Part V of our volume, “Medical and Legal Oversight of Medical Devices,” can be thought of as the part that tries to address the question of “what now?” Previous parts have grappled with the regulation of medical devices as they are developed and come onto the market. Part IV considered the impact that devices may have on patients and family members once they are approved for us. Part V takes that focus a step further to consider how we should monitor, evaluate, and regulate medical devices once they are approved and on the market.
This is an important question because, despite the best efforts of regulators to evaluate products before they reach patients, not all medical devices will prove themselves entirely safe. Sometimes, flaws or challenges in a medical device will only be revealed when there are a wider number of users, beyond the scope of any clinical trial. Therefore, it is critical that the medical system develop methods of flagging concerns with approved devices, and that the legal and regulatory system be able to respond to these concerns. This is a challenging task for both the medical and legal systems, however. It essentially asks how we put the rabbit back in the hat. The rabbit, in this case, being the approval and availability of medical devices post-initial approval.
The authors of the chapters in Part V consider the challenge of monitoring the “rabbit,” following up on concerns regarding the rabbit, and regulating the rabbit from different perspectives. Some chapters focus on the regulatory system as the actor who can properly supervise and deal with the rabbit. Sanket Dhruva, Jonathan Darrow, Aaron Kesselheim, and Rita Redberg open the part with “Ensuring Patient Safety and Benefit in Use of Medical Devices Granted Expedited Approval.” They flag that with a more flexible and streamlined approval process comes an increased chance of unforeseen risks to patients. Therefore, it is necessary to update postmarket requirements to require fuller studies. Efthimios Parasidis and Daniel Kramer likewise turn to the regulatory system to provide sufficient postapproval oversight in their chapter, “Compulsory Medical Device Registries: Legal and Regulatory Issues.” While Dhruva et al. argue for postmarket studies, Parasidis and Kramer support the use of registries to track patient experiences with approved medical devices. They note that registries are perhaps underdeveloped as a tool to monitor medical devices, especially around data governance.
David Rosenberg and Adeyemi Adediran close Part V by considering the interplay of the regulatory system and market pressures in their chapter, “Strengthening the Power of Health Care Insurers to Regulate Medical Device Risks.” Similar to the Dhruva and Parasidis chapters, this piece turns to regulatory solutions to solve postapproval problems. This chapter is different, however, in that it focuses on using regulatory solutions to harness the market power wielded by insurers to adopt or avoid certain medical devices. This chapter highlights for the reader that once medical devices are approved by regulatory agencies, we move beyond a relationship focused tightly on the manufacturer and the regulator, to add in payors and patients.
The other two chapters in Part V are less focused on putting the rabbit back in the hat and more focused on the role of the medical system in monitoring and responding to postapproval issues. Anthony Weiss and Barak Richman look to how the medical profession can incorporate medical technology into physician self-regulation mechanisms, namely peer review. Their chapter, “Professional Self-Regulation in Medicine: Will the Rise of Intelligent Tools Mean the End of Peer Review,” flips the focus in the part. From considering how we can continue to supervise and regulate medical devices, Weiss and Richman instead ask how can we use medical devices to supervise and regulate human practitioners of medicine? Megan Wright and Joseph Fins, in their chapter, “Regulating Post-Trial Access to In-Dwelling Class III Devices,” consider the ethics of risky medical devices embedded in the human body. While Wright and Fins touch on regulatory best practices for following up on study subjects with these implanted devices, they focus strongly on the ethical implications of leaving or removing these devices posttrial.
Overall, the authors of Part V remind us that regulatory approval to bring a medical device to market is not a “happily ever after” or even a final chapter in a story. Instead, approval can be considered a midpoint or inflection point. The subsequent story, of how to monitor, identify problems, and address challenges in approved medical devices, raises significant questions. Our authors grapple with the right mechanisms to tackle these challenges, including the legal and medical systems.
In recent years, legislative mandates and regulatory policy in the United States have sought to streamline testing and approval requirements for novel medical devices with the goal of lowering development costs and accelerating market entry. But increasingly flexible approval requirements mean greater uncertainty as to the extent to which authorized medical devices will benefit patients without unforeseen risks. Some authorized medical devices have later been found to have safety or effectiveness concerns, but once a product is marketed it can be difficult for regulators to take remedial action. There are several reasons for this, including a reluctance to engage in regulatory self-reversal; physician and patient enthusiasm for novel technologies; generous payor coverage that provides higher margins; the challenges of conducting randomized postmarket clinical trials; and the effectiveness of devices in some, but not necessarily most or all, clinical settings. To address these reasons for inadequate regulatory response and better ensure that patients benefit from medical devices approved through special development pathways, we recommend that current expedited development or approval programs be contingent upon 1) timely progress of mandatory postmarket studies and 2) clinical data from these postmarket studies demonstrating that the threshold of reasonable assurance of safety and effectiveness is met for the primary endpoints. Until postmarket studies are completed, improved disclosure to patients is necessary to ensure they are able to provide informed consent.
16.1 Background
The availability of medical devices in the United States is overseen by the Food and Drug Administration (FDA), which evaluates new devices under a framework established by the 1976 Medical Device Amendments.Footnote 1 Under this law, devices are classified into three tiers, with the rigor of regulatory review commensurate with anticipated risk to patients. The highest-risk devices (Class III) are subject to the FDA’s most stringent review process, called Premarket Approval (PMA),Footnote 2 and are required to demonstrate a reasonable assurance of safety and effectiveness to receive marketing authorization. More flexible standards are applied to lower-tier devices (Class I and II), many of which are exempt from review altogether. Since the FDA Modernization Act of 1997, the FDA must consider the “least burdensome” means of evaluating medical devices, defined in FDA guidance as “the minimum amount of information necessary to adequately address a relevant regulatory question.”Footnote 3
While this regulatory framework has helped steward new devices that benefit patients onto the market, it has also allowed for the marketing of unsafe and ineffective medical devices, some of which have remained on the market for years. Even for devices subject to PMA, rigorous high-quality evidence is not necessarily required.Footnote 4 Studies have found low rates of randomization and blinding (i.e. allocation concealment among involved individuals) among clinical trials supporting approval of such devices.Footnote 5 Trials are often single-arm, with comparison to historical (instead of active) controls, which can lead to biased estimates of treatment effects.Footnote 6 Surrogate measures used in pivotal trials often do not translate to meaningful clinical improvements.Footnote 7 “Training patients,” which allow clinicians to gain experience using or implanting a device, are often excluded from reported clinical trial results, widening the gap between labeled efficacy and real-world effectiveness.Footnote 8 Trials often include small numbers of selected patients that may not represent the diversity of real-world patients, for example, due to the exclusion of older adults, women, or those with co-morbidities.Footnote 9 Trial followup is commonly short – an important limitation because many of these devices are permanently implanted, but safety concerns may not be apparent until years after approval.
Evidence limitations for 510(k) cleared devices are even greater.Footnote 10 This commonly used process is based on “substantial equivalence” to one or more predicate (i.e. previously available) medical devices.Footnote 11 Aware that the predicates on which equivalence was based had no requirement for safety or effectiveness, Congress recognized early on that the substantial equivalence requirement of the 510(k) clearance process did not provide full assurance of safety and effectiveness.Footnote 12 In 2011, the Institute of Medicine drew attention to this concern and recommended replacing the pathway,Footnote 13 which has been responsible for the highest proportion of medical device recalls.Footnote 14
Given the limitations in clinical evidence leading to uncertainties of benefit and risk at the time of approval, medical devices might be expected to undergo timely and rigorous postapproval evaluation. Yet only 54 out of 792 (or 7 percent) postapproval studies ordered between 1991 and 2020 were randomized clinical trials,Footnote 15 and of 28 PMA devices approved from 2010–2011, only 13 percent of 204 FDA-required or manufacturer/investigator-initiated postapproval studies were completed between three and five years after FDA approval.Footnote 16 Even eight to ten years after approval, only one-third were completed with final results reported on clinicaltrials.gov or in peer-reviewed publications.Footnote 17 The FDA has never issued a warning letter or penalty because of study delays or inadequate progress of a medical device postapproval study.Footnote 18
Limited pre and postmarket evidence can expose patients to unnecessary harm by allowing the availability of unsafe and/or ineffective medical devices. For example, in 2002 the Essure hysteroscopic sterilization device received premarket approval based on surrogate measures with short (up to two years) followup duration in just 926 women and was subsequently implanted in approximately 750,000 women. Postapproval studies were either not completed or terminated early.Footnote 19 Serious adverse events, including bleeding, pain, and unintended pregnancies were reported by thousands of women. The FDA responded by requiring new studies, and the device was eventually voluntarily removed from the market by its manufacturer in 2018 – sixteen years after FDA approval and just months after the Netflix documentary, The Bleeding Edge, documented the dangers of Essure and other currently used medical devices.Footnote 20
16.2 Increasing Uncertainty about Risks and Benefits of Marketed Medical Devices at the Time of Clearance or Approval
Despite the limited clinical evidence supporting medical device clearance and approvals, legislative mandates, such as the 2016 21st Century Cures Act’s codification of the Breakthrough Devices Program,Footnote 21 increase the potential for uncertainty of risks and benefits. These new flexibilities represent Congressional responses to concerns that device availability in the United States sometimes lags behind access abroad.Footnote 22
But new legislation has not been accompanied by rigorous eligibility requirements that would protect patients. For example, devices may qualify for Breakthrough status if “availability is in the best interest of patients,” providing the FDA with virtually unbounded discretion.Footnote 23 The agency has explicitly acknowledged that accelerating device approvals can reduce certainty of benefit. Agency guidance for the Breakthrough Devices Program, for example, states that the FDA “may accept a greater extent of uncertainty of the benefit-risk profile for these devices if appropriate under the circumstances.”Footnote 24 Devices approved through expedited pathways are more likely to be approved based on lower-quality evidence, such as trials that lack randomization or blinding, use surrogate measures, or are of limited duration.Footnote 25 A recent study of fifteen “breakthrough” devices found that two of these had been cleared under the 510(k) pathway,Footnote 26 a seemingly incongruous designation given that this pathway requires the 510(k) cleared device to be “substantially equivalent” to its previously marketed predicate. The paradox may be explained, if not necessarily justified, by the low and flexible bar to breakthrough designation and the generous definition of “substantial equivalence,” which encompasses devices with “significant changes” in materials, design, energy source, or other features as compared to the predicate, so long as they do not raise different questions of safety or effectiveness.Footnote 27
Due to the high costs of some devices, payor coverage must follow FDA authorization before widespread use is feasible. Payors can therefore serve as important gatekeepers against potentially unsafe or ineffective devices by restricting coverage until higher-quality evidence of benefit is generated. But payor oversight has been scaled back as well. Since late 2019, the Centers for Medicare and Medicaid Services (CMS) has been providing New Technology Add-On Payments for all FDA-designated Breakthrough Devices and increased reimbursement,Footnote 28 while waiving its longstanding (nineteen years) criterion that devices eligible for such add-on payments actually provide “substantial clinical improvement.”Footnote 29 Increasing reimbursement without high-quality evidence of patient benefit means that such data are not likely to ever be generated, as FDA approval and insurer coverage are strong incentives for conducting new high-quality trials.
