote monitoring, and products from one manufacturer require application of a wand over the device to retrieve data, automatic or semiautomatic data collection may improve monitoring frequency and safety signal detection. A recent review of this technology concluded that remote monitoring does result in earlier detection of performance defects compared to conventional monitoring strategies [22]. Nevertheless, barriers to remote monitoring remain. Older CIEDs that are still in service may not have remote monitoring capability. Not all currently marketed devices include this function. Therefore, device clinics must continue to provide access to multiple monitoring options and be familiar with each manufacturer’s proprietary platform. As with conventional monitoring, even if a device has remote monitoring capability, not all patients are offered this service. Others opt out of monitoring or do not use equipment even after it has been placed in their home [23]. Medical practices need software which integrates monitoring software with electronic health record systems, makes it a part of normal workflow, and alerts clinicians and patients when problems are first detected in order to realize more fully the potential of recent advances in monitoring technology.

3.2 Postmortem evaluation of CIEDs

The 2006 HRS Task Force members encouraged postmortem interrogation, removal, and return of CIEDs to the manufacturer, especially in cases of sudden or unexpected death [11]. The group also recommended that patients be asked to consent to postmortem evaluation of a CIED. In a survey conducted by Kirkpatrick and colleagues, only 4 % of 71 funeral directors reported having ever returned an explanted device to a manufacturer [24]. In the same study, 82 % of 150 CIED recipients said they would be willing to have their device interrogated after death, and 79 % indicated that they would consent to having it returned to the manufacturer. Although the study population was small, these findings suggest a high level of acceptance of postmortem CIED evaluation, and that an opportunity exists to increase the number of devices returned for analysis.

3.3 Reporting by manufacturers of CIED performance defects

Device manufacturers must inform the FDA when they become aware of a performance defect in one of their products. Specifically, they are required to, “protect the public health and well-being from products that present a risk of injury or gross deception or are otherwise defective” [25]. One challenge they encounter is that specific thresholds for reporting do not exist. Companies are required to produce annual medical device performance reports. The accuracy of those reports depends upon voluntary completion by clinicians of questionnaires that ask for information about patient status and device function. Medical practices may be unable or unwilling to disrupt workflow or add to overhead to complete this task. Patients may change addresses, physicians, or insurance and become lost to follow-up. Patient and device survival data are unlikely to be missing at random, and analyses of censored data probably underestimate device defects. When reasons other than a design flaw may also explain device malfunction or an event is so infrequent that most clinicians would not recognize it or if a high-risk patient dies suddenly, defects may escape detection. In addition, if a clinician is able to address a performance issue by reprogramming a device or abandoning a lead, manufacturers and regulators may not be notified. Patients may not know that they may report a device problem to the manufacturer and the FDA and may feel they do not have the medical knowledge needed to make a report.

Despite these impediments to communication of device performance issues to the makers of CIEDs, most medical device recalls result from voluntary action taken by manufacturers in response to field reports and/or analysis of returned product. CIED pulse generators are relatively easy to remove and return for analysis. However, return of malfunctioning leads is limited in part because lead extraction is riskier than generator removal, and because faulty leads may be abandoned and capped, rather than removed and replaced. During extraction, leads can break, be partially retained, or otherwise damaged. Manufacturers and regulators must then decide, using incomplete information, whether the reported performance defect is due to a design flaw or should be attributed to patient, operator, or other extrinsic factors.

4 The current FDA approach to medical device surveillance

Each year, several hundred thousand medical device reports (MDRs) of confirmed or suspected performance defects are sent to the FDA Center for Devices and Radiological Health (CDRH) [17]. The CDRH TPLC approach attempts to combine pre- and postmarket data on all of the devices it regulates [26]. In September 2012, the FDA released its plan for “Strengthening Our National System for Medical Device Postmarket Surveillance” and hosted public meetings to engage stakeholders, review existing intra- and extramural resources including registries, and discuss what methods are needed to rapidly generate, synthesize, and appraise evidence [2729]. At the present time, the FDA CDRH uses multiple platforms and strategies for device surveillance, the most important of which are described here.

4.1 The manufacturer and user facility device experience (MAUDE) database

MAUDE was established over 20 years ago to receive and archive MDRs from facilities, manufacturers, distributors, physicians, and patients [30]. MAUDE relies on voluntary reporting by participants and suffers from the absence of prespecified reporting thresholds, a plethora of reporting templates, and outdated software.