The COVID-19 pandemic has accelerated the trend toward lower evidentiary thresholds. For example, in August 2020, the Impella® (Abiomed, Danvers, MA), a mechanical circulatory support device, received Emergency Use Authorization (EUA) for patients who experience complications while receiving extracorporeal membrane oxygenation,Footnote 30 despite limited established efficacy for this indication. EUA is a mechanism authorized by Congress in 2004 that allows widespread preapproval access for drugs or medical devices that “may be effective” in case of declared emergencies associated with chemical, biological, radiological, or nuclear threats. Another similar device, the Impella RP®, received EUA in June 2020 for patients with COVID-19-related right-sided heart failure.Footnote 31 Emergency use was authorized even though a May 2019 “Dear Doctor” letter advised that fewer than 30 percent of patients receiving the device in a postapproval trial for a different indication lived to thirty days, hospital discharge, or to the start of next longer term therapy (this proportion of real-world survival was much lower than in premarket clinical studies, which had demonstrated that 73.3 percent survived to thirty days, hospital discharge, or the start of longterm therapy).Footnote 32 It was subsequently determined that lower survival was among patients who would not have qualified for premarket clinical studies.
16.3 Lack of Regulatory Action for Unsafe Devices
While the FDA has the authority to revoke device approval, the agency has generally chosen to regulate with a lighter touch. In the rare cases when unsafe or ineffective devices have been removed from the market, manufacturers have done so voluntarily in the shadow of mandatory FDA recall authority, sometimes citing declining sales and possibly motivated by litigation concerns. The previously mentioned discontinuation of the Essure hysteroscopic sterilization device by its manufacturer in 2018 is one example.Footnote 33 In other cases, the FDA has imposed new evidence requirements that may have contributed to voluntary withdrawal. For example, after metal-on-metal orthopedic hips were found to have serious adverse events, including the release of metal ions into the bloodstream and adverse local tissue reactions that can lead to pain and device failure,Footnote 34 the FDA issued a final order in 2016 that required removal from market within ninety days if a PMA had not been filed for metal-on-metal hips marketed at that time.Footnote 35 All manufacturers have voluntarily stopped marketing these devices.Footnote 36
In other cases, the agency has not taken regulatory action even when studies with FDA involvement showed that the devices were associated with increased mortality. For example, paclitaxel-coated balloons and stents are sometimes used during endovascular intervention among patients with femoropopliteal peripheral artery disease. A meta-analysis using individual patient data, which followed FDA guidance and had a statistical analysis plan “based on formal discussions with the US Food and Drug Administration with review and approval by industry members,” found these devices were associated with a 4.6 percent absolute increase in in-hospital mortality compared to patients receiving standard balloon angioplasty.Footnote 37 The FDA concluded that additional clarification was needed,Footnote 38 but has not yet taken any regulatory action to restrict use.
The FDA has, at times, revised device labeling or recommended narrower indications in an effort to address safety issues while also minimizing disruptions to the market. For the Essure hysteroscopic sterilization device, the FDA promulgated guidance that included a “patient decision checklist” intended for both patient and physician signature that contained specific information about risks and benefitsFootnote 39. Measures such as this are intended to bolster informed consent so that patients are able to exercise appropriate autonomy when deciding whether to have the device implanted. The Wingspan intracranial stent system (Stryker Neurovascular, Kalamazoo, MI) received a Humanitarian Device Exemption approval by the FDA in 2005 based on a single-arm study that enrolled forty-five patients, with outcomes compared to historical controls at thirty days.Footnote 40 However, a subsequent randomized trial found that the Wingspan device had an increased risk of the composite endpoint of stroke or death in comparison to medical therapy.Footnote 41 Despite these findings, the FDA did not rescind the Humanitarian Device Exemption approval. Instead, the agency left the device on the market so that it would be available as an option for patients similar to those in the initial single-arm study of forty-five patients,Footnote 42 even though that study had significant limitations in rigor. Because the FDA does not regulate the practice of medicine, physicians can continue to use these medical devices off-label for patients who do not meet FDA-recommended criteria. To address safety concerns, the FDA issued a Safety Communication in 2019 (fourteen years after approval) warning of the increased risk of stroke or death when used outside of approved indications.Footnote 43
16.4 Managing Postapproval Safety of Devices
There are numerous reasons why it is challenging for regulators to reverse their decisions for approved medical devices, even in the face of mounting evidence that calls a device’s safety and effectiveness into question. First is a reluctance to engage in regulatory self-reversal. If devices that the FDA determined to meet statutory criteria are later found to be unsafe or ineffective, it can be uncomfortable for the agency to admit that its previous conclusion is no longer valid. This awkward situation can sometimes be avoided while still protecting patients by narrowing the scope of conditions or populations that fall within the labeled indication. Additionally, revoking approval also risks loss of public confidence in initial approvals, potentially deterring the use of unrelated beneficial treatments. The decision to narrow an indication is an acknowledgement that benefits are no longer believed to exceed risks for certain populations or indications and might logically be expected to lead to similar losses of public confidence, but modified labeling tends to draw less attention and is more likely to be perceived as a refinement rather than a reversal.
Similar psychology is at play with patients and physicians, who may have come to rely on the availability of a new device or who are reluctant to believe that a device that they implanted or that is implanted in them could actually do more harm than good. Research in the social sciences on loss aversion suggests that takebacks can be met with greater resistance than refraining from taking an action (in this case, clearing or approving a device) in the first place.Footnote 44 Patients may feel that a potentially beneficial therapy is being withheld from them if it is taken off the market. For physicians, intervention bias in medicine leads to the desire to “do something,” even if doing nothing may result in improved clinical outcomes.Footnote 45 Physicians may think that they are able to selectively use medical devices in patients who will derive clinical benefit, and professional societies may offer such guidance. However, there are important limitations in patients’ and physicians’ understanding of regulatory approvals, and they may not recognize that FDA approval still leaves important uncertainty.Footnote 46
To address this challenge, the FDA could publish guidance documents about benchmarks that must be achieved for a medical device to maintain approval after twelve months on the market. For example, the FDA could mandate that a postapproval clinical trial enroll a certain number of patients and meet specific safety and effectiveness endpoints to remain on the market. Devices that do not meet these parameters could then be withdrawn based upon prespecified, published criteria. As medical devices are often modified through PMA supplements,Footnote 47 or through the 510(k) pathway,Footnote 48 the expectation would be that all new device iterations would also meet these criteria.
A second challenge is that it can be difficult in the postmarket environment to generate high-quality data sufficient to demonstrate that earlier conclusions were wrong. Although randomized controlled trials remain the gold standard for clinical evidence, once a medical device is widely available, regulators rely primarily on observational data. For example, randomized clinical trials of patent foramen ovale occluders studying device ability to reduce the risk of stroke were delayed for several years because there was no incentive to enroll in a randomized trial when the devices were widely available off-trial.Footnote 49 Improved analytical tools have emerged to allow more reliable causal inference from observational data, such as propensity score matching,Footnote 50 instrumental variable analyses,Footnote 51 and the use of falsification hypotheses,Footnote 52 but more may be needed. As the granularity of data and the methods of analysis improve, confidence in observational results can be expected to increase.
Third, outcomes may improve as clinicians gain experience with both the device and its associated procedure, as studies show improved outcomes among patients who receive procedures at hospitals with higher versus lower procedural volume.Footnote 53 However, because training patients are often excluded from pivotal trial data, the “experience factor” has already been at least partially captured at the time of authorization. Making the data from the training patients available and included in premarket authorization would provide a more accurate assessment of expected initial outcomes in clinical practice,Footnote 54 and is necessary to allow patients and clinicians to make adequately informed decisions. Another possibility is for payors to limit reimbursement for certain medical devices to specific hospitals or physicians that have demonstrated expertise and successful outcomes. To protect patients, health systems could implement privileging requirements that require measurable demonstrations of proficiency with such devices, or medical specialty boards could authorize device- or device/procedure-specific certifications. In addition to these private efforts, Congress could expand existing Risk Evaluation and Mitigation Strategies programs to include devices as well as drugs.
A fourth reason is that devices may turn out to be unsafe or ineffective in some clinical circumstances, but still have benefits that outweigh their risks among other indications. For example, coronary stent placement has been shown to improve outcomes in the setting of patients with ST-segment elevation myocardial infarction. However, studies have shown that there is no benefit from coronary stent placement for patients with stable ischemic heart disease.Footnote 55 One way to address this scenario is to broaden use of patient decision checklists, such as with the Essure hysteroscopic sterilization device, to ensure that patients are adequately informed of the FDA-approved indications and the current status of data supporting safety and effectiveness prior to providing consent. Informed consent documents could also include clear FDA-required text, for example, that safety and effectiveness have not been demonstrated for particular indications.
Although the FDA formally has the authority to withdraw products when necessary to protect public health, regardless of manufacturer cooperation, it has rarely exercised this power. In one notorious case, the agency withdrew the metastatic breast cancer indication of bevacizumab (Avastin®) after a confirmatory trial failed to show a benefit in overall survival, leaving the drug itself on the market.Footnote 56 Even though the withdrawal was in reality only a labeling change, the FDA’s decision was extremely unpopular and faced substantial resistance from the manufacturer and public, which led to delays in its implementation despite the recommendation of the FDA’s Oncologic Drugs Advisory Panel.Footnote 57 CMS even stated that it would continue to cover the drug for the breast cancer indication. Use of bevacizumab decreased,Footnote 58 but the experience may have dissuaded the agency from taking similar regulatory actions in the future.
16.5 Managing Ongoing Postapproval Uncertainty of Evidence
The FDA’s growing enthusiasm for expedited approvals increases the need to rely on postapproval medical device studies to better characterize safety and effectiveness. However, such studies may not be completed in a timely manner (or at all).Footnote 59 Noncompletion of postmarket studies within requisite timeframes could also be a basis for revoking FDA approval to better protect public health.
Revoking approval based on lack of study completion is even more challenging than revoking approval based on trial results, since withdrawal for noncompletion of studies necessarily occurs in the absence of required study results and thereby allows hope and belief to override evidence-based practice. If devices are nevertheless withdrawn, patients and physicians may understandably be confused about the meaning of FDA approval: if more evidence was needed to demonstrate safety and effectiveness, then why was the device approved? Once devices are available on the market, generous payments for newer procedures can create a financial incentive for their use. Medical device manufacturers are likely to provide reasonable explanations for why clinical studies have been delayed, such as slow enrolment, and optimistically predict that confirmatory evidence will soon be available. In some cases, manufacturers may have incentives to delay postapproval trials, for example, if concerns remain that confirmatory trials will demonstrate a smaller effect size than in premarket data, or if visible enrolment efforts might engender a perception that a device’s benefit is uncertain.Footnote 60
To promote more timely development of evidence for the effectiveness of medical devices after expedited approval, Congress could ensure that devices have their expedited approvals automatically lapse if postapproval clinical trials are not completed or making adequate progress by FDA-imposed deadlines. For example, if a prespecified number of patients are not enrolled into a trial by a certain date, approval would lapse, and future potential patients would need to be enrolled in a clinical trial (as in a preapproval setting). Similarly, the FDA and other stakeholders would need to make clear through public messaging that timely postmarket evidence generation is necessary to prevent lapse of approval of a medical device. There is international precedent for similar regulatory action: in Japan, manufacturers of some devices must refile for approval with updated data from clinicians, clinical trials, and publications after a requisite time period to ensure that the data continue to demonstrate safety and effectiveness of the device.Footnote 61 If such a measure is implemented, it will be important to provide clear notice to patients of the limited evidence of benefits and risks to ensure that consent to treatment is truly informed.
17.1 Introduction
The Food and Drug Administration’s (FDA’s) strategic vision for monitoring high-risk medical devices emphasizes the role of postmarket registries, which are databases that actively collect and maintain information about individual patient exposures.Footnote 1 Registries are cost-effective relative to traditional clinical trials and can enroll large numbers of patients to provide generalizable observations and identification of rare safety events.Footnote 2 Although they differ in their structure, study goals, and stewardship – with varying involvement of professional societies, industry, academic centers, and regulators – registries in general may facilitate advancements in device use, manufacture, and design.
Registries are particularly useful for cardiovascular devices, which make up a large proportion of novel device approvals but also are commonly implicated in recalls and adverse event reports.Footnote 3 The FDA and the Centers for Medicare and Medicaid Services (CMS) can mandate postmarket registries as a condition of marketing approval or reimbursement, respectively. To help generate timely information on device safety and effectiveness, the FDA and the CMS sometimes require compulsory enrollment with no opt-out mechanism. Although regulators provide guidance and oversight on registry design and use, there has been little evaluation of the legal and ethical implications of compulsory medical device registries. In particular, questions remain regarding the extent to which compulsory registries accord with health privacy laws and ethical standards for human subjects research.
This chapter proceeds in three parts. First, we begin by discussing the emerging and integral role of registries in the FDA’s medical device postmarket repertoire, with a focus on cardiovascular medical devices. Second, we evaluate the applicability of the Health Insurance Portability and Accountability Act (HIPAA), the Common Rule, and state laws to compulsory registries. Third, we propose additional guidance for registry development, including rules for enrolment, consent, data use, and access to data.
17.2 The Role of Registries in Postmarket Analysis of Medical Devices
17.2.1 Limitations in the FDA’s Evaluation and Monitoring of Devices
The FDA employs a risk-based regulatory framework that classifies medical devices into three categories: Class I (low risk) devices are those that pose a minimal potential for harm, such as tongue depressors and stethoscopes; Class II (medium risk) devices have a higher potential for harm, such as syringes and electrocardiograph machines; and Class III (high risk) devices have the highest potential for harm, such as pacemakers and defibrillators.Footnote 4
All three classes of medical devices are subject to “general controls,” which include, inter alia, registration, prohibitions against misbranding and adulteration, and adherence to good manufacturing practices.Footnote 5 For Class I and Class II devices where general controls are insufficient to provide a reasonable assurance of safety and efficacy, “special controls” are required, which may include, inter alia, postmarket surveillance, patient registries, and 510(k) premarket notification.Footnote 6 For Class III devices where special controls are insufficient to provide a reasonable assurance of safety and efficacy, a premarket approval (PMA) application is required.Footnote 7
The 510(k) pathway principally seeks to establish that a new device is “substantially equivalent” to a device that the FDA has already cleared for marketing. As the FDA explains, the 510(k) pathway “is comparative” whereas the PMA pathway involves “an independent demonstration of safety and effectiveness.”Footnote 8 The 510(k) process “was specifically intended for devices with less need for scientific scrutiny, such as surgical gloves and hearing aids.”Footnote 9 Over the years, however, the breadth of devices eligible for the expedited review mechanism has been expanded significantly, and only 1 percent of medical devices utilize the more rigorous PMA pathway.Footnote 10
Apart from utilization of the 510(k) pathway for some high-risk devices, other high-risk devices come to market as PMA supplements – a subset available where a new device contains changes to an already approved device.Footnote 11 PMA Supplements may be required when changes impact the safety or effectiveness of a device, including but not limited to new device indications, labeling changes, use of new manufacturing processes or facilities, changes in sterilization procedures, packaging changes, or changes in design specifications or components.Footnote 12 For devices that come to market as PMA supplements, the FDA generally does not require clinical trial data.Footnote 13 In recent years, several high-risk cardiac devices approved as PMA supplements – some of which were implanted into hundreds of thousands of patients – have been recalled due to serious safety concerns.Footnote 14
We summarize the distinctions between the 510(k) and PMA pathways here to highlight the fact that it is common for high-risk medical devices to come to market without providing the FDA with clinical trial data that demonstrates the device’s safety and effectiveness. In part these accelerated pathways to market are due to budgetary constraints – specifically, Congress has not allocated sufficient funds so that regulators have the resources to oversee and review clinical trial data. A second relevant factor is that there are significant budgetary and scientific barriers to applying robust scrutiny to a large number of devices from conception through real-world utilization (often referred to as the “total product life cycle”).Footnote 15 In other words, the cost and time to provide meaningful safety and efficacy data would translate to longer periods of time before which a new device could come to market.
These resource constraints are exacerbated by statutory requirements that the FDA utilize “the least burdensome appropriate means of evaluating device effectiveness that would have a reasonable likelihood of resulting in approval.”Footnote 16 This legal requirement – which is not found in regulations governing FDA review of pharmaceuticals or vaccines – was enacted by Congress, largely at the request of lobbyists and medical device manufacturers.Footnote 17 It forces the FDA’s hands by requiring that the agency think creatively on how to solicit the least amount of information that can illustrate device safety and efficacy. As a practical matter it translates to device approvals that, for the most part, do not require clinical trial data. The least burdensome standard applies even for high-risk medical devices such as implantable cardioverter-defibrillators (ICDs), pacemakers, and artificial heart valves.
Several observers have highlighted limitations in the current legal and regulatory framework, particularly in premarket review.Footnote 18 These critiques also extend to the postmarket surveillance scheme which, despite evolving emphasis on new strategies,Footnote 19 continues to rely significantly on passive surveillance of marketed medical devices, a mechanism that fails to adequately capture postmarket safety and efficacy concerns.Footnote 20 While passive surveillance has been able to capture some instances of patient harms due to faulty devices, underreporting is widespread, and reports submitted to the FDA’s passive surveillance database are often submitted late and lack critical information on adverse events.Footnote 21 In instances where the FDA mandates postapproval studies, studies have found that progress is often inadequate and many requirements go uncompleted.Footnote 22 Inadequate postmarket surveillance is not limited to medical devices, but also plagues postmarket evaluation of pharmaceuticals and vaccines.Footnote 23 For truly novel, transformative, and influential therapeutics, then, a robust postmarket surveillance strategy is of great importance to regulators, payors, and the public because it helps produce meaningful evidence to continuously evaluate the safety and efficacy of marketed medical products.
17.2.2 General Structure and Function of Regulatory Registries
When structured and utilized properly, registries can provide valuable information to support postmarket analysis on safety and efficacy. As noted above, the FDA can mandate registries either as a condition of approval for high-risk device (a so-called postapproval study) or as a “522 study,” which can be applied at any point in a product lifecycle.Footnote 24 Timely completion of these studies is the responsibility of device sponsors and, in theory, the FDA can withdraw marketing approval or clearance for failure to do so.
Registries defined by exposure to a specific device or procedure can generate datasets with large sample sizes that include a more diverse set of patients than those in premarket studies. Registries can include or be linked to additional clinical data, which allows for identification of information related to disease severity and comorbidities and may provide information on the device utilization outside the context of pivotal clinical trials or established guidelines. For example, studies have uncovered divergence from guidelines-based indications for ICDs and cardiac resynchronization therapy.Footnote 25 Registries can provide important insights regarding off-label use of devices, such as transcatheter aortic valve replacement (TAVR), that may guide future regulatory decisions about expanded indications.Footnote 26
Registries also play an important role in coverage decisions and subsequent requirements for evidence generation. Once FDA approval is earned, sponsors of new devices typically submit applications to the CMS to determine whether the product meets the statutory requirement of “reasonable and necessary” for reimbursement.Footnote 27 Both terms remain somewhat nebulous but together are generally understood to reflect a totality of evidence supportive of clinically meaningful benefits with an acceptable safety profile.Footnote 28
While many services (including use of new devices) are covered by the CMS automatically, in select cases, manufacturers, clinicians, or the CMS request a national coverage determination, which grants, limits, or excludes Medicare coverage nationwide.Footnote 29 A small proportion of services thought to be particularly novel, influential for Medicare beneficiaries, or otherwise identified as important from the CMS’s perspective are provided conditional reimbursement – “coverage with evidence development.”Footnote 30 In these cases, payment for services occurs only in concert with a prospective study approved by the CMS as meeting specific scientific goals relevant to safety, effectiveness, or utilization among its beneficiaries. Over the past fifteen years, more than two dozen devices or services have been subject to coverage with evidence development decisions. This includes truly novel and (for Medicare patients in particular, most of whom are aged greater than sixty-five) clinically impactful transcatheter treatments for valvular heart disease, devices for stroke prevention, and new “leadless” designs for implantable pacemakers.
17.2.3 Compulsory Registries for Cardiovascular Devices
FDA review and CMS reimbursement have brought together agencies with overlapping public health mandates to help establish several pivotal cardiovascular devices registries.Footnote 31 While the individual details and methods vary, in general these registries have met the needs of regulatory agencies to develop additional evidence specific to its intended patient population, while also providing a platform for postmarket surveillance studies assessing safety, off-label utilization, real-world outcomes, and potential expansion of indications. The exact purpose, structure, and stewardship of “regulatory registries” – that is, those created primarily to meet requirements of the FDA, CMS, or both – varies according to device. Here we describe two influential cardiovascular device regulatory registries that share the feature of compulsory enrollment.
The National Cardiovascular Data Registry (NCDR) ICD Registry was created in 2005 in concert with expansion of CMS coverage guidelines for primary prevention ICDs, which are ICDs implanted in patients without a history of cardiac arrest or sustained ventricular arrhythmias.Footnote 32 A clinical trial published in 2004 demonstrated a survival advantage for ICD implantation in patients with heart failure from left ventricular systolic dysfunction regardless of etiology, widely expanding the pool of patients eligible for an effective but expensive intervention.Footnote 33 The ICD Registry was developed by the American College of Cardiology (ACC), which manages a suite of registries under the NCDR umbrella, and the Heart Rhythm Society (HRS), a professional society for cardiac electrophysiology, with guidance from the CMS and FDA. Notably, the CMS coverage memo requires only that data be collected for Medicare beneficiaries. However, the majority of the approximately 1,500 participating sites submit data on all patients who receive ICD implants, and thus the ICD Registry serves as an excellent storehouse of postmarket information.