4.2 The medical product safety network (MedSun)

Established in 2002, MedSun is an FDA partnership with approximately 280 health care facilities, mostly hospitals, trained by the agency to report instances of suspected abnormal device function [31]. FDA training is credited with the submission, on average, of higher quality, more complete MDRs than are submitted to MAUDE. MedSun generates approximately 5,000 MDRs yearly. Most MedSun reports are about class II devices. Its contribution to CIED surveillance is limited.

4.3 Post-approval studies

CDRH at times requires additional studies as a condition of device approval. Since 2008, the FDA has required post-approval studies of all new or substantially modified ICD leads. These studies are designed to detect early signals of poor device performance by collecting data from at least 1,000 patients for 5 years after implantation. These requirements were not in effect when the Medtronic Fidelis and St. Jude Medical Riata ICD leads were approved [16].

4.4 Postmarket “522” studies

Section 522 of the Food, Drug and Cosmetic Act gives the FDA the authority to order a study of an approved device if a safety issue has been identified [32]. Analysis of existing data, observational studies, patient registries, and randomized controlled trials has all been used to address FDA concerns [33]. The FDA recently ordered a 3-year 522 study of the St. Jude Medical Riata lead, “to determine how frequently and how soon after implantation Riata insulation fails; how often and how soon both inner and outer layers of insulation fail and at what point migration or externalization of the electrical conductors cause ICD lead malfunction or other problems; and risk factors that contribute to insulation failure or externalization of the electrical conductors.” [34]. The FDA has also ordered the company to conduct postmarket studies of the QuickFlex LV CRT leads, QuickSite LV CRT leads, Riata ST Optim, and Durata ICD leads.

5 Newer FDA medical device surveillance strategies

5.1 Unique device identification (UDI) initiative

Just as vehicle identification numbers have facilitated tracking of product performance in the automotive industry, a unique device identification system is expected to improve tracking of medical devices. In July 2012, the FDA released for public comment a proposed rule that would establish a UDI system [35]. Once operational, it is expected that UDI data will be incorporated into electronic health records, claims databases, and other clinical information systems. Participants in the Medical Device Epidemiology Network (described below) are expected to use the UDI in database searches when responding to FDA queries about possible device performance failures.

5.2 The sentinel and the medical device epidemiology networks (MDEpiNet)

The FDA Amendments Act of 2007 instructed the agency to establish a public private partnership and develop methods to access disparate and widely distributed data sources to better evaluate the safety of drugs and devices [36]. First created, the FDA Sentinel Initiative is a distributed network of databases that collectively contain health information about over 100 million individuals. Participating health systems, health plans, pharmacy benefit managers, and government agencies including the Departments of Veteran’s Affairs and Defense and the Center for Medicare and Medicaid Services (CMS) retain exclusive access to their data. The Mini-Sentinel Pilot project tested the network’s ability to respond to a drug safety query [37]. FDA submitted test queries through the coordinating center at the Harvard Clinical Research Institute (HCRI). Participants queried their databases and provided de-identified results to the HCRI. Summary statistics were forwarded to FDA, allowing the agency to decide if the evidence warranted further investigation. Mini-Sentinel demonstrated the feasibility of this approach. The FDA Safety and Innovation Act of 2012 mandated expansion of Sentinel to medical devices [38]. The MDEpiNet has been expanded to fulfill this mandate and as a public–private partnership will develop methods of analysis for use across diverse databases to investigate safety signals related to devices [39].

5.3 Longitudinal Patient Registries

The Agency for Healthcare Research on Quality User’s Guide to Registries defines an outcomes registry as “an organized system that uses observational study methods to collect uniform data (clinical and other) to evaluate specified outcomes for a population defined by a particular disease, condition, or exposure, and that serves one or more predetermined scientific, clinical, or policy purposes” [40].