Several specific analytic questions were posed by the CMS as the guiding scientific goals for the ICD Registry. The overall principle was summarized in the original 2005 memo from the CMS, which indicated: “We are concerned that the available evidence does not provide a high degree of guidance to providers to target these devices to patients who will clearly derive benefit.”Footnote 34 Specific hypotheses posited to refine that position through the ICD Registry include Table 17.1:Footnote 35
Table 17.1 Hypotheses
|
Data collection is performed for over 100 data elements incorporating patient characteristics, procedural details, laboratory tests, and complications that occur within the index hospitalization. These data include multiple individual identifiers, which have facilitated linkages to other datasets such as administrative claims data as well as industry data.Footnote 36 Over one million patients have had data entered into the registry, including hundreds of thousands of patients who are not Medicare beneficiaries. There is no consent obtained and no mechanism for patients to opt-out or to view their own data. Of note, an updated Medicare coverage memo issued in 2018 ended the requirement for entry into the ICD registry as a condition of reimbursement.Footnote 37 Data from the ICD Registry has been relied upon in several publications that have analyzed safety and efficacy of ICDs, though the impact of the ICD Registry on CMS reimbursement has been less clear.
Similar motivation supports the Transvalvular Therapeutics (TVT) Registry, a partnership between the Society of Thoracic Surgeons (STS) and ACC that has been approved by the CMS to meet coverage requirements related to TAVR and transcatheter mitral valve repair. These two device types have been transformative therapies over the past several years, bringing minimally invasive options to patients previously considered prohibitive or high risk for surgical intervention and increasingly extending towards wider populations of potential recipients. The FDA worked with the CMS to structure the registry. The CMS coverage memo for TAVR echoed elements of that issued for ICDs, including the following specifications (among others) and articulated study goals Table 17.2:Footnote 38
Table 17.2 Specifications and study goals
|
Again, patients must be enrolled, or the facility risks nonreimbursement. In practice, this means that device recipients are automatically enrolled without consent or an opt-out mechanism. Notably, the ICD registry case report forms can generally be completed entirely from electronic or similar data sources, without the need to speak with patients. The TVT Registry form includes many of the same demographic, clinical, and procedural details as the ICD registry but also captures quality of life information. These additional data points require a brief interview with patients.
17.3 Legal Framework Governing Compulsory Medical Device Registries
There is no uniform legal framework applicable to all registries. Rather, the reach of the law – including health privacy laws and regulations governing research with human subjects – depends on the structure and function of a registry, as well as the registry steward. This is problematic, since a wide range of stakeholders creates and uses registries, including academic medical centers, not-for-profit entities, professional societies and organizations, private companies, health care payors, provider organizations, and medical device companies.Footnote 39 Divergent protections can result in use of health data in ways that contradict the expectations or interests of patients, which may exacerbate lack of trust in data use and the health care system.
17.3.1 The Scope of HIPAA Protections for Registry Data
HIPAA protections apply solely to covered entities (i.e., health care providers, health plans, and health care clearinghouses) and the business associates of these entities.Footnote 40 Several registry stewards fall outside of HIPAA’s reach entirely, so long as they do not collaborate with a covered entity, including medical device companies, patient advocacy groups, and professional societies. Registry data submitted directly from a patient to a registry steward is also not encompassed by HIPAA’s protections.Footnote 41 And, HIPAA’s limitations apply solely to protected health information, not to the collection and use of deidentified data.Footnote 42
For entities that fall under the HIPAA umbrella, the HIPAA security rule requires implementation of a reasonable security plan and security risk assessments.Footnote 43 In addition to the protections mandated under the HIPAA security rule, the HIPAA privacy rule affords protections to individuals whose health information is handled by an entity bound by HIPAA. The privacy rule requires patient authorization if health information is to be used in research, but authorization is not required if the information is to be used for public health activities.Footnote 44 Via this exception, patient authorization is not necessary for public health surveillance registries that do not include research.Footnote 45 This includes registries created to track the quality, safety, or effectiveness of FDA-regulated products.Footnote 46 Overall, HIPAA allows covered entities and their business associates to disclose identifiable patient information without patient authorization in cases where a registry: 1) furthers public health activities, including public health surveillance and review of an FDA-regulated device; 2) supports health care operations; or 3) is created pursuant to a legal mandate of health oversight officials, such as for CMS reimbursement.
If public health research is conducted using registry data assembled for public health practice, HIPAA permits disclosure of identifiable patient information without consent for a limited dataset, so long as an institutional review board (IRB) or privacy board issues a waiver of consent and the data source and registry steward enter into a data-use agreement.Footnote 47 In considering whether a waiver of consent is appropriate, relevant factors include whether 1) the research involves more than minimal risk, 2) adequate data protections are in place, 3) the research could not practically be conducted if patient authorization is required, and 4) the research could not practically be conducted without identifiable information.Footnote 48 Notably, a limited dataset cannot contain certain data points, such as names, device identifiers, and biometric identifiers; accordingly, limited datasets may be of diminished relevance to device registries, and particularly for cardiac device registries where device and biometric identifiers are essential.
Under HIPAA, patient authorization is also not required for health care treatment, payment processing, or health care operations.Footnote 49 Accordingly, registries used solely to tailor treatments for patients would not need patient authorization, nor would registries that facilitate health care quality improvement, outcomes evaluation, and development of clinical guidelines.Footnote 50 This includes registries created by hospitals or health care providers to track patient outcomes against clinical care standards.Footnote 51
Taken together, HIPAA allows covered entities and their business associates to disclose identifiable patient information without patient authorization in cases where a registry: furthers public health activities, including public health surveillance and review of an FDA-regulated device; supports health care operations; or is created pursuant to a legal mandate of health oversight officials, such as for CMS reimbursement. The public health surveillance exception is particularly relevant in the context of compulsory regulatory registries. Also relevant is the exception whereby identifiable patient data can be disclosed for research purposes if the research could not reasonably be achieved if patient authorization is required. As to the latter, such an argument in the context of a compulsory registry may not withstand scrutiny in cases where direct patient contact in a clinical setting could be expanded to include, for example, verbal or written consent to use of patient data in a registry. The lack of a uniform legal framework to apply across all medical device registries leaves registry stewards to act on an ad hoc basis, which may lead to inconsistent protections across the population.
17.3.2 Applicability of the Common Rule to Registries
In instances involving research based on registry information, federal protections governing research with human participants may apply. As a threshold matter, the Common Rule applies to 1) federally funded research sponsored by one of the seventeen federal agencies that have adopted the Common Rule or 2) studies that will be submitted to the FDA in the context of device approval or monitoring. Some institutions – such as academic medical centers – have adopted the Common Rule to all research conducted at the institution, regardless of funding source. Given the breadth of registry stewards, however, there may be instances where a registry steward or data user is not legally bound by the Common Rule. In such instances the steward or data user has the discretion as to whether, and to what extent, to follow the federal guidelines.
The Common Rule’s protections apply solely to research, which is defined as a systematic investigation that is designed or developed to contribute to generalizable knowledge.Footnote 52 At the outset, it is important to note that the Common Rule does not apply to registries that do not include individually-identifiable information.Footnote 53 Moreover, under the statute, research does not include public health surveillance and the provision of health care.Footnote 54 These exceptions are particularly relevant in the context of regulatory registries, since registries are often created to monitor public health or comply with FDA postmarket requirements.
At the same time, if identifiable information is used for public health research – rather than public health surveillance – the Common Rule would apply and patient consent would be required, unless an IRB or privacy board determines that a waiver of consent is applicable.Footnote 55 Along these lines, the Common Rule’s protections apply to registry research in the context of an FDA-regulated device; as with public health research, the Common Rule would apply and patient consent would be required, unless an IRB or privacy board determines that a waiver of consent is applicable.Footnote 56
For registries that fall within the purview of the Common Rule, regulations require that the registry steward and registry data user obtain informed consent from identifiable individuals who are included in the registry.Footnote 57 A waiver of informed consent may apply if the research poses a minimal risk to the research subjects, cannot be practically conducted without a waiver, does not use registry data in identifiable form, and will not adversely affect the rights and welfare of the research subjects.Footnote 58
In instances where informed consent is required, the research participant must be informed of the risks and benefits of the research. This includes information related to privacy protections and the risks of loss of confidentiality.Footnote 59 However, pursuant to revisions to the Common Rule enacted in 2016, “broad consent” is now permitted in instances where researchers are conducting downstream research using identifiable personal information. Under the broad consent principle, at the point of initial consent, all that is required is a general description of the type of research that may be conducted, the identifiable information that may be used, timeframe for research, any plans to share information, and contact information for the researchers.Footnote 60 Thus, at the time of initial collection, the registry steward can utilize a broad consent document that covers future uses of the patient’s information which, as a practical matter, provides little guidance to the patient on how, precisely, their information will be utilized.Footnote 61
The Office for Human Research Protections explains that primary and secondary purposes of an activity are relevant factors to consider in determining whether a project qualifies as research under the Common Rule.Footnote 62 As such, registries created for research purposes, in whole or in part, would fall under the Common Rule if the entity creating the registry is bound by the Common Rule’s protections.Footnote 63 This is distinct from the HIPAA privacy rule, which indicates that the protections apply only if research is the primary purpose behind use of patient information; otherwise, HIPAA classifies the data use as health care operations.
IRB review of the registry protocol would include the research purpose of the registry, informed consent arrangements (or an explanation of why informed consent is not necessary), and privacy and confidentiality safeguards.Footnote 64 Compulsory registries that are required by law fall under the Common Rule’s umbrella only if the registry is used for research. In such cases, consent would be required unless a waiver has been authorized by a governing IRB.Footnote 65 Taken together, although the Common Rule affords protections for research that utilizes registry data, registry stewards and downstream data users must be mindful of the ethical implications of consent waivers and other exceptions to the research guidelines. Just because use of registry data without patient consent may be legal, it does not mean that such use is ethically appropriate.
17.3.3 Additional Laws
Apart from HIPAA and the Common Rule, we also note briefly that several other laws may apply to the creation and use of registries. The additional laws include federal statutes, state statutes, state common law, and, in the case of registries incorporating data derived from patients outside the United States, laws from other nations. For example, coupled with the Common Rule’s application to research involving registries, there are supplemental federal protections and guidelines for research involving prisoners, pregnant women, children, and patients in federally funded substance abuse programs. In addition, the NIH can issue a certificate of confidentiality for a specific project that requires confidentiality beyond the general legal requirements.Footnote 66
Also relevant is the Federal Trade Commission (FTC) Act, which prohibits unfair or deceptive trade practices.Footnote 67 Registries fall within the FTC Act’s reach, and it would be a deceptive trade practice to provide individuals with false or misleading information regarding data collection or use.Footnote 68 State laws, such as California’s Consumer Privacy Act, may also dictate rights to bearing on registry design, as will the laws of other nations, such as the European Union’s General Data Protection Regulation, if data are collected from or shared within its jurisdiction.
17.4 Proposed Guidelines for Development and Use of Compulsory Registries
The benefits of compulsory registries are tangible and significant. In light of the significant evidence gaps in premarket review, we believe that the potential benefits to be gained from compulsory registries likely outweigh the risks to participating subjects. Yet we are mindful of the implications of compulsory registries on patient autonomy and respect for persons, and thus recommend that a robust informational-disclosure dialogue be implemented as a component of informed consent for clinical care where enrollment in a registry is a requirement for use of the medical device.
Insofar as all patients will have procedural consent obtained prior to device implant, there is an existing mechanism for clinical contact. Moreover, some registries already incorporate detailed patient interviews purely for research purposes, such as the collection of patient-reported outcome measures for key variables such as quality of life. Incorporating a verbal or written consent into the clinical point of care or patient interview would not pose an unreasonable burden on physicians or investigators.
In addition to robust consent and data-linkage protocols, we likewise recommend that registry stewards and data users be held legally accountable for maintaining the security of patient data and providing patients with clear information on data collection and use. This includes providing detailed information on registry sponsorship and specific uses of registry data prior to patient authorization to clinical use of a device. Accountability also includes a privacy-by-design feature whereby registry stewards must affirmatively obtain consent from patients if patient data is to be used beyond the original scope, allowing patients to opt-out of such downstream uses. To further accountability, patients should have easy access to a tracking system that details data use and downstream research.
To promote public trust in compulsory registries, stewards should task a standing advisory committee to track operational and ethical issues. At least one member of the committee should be trained as an ethicist and not have a relationship with the registry steward or downstream data users. The committee should also be a forum whereby patients can raise questions or concerns about the registry. To the extent these criteria are met by an existing institutional review board, there may not be a need to create a separate committee.
17.5 Conclusion
Compulsory registries promote patient outcomes and facilitate robust lifecycle analysis of medical devices. Insofar as laws and regulations have significant gaps in instances where patient authorization is required prior to collection and use of patient data, providers and registry stewards have an ethical obligation to inform patients about data collection and use. Contemporary data protection and research laws afford limited protections for individuals, but these existing laws need not dictate ethical guidance. This is particularly true in the context of health information, which is widely viewed as one of the most sensitive informational areas. Instilling supplemental privacy safeguards and data-use limits may be appropriate when patients are compelled to include their personal information into a registry as a condition of receiving medical care.
18.1 Introduction
Medicine has a longstanding compact with society to prioritize the needs of patients above other aims, a concept as old as the Hippocratic Oath. As an extension of this professional commitment, physicians have been granted the discretion and authority to police their own – to establish professional standards of conduct and to enforce those standards in the practice of medicine against fellow professionals.
Peer review is the culmination of this social compact. Peer review is a decentralized process in which the formal medical staff structures across the country assess adverse medical events, determine whether a colleague’s conduct fell short of a professional standard of care, and, when necessary, discipline errant physicians. It is led by physicians, and though the objective is to learn from all errors and improve care throughout the delivery system, its primary focus is on physician conduct. It is so central to the modern practice of medicine that it has shaped the organization of hospitals and embodies the central values of being a medical professional.
However, the contours of peer review are products of – and are thus sensitive to – medical technology. Physician self-regulation and its prevailing current structures are largely driven by the realities of knowledge. Historically, it was held that only physicians had the expertise to distinguish proper from improper care, and thus should be the sole authorities to assess the competence of any individual member of the profession.Footnote 1 But modern medicine is now guided by sources of knowledge that no longer lie in the sole possession of physicians. Shifts toward team-based care models rely on the expertise of nonphysicians and skills that lie beyond physician capabilities. Electronic medical records are now the primary repositories of patient medical information and are becoming more capable than physicians at detecting and correcting errors or deviations in practice. And digital technologies increasingly have the capacity to synthesize data to generate diagnoses and medical recommendations, which compare favorably to human experts. These new technologies and systems are challenging the supremacy of physician expertise in medicine, and consequently are eroding the underlying justifications for peer review.
And true disruption might be at hand. Artificial intelligence (AI) and machine learning (ML) algorithms are slowly enabling computer-driven medicine to perform the core functions of peer review: identifying medical errors, learning from adverse outcomes, and instituting reforms.
Even though it is widely known that new capabilities and skills from nonphysicians are playing an increasingly significant role in the practice of medicine, there has been little thought to whether these technological disruptions will necessitate broader changes in the organization and delivery of medicine. This chapter begins that inquiry by exploring the implications that AI and digital technologies have for peer review. We suggest that these technologies do much more than supplement the medical staff’s ability to evaluate and improve medical practice. They have the capacity to disrupt traditional centers of authority that underlie peer review and may thereby force a reorganization of medicine and a recalibration of health care regulation.
18.2 Peer Review Explained
The governance relationship between so-called learned professions and the rest of society has been likened to a social contract. In exchange for professionals investing in valuable expertise, inculcating a commitment to, and offering guidance to state officials, the state defers to professional expertise in both substance and practice.Footnote 2 Professionals thus are tasked by the state to define standards of conduct and to discipline themselves accordingly. Economists typically describe this arrangement as a product of information asymmetries – that lay people, even elected officials, have inadequate knowledge to scrutinize the conduct of scientific expertsFootnote 3 – and sociologists observe that this social arrangement reserves for professionals a privileged status and fierce autonomy that few other laborers enjoy.
Physicians historically have been the archetype of the learned professions, enjoying greater autonomy and self-governance privileges than that afforded to other professional societies, and the expanse of physician self-governance is reflected across numerous public and private mechanisms. States authorize medical societies to establish state licensure regimes and malpractice standards, thereby allowing the profession to define qualifications and minimal standards. States also provide their plenary powers to both licensure boards (thereby prohibiting nonphysicians from engaging in “the practice of medicine”) and courts (thereby disciplining those who fail to meet medical board standards). The incorporation of professional standards into the law, and the use of the state’s police power to enforce those standards, allow medical professionals to maintain a robust and self-sustaining system of self-regulation.
But it might be said that the cornerstone of physician self-governance – and the pinnacle of collaboration between medical professionals and the state – lies in peer review. Both medical boards and state authority defer to institutional peer reviews, and numerous laws cloak the peer review process both to secure its sanctity and reinforce its authority.
This legal arrangement has been described as a unique reflection of successive “devolutions” of medical authority, from federal government to state government, from state government to physician-led state licensing boards, and from these boards to the local medical staff structures within individual hospitals.Footnote 4 In many respects, it is a historical compromise reflecting the need to protect the public from deviant medical practice, while recognizing that medicine is an imperfect science with significant regional variation. It is also a means to achieve the end of public safety while avoiding nonphysician (government) intrusion in the doctor-patient relationship, a politically challenging issue.
While voluntary and informal review of physician practice by peers may occur in a number of settings, formal peer review as considered in this chapter is largely a hospital-based function, conducted to meet accreditation requirements specified by the Joint Commission (TJC). Despite some guidance put forward by TJC for medical staff practice, peer review remains highly variable in its structure and process, with relatively little standardization across the country. Peer review is typically conducted by a multi-disciplinary committee of physicians, and supported by hospital personnel, including nurses, safety experts, and attorneys. The peer review committee reports to a physician-led medical executive committee (MEC), which has local responsibility for the oversight of medical practice at a hospital. The committee may serve a number of roles, including initial review of physician qualifications for inclusion on the medical staff. But the committee primarily reviews adverse patient outcomes for evidence of physician negligence or incompetence, typically comparing the actions of the responsible physician to a community standard, looking for gross deviation from usual practice. In some cases, the committee will undertake an exercise, known as root cause analysis, meant to uncover the specific factors which resulted in the adverse patient outcome. The work product of the committee may range from recommendations to the hospital’s governing body (via the MEC) to rescind physician privileges, to dismissal of any concern related to the adverse incident.
Peer review is much more than a political compromise or a social compact of convenience. There are benefits that justify empowering local inquiry, by and for professionals, in assessing medical error. First, because of the complexity of medicine, inquiries into errors are best done by physicians most familiar with the context in which the error took place. In addition to the natural complexity of medicine, which justifies deferring to the expertise of practitioners over the judgment of regulators, there is further reason to defer to physicians who are best acquainted with the surrounding environment, facilities, and personnel in which scrutinized care took place. For this reason, many applications of medical malpractice law recognize regional variations in the practice of medicine and apply locally determined standards of care.
Second, and more significant, local peer review is designed to enhance the benefits of scrutinizing adverse outcomes. Since the ultimate objective is to learn from mistakes and improve the quality of care, the priority of any review process is to acquire accurate information, which plausibly is best done between colleagues within a cloak of trust, reciprocity, and collective learning. The tacit nature of information, the sensitivity of disclosing information related to potential errors, and the formal and informal support structures that embed the disclosure of this information all counsel towards providing discretion and authority to local medical boards.
These purported benefits of peer review are also reflected in the law, as many doctrines explicitly protect the peer review process. For example, any materials generated during the peer review process, including admissions of error, are shielded from discovery in any subsequent malpractice suit. Moreover, if internal documents or materials related to a medical error are not part of a peer review, they then do become subject to discovery. Perhaps most important, the association of peer review with high-quality medicine is enshrined in state licensure law and accreditation standards. A hospital needs to institute a peer review process, administered by a physician-led medical board, to be permitted to care for patients and receive Medicare funds. Peer review is not just thought to be important for maintaining high-quality medical care, it is deemed to be an essential feature of quality assurance for medical facilities.
18.3 Criticisms and Shortcomings of Peer Review
Peer review is not without its detractors, however, both within and outside the field of medicine. There are three broad categories of criticism: 1) It promotes a singular societal aim (safety) over other health care ends of importance to the public (like innovation, cost, and access); 2) Like all human processes, it is liable to bias, self-preservation, and abuse; 3) It is ineffective in achieving its primary aim of promoting safe care.
Perhaps owing to the ancient dictum, primum non nocere, peer review almost solely focuses on safety of health care to the exclusion of other valuable aims. While nominally, peer review purports to drive learning and improvement, reviews of its success in this area have been disappointing.Footnote 5 This is, in part, inherent to the nature of the process – the widespread dissemination of lessons learned (needed for innovation) stands in opposition to the privacy needed to maintain a trusting arrangement for local review. Furthermore, peer review has limited impact on other aspects of the “iron triangle” of health care, including cost containment and access.Footnote 6 Given the increasing importance of these parameters in promoting both affordability and public health, the local self-regulatory nature of peer review may be insufficient.