The Society for Thoracic Surgery Database, The Interagency Registry for Mechanically Assisted Circulatory Support, and the National Cardiovascular Data Registry (NCDR) contain data on millions of patients who have undergone cardiovascular procedures, including device implantation [4143]. The NCDR-ICD was established in 2005 in response to a National Coverage Decision by the CMS that it would provide coverage for implantation of ICDs for primary prevention of sudden cardiac death in high-risk Medicare beneficiaries so long as hospitals meet a Coverage with Evidence Development requirement to submit data to an approved registry or a clinical trial [44]. Data are submitted by participating facilities at the time of an implant but the registry is not designed to collect patient and device data between procedures. Quarterly reporting of key indicators, benchmarked to national outcomes and facilities of similar size and case volume, is provided to each participating facility and to CMS. Business associate agreements between the NCDR and each hospital govern data sharing and ensure protection of patient-specific information. Data sharing between hospitals and with other stakeholders is not permitted. The NCDR-ICD receives data related to approximately 90 % of all ICD implants in the USA. Its potential as a platform to strengthen ICD surveillance is obvious. To date, only limited funding for research and device surveillance has been available. The FDA has announced plans to host workshops to examine how it can leverage registry experience and expertise, establish common data elements, develop and share methods, promote interoperability among registries, set priorities, and pre-certify registries for use in post-approval studies [28, 30]. Challenges to be addressed include the need for greater clinician engagement, registry sustainability, and creation of governance models that ensure transparency and inspire trust.

6 Modernization of adverse event reporting and analysis

The FDA plans to promote automated adverse event reporting as a part of normal clinician workflow, expand its capacity to receive data electronically, and develop mobile reporting applications [28]. The agency will replace MAUDE, after more than 20 years in service, with the FDA Adverse Event Reporting System (FAERS). FAERS will increase FDA’s capacity to receive and analyze adverse event reports and to rapidly identify safety signals. The FDA plans to develop data storage standards, tools for data mining text, methods to combine data from diverse sources, and a master plan for data management. Early results from the Data Extraction and Longitudinal Time Analysis (DELTA) system suggest that automatic safety signal detection for cardiovascular devices is possible [45].

The FDA’s initiatives ought to benefit from recent investments in health information technology (HIT). In response to provisions of the Health Information Technology for Economic and Clinical Health and the Patient Protection and Affordable Care (PPACA) Acts, the Office of the National Coordinator for Health Information Technology (ONC-HIT) has created incentives for adoption and meaningful use of HIT [4648]. The final rule for the second stage of meaningful use of HIT was issued in August 2012 [49]. Eligible providers seeking incentive payments for meaningful use of HIT have the option of identifying and reporting specific cases to a specialized registry (other than a cancer registry). Although medical device surveillance is not specifically referred to in the rule, clinicians monitoring CIEDs could use registry participation to fulfill one meaningful use requirement. The FDA and the ONC-HIT should coordinate their efforts and use every opportunity to create incentives for HIT vendors to include UDIs, automatic device surveillance, and registries in electronic health record platforms.

7 Evaluation of postmarket device surveillance systems

In addition to infrastructure and new methods for medical device surveillance, critical assessment of each component of a surveillance program is necessary in order to decide if continuing investment of resources is warranted. Other than its intention to replace MAUDE, the FDA has not announced any plans to phase out other existing programs while it implements its plan to strengthen postmarket surveillance. Comparative effectiveness research focused on different approaches is needed to ensure that funding decisions of regulators and payers and clinical decisions of physicians and patients are justified by scientific evidence.

Surveillance strategies should be evaluated using performance measures such as the time from the first confirmed instance of a performance defect to regulatory action, the number of recipients affected, the number of corrective actions taken, the severity of complications resulting from device malfunctions and corrective actions, and the economic and noneconomic costs associated with an advisory. The principle underlying safe use of diagnostic radiation, aiming for a level of exposure As Low As Reasonably Achievable, should be adapted for use in the medical device arena, to minimize patient exposure to performance defects, device failures, inappropriate shocks, and corrective procedures. Evaluation metrics should also account for the greater risk of extraction the longer a patient has a lead in place. Methods that shorten the time from first signal to action benefit not only those who have a lead removed and replaced, but also those who have avoided receiving that lead in the first place. Algorithms to improve preclinical detection of CIED malfunction, such as those developed at the time of the Medtronic Fidelis lead advisory, should continue to be developed and tested [4].