The process of peer review itself has been criticized as being biased,Footnote 7 with concerns of both underreporting and overreporting of poor physician practice. Peer review requires physician colleagues to assess each other’s work product (patient care) in a reciprocal manner. Despite best intentions, policing one’s peers can prove difficult, with a variety of conflicts of interest and social connections serving as barriers to even the most well-meaning and thoughtful peer review committees. Underreporting is an unsurprising result of a process in which unpaid physician committee members are asked to make potential career-ending calls on classmates, friends, and patient-referral sources.Footnote 8 On the other hand, peer review has also been used in an overaggressive manner, as one group of physicians attempts to drive out a competing member of another group. Indeed, this anticompetitive practice (sometimes known as “sham” peer review) was the basis for a $2.2 million settlement for the plaintiff in the 1986 antitrust lawsuit brought by Dr. Timothy Patrick against Dr. William Burget and the Astoria Clinic. This settlement sent a strong message to curtail the practice of sham peer review, but further dampened physician interest in participating in peer review, for fear of legal downsides. This led to the passage of the Health Care Quality Improvement Act (HCQIA), which attempted to rectify this by providing legal immunity to those physicians who participate in peer review in good faith.
The efficacy of peer review in evaluating the adequacy of medical care has also come under criticism. Despite the push toward evidence-based medicine, large swaths of care remain outside the evidence-based map, leaving ample room for variation in practice. In addition, many medical errors are the result of cognitive biases that are challenging to elucidate and difficult to mitigate.Footnote 9 Indeed, this point highlights the real challenge of the peer review process. To work well, peer review requires: self-reflection and insight, accurate recollection, capacity for thoughtful interrogation, cooperativity, and clear communication within a trustworthy circle of colleagues. Deficiencies in any aspect of this set of conditions may limit the adequacy of the evaluation and diminish its impact of provision of safe care.
18.4 The Promise and Challenges of Artificial Intelligence (AI)
Against this backdrop, a new era of intelligent tools, including advanced decision support with artificial intelligence, pose an attractive alternative to the peer review process. Artificial intelligence (AI) and machine learning (ML) are predictive modeling approaches which combine data in unique ways, via algorithms, to identify optimal solutions.Footnote 10 With improvements in programming, massive increases in computing power, and digitization of nearly all aspects of health care delivery, AI/ML has achieved a series of impressive results, now surpassing human capabilities in many domains, including image recognition. The latest iteration of this type of technology, known as deep learning, shows the capacity for perpetual enhancement in accuracy, self-modifying the algorithms it uses based on the relation of outputs to inputted data.
There is significant enthusiasm for the use of AI/ML within health care. The myriad of potential applications currently fall largely into two domains: image analysis and clinical decision support. The former application has multiple use-cases within health care, from visual analysis of radiographic images,Footnote 11 to pathologic tissue diagnosis,Footnote 12 to interpretation of retinal scans.Footnote 13 These solutions are focused primarily on the diagnostic aspect of health care – distinguishing normal from abnormal and applying the taxonomy of human pathology to abnormal findings.
AI is also being explored to assist with the cognitive decision making so critical to the work of many physicians. Taking disparate bits of information from multiple sources (e.g., patient history, physical exam, diagnostic tests) and determining a diagnosis, prognosis, and therapeutic plan. The range of potential decision-support applications is as broad as the expanse of all of medical practice, and includes recent examples, including early warning prediction of intraoperative hypotension,Footnote 14 mortality after heart failure,Footnote 15 and suicidal behavior after hospital discharge.Footnote 16 Increasingly, applications are moving beyond diagnosis or prediction, to autonomously enacting therapeutic decisions. For example, closed-loop neurostimulatory devices are being designed to both identify epileptic signals in the brain and immediately treat them via electrical stimulation.Footnote 17 Chatbots powered by AI are also now being deployed, to field patient chief complaints and triage them, in some cases recommending basic treatments.Footnote 18
Despite this potential, AI and other digital diagnostics are largely kept out of the peer review process. One reason for this is the traditional structure of peer review. Peer review is focused on physician conduct, more than systems-centered care, and is by design reactive to significant adverse outcomes. Of the many errant and potentially harmful actions that take place regularly in a large and complex hospital, very few attract the attention of peer review scrutiny. Those that do are examined by fellow physicians through traditional means of physician judgment, without the deep computational support that is more commonplace in advanced analytics.
Another reason is that digital monitoring, the foundation on which AI technology is built, has developed around a quality assurance infrastructure that is largely parallel to physician-centric peer review. Modern hospitals institute software that monitors the administration of patientcare that contain safeguards against predictable errors. For example, when medication orders are entered into most hospital computer systems, software reviews these orders for obvious deviations or harmful interactions. This electronic monitoring occurs downstream of most physician conduct and has the capacity to scrutinize daily conduct that usually is outside the domain of peer review.
The real-time digital quality assurance alerts and the retrospective human-led peer review represent largely parallel solutions to improving patient safety, both required for Joint Commission accreditation. Even now, digital monitoring exhibits capabilities that reach much deeper and more objectively into the practice of medicine. Whereas peer review examines perhaps a few dozen adverse outcomes per month, order alert software monitors thousands of inputs daily. Moreover, these software systems can collate and analyze these events to identify common sources of error.
The superimposition of AI offers enormous opportunity for deeper analysis and more sophisticated monitoring that could improve quality assurance. First, AI and data analytics could do more than provide simple alerts to known errors. Deep learning algorithms could examine population health data and tailor recommendations to an individual patient’s needs, identify improvements to accepted medical protocols, or anticipate systemic sources of provider error. Second, AI algorithms could institute real-time guidance to providers at the point of care, both anticipating moments of likely errors and interjecting with medical treatments that data analysis determines is superior to a human’s judgment. Perhaps ultimately, such deep learning algorithms could remove the role of human judgment and human implementation altogether. AI could, on its own, identify and implement a tailored treatment, assess preliminary results, change course to alternatives if necessary, and integrate data from a patient’s progress to a broader body of knowledge. The role of humans would be to peripherally monitor the AI’s implementation, perhaps play some supervisory role, and ensure that the algorithms have the data and resources they need to operate effectively.
18.5 Peer Review and Artificial Intelligence
The very aspects of AI that make it an exciting adjunct or alternative within health care are those that pose the greatest challenges to peer review. As much as software and digital capabilities can enhance and improve the quality assurance mechanisms that culminate in peer review, those same technologies might also prove to undermine peer review.
The self-learning, black-box nature of AI makes it difficult to interrogate to human knowledge, even if one has expertise in computer technology. As AI puts together information in novel and unique ways, the resulting algorithms may in fact be more accurate. But this reliance on alternative models makes it impenetrable to the physician review process which is framed by longstanding medical Western Medical tradition of the factors to be incorporated in diagnosis, prognosis, and treatment. Physicians simply do not have an understanding of the algorithms used to generate these decisions; decisions that may be swayed by information a physician may not have considered prima facie relevant. Whereas peer review is designed to be by physicians and for physicians, the skillset required to monitor AI-guided medicine would more likely involve computer scientists and software engineers.
In addition, since AI may identify patterns in existing data that will guide treatments where there is current clinical equipoise, the best course of treatment may no longer be determined solely on the basis of published literature. A peer review committee determining if an error occurred will have less information than the machine that directed treatment. Since AI will constantly be at the forefront of medical practice, any human oversight of AI and deep learning algorithms will be, by definition, deficient.
Therefore, however helpful AI might be, it is critical to recognize it contains qualities that are starkly different from human physicians (to state the obvious, machines are different from humans). In some ways, AI is like an uncooperative physician, one whose logic is impenetrable, and speaks another language entirely. To the degree that this new member of the medical staff is guiding care which may in some cases be harmful, it will fall upon local peer review processes to mitigate their impact on other patients. It is uncertain how the traditional approach to reciprocal peer review can effectively manage this task.
18.6 Implementing AI into Hospital Quality Assurance
In short, while the hype and promise surrounding AI is clear, it is also clear that health care AI will not readily fit into existing hospital governance. The oversight of AI will depend heavily on the nature of its use, whether AI is a tool used by providers or whether it becomes a provider in and of itself.Footnote 19
The introduction of AI processes could be maintained and modified to become safely incorporated into the current delivery system. For example, local medical staff policies could be implemented to limit the use of AI, by precluding its use in making immediate patient-facing decisions or requiring physician signoff on all diagnostic or therapeutic decisions. In so doing, human physicians would retain their current authority over medical care and buffer patients from harm related to aberrant AI decisions. The medical staff could also monitor the novel use of AI for a period of time until deemed safe, akin to the Joint Commission-mandated initial focused professional peer evaluations (FPPE) used to evaluate a new member of the medical staff. Furthermore, the medical staff could invite assistance from outside technical experts, such as computer programmers, to assist in interrogating the algorithm used in association with a medical error. This would be akin to the common practice of asking a biomedical engineer to inspect a machine (like an intravenous pump) after a serious event, to help distinguish operator error from device failure. These are not mutually exclusive – some combination of these solutions (and more) could be deployed while still maintaining local regulatory control through the classical peer review process.
A scenario that we think is more likely is that AI applications will chip away at traditional physician-centered medical care and even physician self-governance. The entry of AI, eventually, is likely to surpass the capabilities of local-level provider control, and the physician’s primacy in the delivery of health care will wane. One might say that this trend has already begun with greater corporatization in health care and the expansion of nonphysician health care providers. Uptake of AI adds a degree of technical complexity that likely lies outside the physician’s expertise, and the logic underlying physician self-governance will collapse. More immediately, AI would spell the end for peer review as a practice. The complexities of an autonomous intelligent health care machine may simply be too challenging for a quaint process from the late-nineteenth century.
A related challenge will confront the traditional powers that now govern peer review. What is now a physician-dominated process – often exclusively so – must also be able to handle AI-based approaches to health care. A hospital’s most critical personnel decisions, including the allocation of admission privileges, remains (by tradition, law, and otherwise) under the control of the physician-led hospital medical staff. If AI and the peer review process become at odds, the continued use of either one will be determined ultimately by the medical staff. The medical staff would have to prove it was capable of managing the growing significance of AI, or it would have to narrow its authority over hospital operations.
This in turn begs the broader question of whether AI will force a change in the organization of the hospital. Peer review is central to a hospital’s governance structure precisely because it is a window into the hospital’s authority structure. If the locus of control changes in reviewing patient care and scrutinizing provider conduct, the locus of authority will also change in health care delivery. It certainly is hard to imagine that the superiority of AI (if, indeed it earns that superiority) over human-governed quality assurance is compatible with an American hospital’s traditional governance structure. If AI changes peer review, it stands to reason that it will require changes to the hospital as well. The power to define standards and determine when those standards are not met – and by whom – is the power that controls the care delivered within the hospital.
A reallocation and reorganization of power in the hospital would certainly signal broader changes in health care delivery. A hospital system would have little trouble leveraging its AI capabilities to reach a broader scale of patients, and hospital administrators untethered to physician limits might reconfigure delivery. Although removing the human element from medicine will introduce meaningful drawbacks – and presumably, AI and digital services will never duplicate human wisdom of physicians – one could imagine how an operations-centric, AI-centric delivery system could advance the aims of population health. An AI-focused health system might monitor and sustain the health of a large population better than one that services patients with physician visits and hospital-based procedures.
18.7 Legal and Policy Implications
Just as AI might change the operation and governance of the hospital, it will also likely demand changes to health care regulation. The nation’s current angst over increased health insecurity, political demands for distributive justice, and the fiscal strain of health care expenditures already fuel public demands for delivery reform. There may be an acute need to develop policies that thoughtfully usher in AI medicine while assuring a worrying public.
Even as AI grows to replace human roles, it will be treated as a device or product and thus will receive the same legal treatment as other technologies currently in use. For example, AI tools are likely to be subject to product liability law, unlike malpractice law that governs physicians. Moreover, even when physicians maintain responsibility for the care provided by AI tools, the law will have difficulty navigating between the two tort regimes.Footnote 20 This might also mean a greater regulatory focus on the corporations providing health care, rather than on individual providers or specific tools, and thus might stimulate demand for enterprise liability regimes that will feature employed, rather than independent, physicians.
The shift to product liability or enterprise liability, and an erosion of individual professional liability, could be further fueled by a need to scrutinize the core of malpractice law. If malpractice liability law evolved under a professionalism paradigm – one that nurtured and protected peer review and deferred to professional sources of authority – then the supremacy of AI and the inadequacies of human knowledge will require a rethinking of medical malpractice law. Specifically, if expertise lies more within the domain of the computer engineer than the physician, and if general standards of practice succumb to deep learning algorithms, then the malpractice law’s deference to professional standards would be inapposite. Changes would include altering Daubert rules that define expertise, discovery rules that determine what evidence is authoritative and what is not, and substantive rules of tortious negligence and the sources of knowledge from which they are derived.
There would also be a shift towards federal medical device regulation, governed under FDA and intellectual property laws, and away from local regimes that govern the practice of medicine. And because digital products will naturally disseminate in a national (and international) market, there will be diminished tolerance for localities with their own rules and quality standards. We should expect to see a continued loss of local control over medicine, both in law and in practice, and a corresponding shift from local to state and from state to federal oversite of health care delivery.
Licensure regimes might change as well. Instead of licensing boards scrutinizing which humans warrant credentials, the FDA would approve machines and algorithms, and AI certifications would emerge in a to-be-developed federal product approval process. This could both dilute the effect of state medical boards, perhaps the longest surviving institution in modern medicine, and usher in the rise of health care corporations with national reach. Health care systems, insurance companies, and technology companies are the most likely to deploy AI tools, and these providers would be responsible for the outcomes, including safety, cost, and access, much as Ford is responsible for the cars they sell nationwide.
And even if peer review continues, the specific laws surrounding it will be ripe for reform. For example, discovery rules include immunities that protect disclosure during the peer review process, under the logic that peer review requires that immunity. But if quality review relies on digital analytics, which in turn requires the exchange of data across hospital systems, then the logic of limiting discovery is undermined. In fact, even without the presence of malpractice suits, AI would function best with vigorous reporting requirements, subject to appropriate privacy rules. The laws originally designed to nurture peer review will be redesigned to nurture alternative mechanisms to assure medical quality.
This might suggest a broader need to rethink health care regulation more generally. Regulators should be receptive to the possibility that health care regulation should accommodate the needs of AI, rather than the reverse, and deem the emergence of AI as an occasion for a broader reframing of medical regulation.Footnote 21 If AI genuinely represents a potential to improve the quality of medical care, provide the digital products that can achieve population health, and avoid the shortcomings of physician self-governance, then policymakers and medical leaders should plot out the legal rules that would support the growth, improvement, and accessibility of new digital products.
At the same time, there will be significant hesitation to move both the delivery of health care and the regulation of health care away from a professional paradigm. Despite its imperfections, locally controlled, physician-led peer review is a process that has served an important role in ensuring patient safety for more than a century, adapting to innumerable changes in health care practice over that time.
In the end, these questions might be answered by political intuitions and popular perceptions. Peer review has remained a pillar of medical practice because it has succeeded in maintaining the trust of the public, and a move to algorithms could undermine that trust. But the integrity of peer review has been showing cracks of its own and may not continue to win the confidence of a digitally connected public. Perhaps the incorporation of AI tools will require a very different set of rules to maintain public trust.
The authors acknowledge the support of the NIH BRAIN Initiative project, “Cognitive Restoration: Neuroethics and Disability Rights,” 1RF1MH12378-01. The views expressed herein are the authors’ own and not necessarily those of their institutions or the NIH.
19.1 Introduction
Research participants in clinical trials often have an interest in maintaining access to a drug or device after the trial has concluded. In the case of clinical trials of in-dwelling Class III medical devices, which are life-sustaining or risky devices that are implanted in the body, participants’ interest in posttrial access is considerable. Such devices can be harmful to the participant if not properly maintained or removed, and if the device is beneficial to the participant, they may desire surveillance and maintenance to ensure proper device functioning, as well as access to replacement devices.
To date, the Food and Drug Administration (FDA) has not provided clear guidance about the posttrial access obligations clinical trial sponsors and investigators have to research participants. And while litigation about posttrial access in the case of investigational drugs has resulted in courts finding that there is no legal duty for pharmaceutical companies to continue to provide access to the tested drug to study participants, it is not clear how this body of law would apply to in-dwelling Class III medical devices or what normative obligations sponsors and investigators of such device trials have.
Legal and ethical clarity on this issue is of critical importance, given that such devices, unlike drugs, will remain in a participant’s body and may require ongoing maintenance, surveillance, replacement, or explanation.Footnote 1 Further, prospective research participants may decline to enroll in studies assessing the safety and efficacy of embedded Class III medical devices if they are not guaranteed posttrial access. Such recruitment problems could inhibit production of scientific knowledge and delay effective medical devices from making it to the market, thus harming innovation.
This chapter first explains the FDA approval process for Class III medical devices and the resulting issue of posttrial access to in-dwelling devices. The chapter then explores the law and ethics of posttrial access for drugs, devices, and biologics, highlighting the dearth of legal guidance. The chapter then discusses the case of posttrial access to deep brain stimulation (DBS) and patient perspectives on this issue. We conclude with a call for transparency about the type and degree of posttrial access as part of the preimplantation informed consent process as well as mandating that sponsors fund device maintenance or explantation after the conclusion of the trial.
19.2 Posttrial Access to In-Dwelling Class III Medical Devices?
The FDA categorizes medical devices based on the type and degree of risk from the device. Class III medical devices, such as pacemakers or deep brain stimulation (DBS), are the highest-risk category, and thus receive more scrutiny from the FDA. Class III medical devices are defined as those that are “life-supporting or life-sustaining, or for a use which is of substantial importance in preventing impairment of human health, or … present a potential unreasonable risk of illness or injury.”Footnote 2 Such devices must, through scientific evidence, demonstrate “reasonable assurance” of safety and efficacy prior to FDA approval.Footnote 3
Not all clinical trials of Class III medical devices will successfully demonstrate safety and efficacy, necessary conditions for receiving FDA approval.Footnote 4 Other clinical trials may be successful on these measures, but the study sponsor and investigators may ultimately decide not to bring the device in question to market.Footnote 5 Both scenarios can strand research participants who may have an interest in maintaining access to the device if the intervention is or perceived to be efficacious, especially if the device is implanted in the research participant’s body (also referred to as invasive or in-dwelling devices).Footnote 6 Posttrial access may include routine device maintenance, repair or replacement if the device malfunctions, or device removal, all of which often require specialized skills that only study investigators have. Uncertainty about posttrial access to in-dwelling devices may dissuade prospective participants from trial enrollment, potentially thwarting the progression of device development from bench to bedside.
As the next section demonstrates, there has been little regulatory guidance from the FDA about the posttrial obligations owed to participants by device or drug sponsors and little case law clarifying this issue.Footnote 7 Furthermore, while industry norms tend to govern posttrial access to pharmaceuticals – often offering limited posttrial accessFootnote 8 – these norms are neither established nor directly analogous to questions of posttrial access to in-dwelling Class III medical devices, such as DBS.
19.3 Law of Posttrial Access
Statutory and regulatory guidance for posttrial access to drugs, biologics, and devices is sparse. Most of the attention has instead focused on the passage of state and federal right-to-try laws and expanded access (i.e., compassionate use) to investigational drugs and devices through the 21st Century Cures Act.Footnote 9 Conceivably, a former study participant could seek posttrial access through one of these other routes,Footnote 10 but the legal and ethical rationales for permitting or prohibiting such access will differ. Right-to-try laws allow terminally ill patients and their physicians to request access to early-stage drug trials, although the study sponsor does not have to grant access.Footnote 11 Right-to-try laws exclude medical devices.Footnote 12 The FDA also has an expanded access program for seriously ill patients who have no other treatment options to access investigational medical products, including medical devices, that have not yet demonstrated safety or efficacy.Footnote 13
While accessing investigational drugs and devices outside of enrollment in a clinical trial may be possible through right-to-try laws and expanded access FDA pathways, these options do not directly address the situation of someone who formerly had access to an investigational drug because of their participation in a safety or efficacy study and desires continued access. Moreover, if the device has already made it to market, the expanded access pathway is no longer relevant.
Case law provides some insight into the responsibilities study sponsors and investigators have to provide posttrial access to pharmaceuticals to clinical trial participants. In the mid-2000s, in two well-known court cases,Footnote 14 participants in a study testing a new drug for Parkinson’s disease wanted continued access to the drugs that they believed were beneficial, but the study sponsor ended the trial because there were safety concerns and limited evidence of efficacy. The study sponsor also refused to provide posttrial access even under the compassionate use option.Footnote 15 The study participants argued that there was a contractual duty for the sponsor to provide posttrial access, that they relied on the sponsor’s promise to provide the drugs post trial, and that the study sponsor had a fiduciary duty to participants that required posttrial access.Footnote 16 Their breach of contract claim failed because the agreement study participants had was with investigators (i.e., the informed consent document) rather than the study sponsors; their promissory estoppel claim failed because again, there was no promise made by the study sponsor to study participants; and the breach of fiduciary duty claim failed because the court declined to find the study sponsor to be a fiduciary, or a person who is expected to act in the best interest of another. The court also considered policy reasons for providing posttrial access, namely that patient enrollment will decline if this is not ensured, but also considered policy reasons against a mandate for sponsors to provide posttrial access, namely that companies would be less inclined to sponsor drug trials in the future.
Posttrial access continues to be governed by private agreement rather than public regulation,Footnote 17 which means that study participants are only entitled to what study sponsors and investigators are willing to explicitly agree to, which may not accord with participant preferences or ethical principles such as benevolence, nonmaleficence, and justice. For example, prior to one study of DBS for severe depression, the study sponsor, a medical device company, agreed to pay for the surgery to remove the device and continue supplying batteries.Footnote 18 While many participants reported experiencing a benefit from the device, the trial was unsuccessful, and participants who wished to retain the device were left to cover the costs of future maintenance and care themselves, a situation that raised many ethical issues.Footnote 19 Given the absence of legal guidance, ethical practice becomes more important. The next section addresses normative dimensions of posttrial access.