8 Future challenges and opportunities to improve medical device performance and surveillance

Highly reliable performance evaluation programs assume that there will be defects in the systems that they monitor and employ multiple strategies to ensure early detection of performance defects and reduce the likelihood of harm. James Reason has suggested that errors leading to patient harm result from a series of defects in safety and monitoring systems, and has argued against assignment of blame to any one step, person, company, or agency [49]. Instead, he recommends a “management program aimed at several targets: the person, the team, the task, the workplace, and the institution” that is defined by its “preoccupation with the possibility of failure” [49]. In such a program, participants share responsibility for preventing harm. Reason’s conceptual model of harm prevention can be applied to device surveillance systems to effectively engage all participants, including patients and clinicians.

The unique device identification system will be developed for new devices, but implementation may take 7 years and it will not be applied retroactively to devices already implanted in patients [35]. The merits of linking a UDI with a unique patient identifier (UPI) are obvious. The USA is the only developed nation without a UPI system [50]. Opposition has been based upon concerns about patient privacy and skepticism about the role of the government in the lives of individuals. It is possible that provisions of the PPACA that prohibit denial of coverage based on preexisting medical conditions, place limits on rates charged to higher risk individuals, and expand access to insurance to millions of Americans will create a more favorable environment in which to reconsider a UPI system. Until then, a voluntary nongovernmental approach could perhaps be considered. Recipients of medical devices could be given the option of a UPI that is linked to the UDI system. In the meantime, clinicians and health systems should consider using currently available levers to increase patient participation in remote CIED monitoring when possible, and conventional monitoring when it is not. Protocol-driven enrollment in remote monitoring could become a requirement of discharge from a facility after CIED implantation. Just as facilities and clinicians are striving to prevent post-discharge adverse events such as readmission for congestive heart failure or device infection, similar strategies including early telephone and in-person follow-up could be used to increase participation in CIED monitoring. Protocol-driven reminders can also be used to improve participation rates and initiate corrective action when remote or in-person device surveillance is overdue. Local initiatives at the practice or institutional level have the advantage of being under local control and do not depend on regulatory or legislative actions.

Individuals and organizations act to earn incentives and limit downside risks. The National Quality Forum (NQF) should call for performance measures that incentivize device monitoring and device registry participation. CMS could then include NQF-endorsed measures in its Physician Quality Reporting System, public reporting (Hospital- and Physician Compare), and value-based purchasing programs. The Joint Commission (JC) could require device surveillance systems as a condition of hospital accreditation. The JC could also create recognition programs for accredited facilities that routinely return explanted products to manufacturers. Protections from litigation and limits on liability both have the potential to increase rates of return of explanted products by facilities and clinicians. Participation in device registries such as the NCDR could become a requirement for accreditation or special recognition. The NCDR should leverage its experience with ICDs and develop a module for permanent pacemaker procedures. The FDA should contract with the NCDR for CIED surveillance data and pre-certify the NCDR-ICD as a platform for postmarket studies [28]. The American Board of Medical Examiners could incorporate medical device management into practice improvement requirements for maintenance of certification in clinical cardiac electrophysiology [51].

As the postmarket medical device surveillance system is strengthened, and gaps are addressed, clinicians may want to consider preferentially implanting CIEDs with longer track records. Recently approved devices could be reserved for patients who require the newer features. Both patients and clinicians need better access to near real-time information about the device they are about to use and depend less on annual performance reports. Patients, physicians, and other stakeholders should advocate for creation of incentives so manufacturers continue to produce CIEDs with excellent long-term track records. Health technology assessment groups and payers making coverage decisions could require evidence that a new device improves clinical and patient-reported outcomes compared to existing options. All stakeholders need to balance the demand for innovation with the need for reliable device performance.

9 Conclusion

A robust surveillance system depends upon rigorous evaluation and continuous improvement of the devices it monitors, the infrastructure it builds, and the methods it applies. Reliable systems require investments to develop state of the art infrastructure and methods. A commitment to invest in high-performing assets and a willingness to decommission underperforming infrastructure and tools are essential. Together, engagement of all stakeholders, a unique device identification system, incentives to promote return of explanted devices, adoption of electronic health records, greater participation in registries, more remote monitoring, and integration of monitoring as a part of normal workflow have the potential to dramatically improve medical device surveillance in the USA.

Copyright information

© Springer Science+Business Media New York 2013