19.4 Ethics of Posttrial Access
There has been extensive academic commentary about what, if any, ethical duties are owed to clinical trial research participants. Multiple commentators have argued that if participants have experienced a benefit from an investigational intervention during a trial, the principles of nonmaleficence and beneficence demand that they should continue to have access after the trial concludes.Footnote 20 Similar arguments for access to trial benefits using the principle of reciprocity have been made given that participants have undergone risk to advance scientific understanding and benefit future patients.Footnote 21 These two arguments are embedded in an earlier version of the Declaration of Helsinki, adopted by the World Medical Association: “[a]t the conclusion of the study, patients entered into the study are entitled … to share any benefits that result from it, for example, access to interventions identified as beneficial in the study or to other appropriate care or benefits.”Footnote 22 Finally, some have argued in the case of DBS trials, the principle of nonabandonment brings with it “a longitudinal fiduciary obligation to provide [research participants] with support,” specifically “a responsibility to provide on-going care and a fiscal responsibility for any associated costs” on the part of study investigators and sponsors.Footnote 23
Some argue that determinations about posttrial access should be left to the discretion of investigators and sponsors because there may be uncertainty about whether there is true efficacy of the intervention or whether the benefits outweigh the harms for individual clinical trial participants and for the population at large.Footnote 24 Some also contend that research participants may have already received benefits from study participation, addressing reciprocity obligations.Footnote 25 Offsetting this argument in the case of Class III devices, any benefits may also be accompanied by potential burdens or complications associated with in-dwelling devices or their removal.Footnote 26 Furthermore, while participants likely expect posttrial access to the investigatory technology in part because of a therapeutic misconception,Footnote 27 the appropriate remedy to this problem is not posttrial access.Footnote 28 Instead, scholars have generally called for greater planning and transparency about posttrial access.Footnote 29 Finally, mandating posttrial access may raise research costs, ultimately disincentivizing research, resulting in collective societal harm.Footnote 30
While scholars have engaged in the above-described debates, there have been many fewer studies of the viewpoints of participants about duties owed to them stemming from trial participation, especially in the context of in-dwelling devices. Existing empirical research of clinical trial participant attitudes about duties owed to them by investigators, research sponsors, drug and device manufacturers, and regulators has found that many, although not all, participants think there is a duty to provide posttrial access to a drug (or its equivalent) that provided them with benefits after the trial and until approval,Footnote 31 and after approval at a fair market or reduced price, for an indefinite period of time.Footnote 32 This small body of empirical research about posttrial expectations focused on reports from drug trial participants. It is unknown how device trial participants view these questions.
19.5 Deep Brain Stimulation and Posttrial Access
Posttrial access to DBS, which is classified by the FDA as a Class III medical deviceFootnote 33 because it has a greater level of risk to patients,Footnote 34 is a particularly pressing issue. DBS is a “programmable and adjustable implant of electrodes into specific deep brain structures that delivers electrical impulses to alter circuit function and overcome abnormal activity.”Footnote 35 Once properly implanted, the electrodes can be stimulated with varying levels of voltage to produce desired cognitive, emotional, and physical effects. A battery for the electrodes is placed in a patient or research participant’s chest. DBS has been shown to be effective and is approved by the FDA for Parkinson’s diseaseFootnote 36 and has been tested for treatment-resistant neuropsychological disorders, such as depression.Footnote 37 The application of DBS for traumatic brain injuries is currently being explored.Footnote 38
While all research participants have interests in posttrial access to investigational drugs or devices, the stakes are higher for persons enrolled in trials of invasive Class III medical devices, such as DBS, given the higher level of risk of the device coupled with the reality that participants cannot remove the devices. So unlike in clinical drug trials in which the participant can discontinue use of the study drug after the conclusion of the trial, participants in in-dwelling medical device trials have a device permanently embedded, barring a procedure to remove it.Footnote 39 Participants in such trials thus have a pressing interest in the device’s safety, and if efficacious, ensuring that it remains available to them. This is especially the case for invasive neural technologies. Not only do implanted neural devices share the features of risk and permanence of other in-dwelling medical devices such as cardiac pacemakers, but they affect the brain, and implicate cognitive abilities, personality, identity, and agency in a way that other investigational devices and drugs may not,Footnote 40 which again raises the stakes of posttrial access.
Because both electrodes and the battery are implanted devices that cannot be removed by the participant, questions arise about how these devices will be surveilled, maintained, replaced, or removed after the clinical trial assessing their safety and efficacy concludes, especially if the device never receives FDA approval, never makes it to the market, or the device manufacturer stops making the device.Footnote 41 Indeed, there are significant hardware risks with DBS “including infection, malfunction, erosion, and migration or fracture of leads, which may require additional surgery or explantation.”Footnote 42 Furthermore, many persons implanted with DBS need access to specialized neurosurgeons and neurologists for battery replacement and device programming, which they may not have access to once the study concludes.Footnote 43 Even if DBS research participants have continued access to study investigators, “they may be left with costs for device maintenance, continued access, or explantation,”Footnote 44 which often are not budgeted for in grants that fund this research and which health insurance likely will not cover.Footnote 45 As one of us asked over a decade ago when writing about research participants in trials of investigational neuromodulation technology, “What is their fate? What happens to these patients when the trial ends? Who provides on-going care? Who pays for battery replacement? Who removes a broken device? Who adjusts stimulation parameters … in perpetuity?”Footnote 46
Recent ethical guidance from the NIH BRAIN Initiative Neuroethics Working Group, of which one of us (JJF) is a member, about posttrial access to neural devices such as DBS addresses some of these questions. The guidance includes the following recommendations: planning in advance of a study for research participants’ posttrial access needs, regardless of whether the device is safe and effective, including planning for cost; ensuring that posttrial access issues are addressed in the process of obtaining institutional review board approval for the study and that plans are communicated to research participants in the informed consent process; and requiring greater obligations posttrial from study sponsors and investigators if the device is beneficial or risky, the study participants are vulnerable, the provision of access would not be costly, the device is too complex for general health care professionals to manage, or the device contains “built-in obsolescence and proprietary hardware and software, effectively locking patients and clinicians into ongoing relationships with a manufacturer.”Footnote 47
While the Working Group offered ethical guidance about posttrial access to neural devices, their suggestions do not have the force of law and it is unclear to what extent study sponsors and investigators are heeding these suggestions. Indeed, the Working Group notes that the “locus of posttrial responsibilities is currently determined on a case-by-case basis.”Footnote 48 And importantly, the views of research participants enrolled in clinical trials of invasive neural devices such as DBS regarding posttrial access remain unclear.Footnote 49
Presently, as part of an ongoing larger study, we are studying the perspectives and experiences of research participants in a DBS clinical trial for patients with moderate to severe traumatic brain injury,Footnote 50 and one question we ask is about participants’ concerns about posttrial access to the investigational device.Footnote 51 Preliminary results from research participants provide a window into the posttrial access hopes and concerns of those enrolled in invasive neural device clinical trials.
Study participants have questions and concerns about posttrial access. Some participants ask about device support and maintenance prior to agreeing to participate in the trial. One participant described proactively engaging their health insurer to determine whether their insurance policy would cover the cost of battery replacement, for example, and also asking investigators about the length and degree of posttrial support. While he understood budgetary constraints of guaranteeing posttrial access in perpetuity, he thought that if there was a benefit from the device, participants should have ongoing access to support and maintenance.
Another study participant became concerned about posttrial access after they had already been implanted. The participant indicated a desire that the device be turned up, and that it never be turned off because it provided such a benefit. The study participant did not want to go back to a time when the device was not available. In fact, the participant emphasized that future study enrollees be warned that they may have a negative experience if their device is turned off during or after the study because they will revert to their old self. That is, if participants experience a beneficial change due to DBS, then if they no longer have access to a functioning device (e.g., dead batteries, faulty electrodes, etc.), they may feel harmed. This participant’s informal caregiver also echoed the study participant’s concerns, emphasizing that the positive effect of DBS on the participant’s life has been so profound that they hoped that the device is never turned off.Footnote 52
While more data about the views of participants and their informal caregivers is needed, these preliminary insights speak to the need to hear the voices of those most proximate to these trials. As we continue our study, we plan to add more research participant perspectives to the policy and ethical debate over posttrial access to implanted Class III medical devices.
19.6 Adapting the Regulatory Regime for Innovative Medical Device Technologies
The FDA recently released draft guidance calling for patient input into clinical trial design for medical devices.Footnote 53 The guidance about patient engagement is meant to “mitigate some of the practical challenges to robust clinical investigations, including challenges concerning study/research participant enrollment and retention in the study” through strengthening the informed consent documents and prioritizing clinical endpoints patients care about, for example.Footnote 54 Another important part of patient feedback on clinical trial design is what patients/participants and families think investigator and sponsor responsibilities are with respect to posttrial access to a functioning embedded medical device. These data can help inform policy creation.
As data collected from participants in our study has shown, individuals have an interest in maintaining access to a safe and effective device after their participation in a clinical trial ends. But while their perspectives are important, they are just one part of the regulatory puzzle. The views of investigators, sponsors, and manufacturers also need be considered, as the social compact centering around device implantation transcends the narrow purview of informed consent, especially if there are conflicts between participant preferences and the sponsors or manufacturers bringing innovative devices to market given economic constraints.
We argue for a bifurcated conception of responsibility for posttrial access. With respect to investigators, we argue, that at a minimum, they owe a duty of complete transparency to participants and prospective participants about posttrial access to surveillance, maintenance, upgrades, or removal of Class III implanted devices as part of an ongoing informed consent process. Transparency about posttrial access necessitates advance planning. Our argument for planning and transparency is in line with the recommendations from the NIH BRAIN Initiative Neuroethics Working Group discussed previously.Footnote 55
Duties to plan for and be transparent about posttrial access are not limited to medical device trials. Indeed, these duties apply to all clinical trials, including pharmaceutical trials, because disclosing posttrial access plans to prospective participants is necessary to obtain genuine informed consent. But ensuring ethical clinical trials of Class III implanted devices, especially neural devices, demands more of investigators and sponsors to ensure that study participants are not harmed and are treated justly, given participants’ posttrial access preferences and expectations; the greater risk to study participants, many of whom are vulnerable because of their existing medical conditions, from implanting a device; the probable permanence of the device and the need for ongoing access to specialized medical care and device maintenance; and the potential for changed personality and identity with a brain-based medical intervention.
Thus, with respect to study sponsors and the medical device industry, we argue that there is a correlative set of responsibilities to participants and their families to ensure ongoing access to repairs, maintenance, and the costs of explantation (should participants desire device removal) for embedded neural devices. Funds should be put aside at the start of a trial to ensure such access after a trial has concluded. While this requirement may seem financially onerous, these costs would likely be a small fraction of the total expenditures related to research and device development. Indeed, these ongoing costs should be understood as central to maintaining the integrity of this work, as part of the cost of doing business, and a concrete set of ethical obligations given the unique challenges of in-dwelling neural devices.Footnote 56 With this recommendation, we move beyond what the NIH BRAIN Initiative Neuroethics Working Group proposes, given the many qualifications contained in their arguments about posttrial access (e.g., conditioning investigator and sponsor responsibility to provide access on criteria such as study participant vulnerability, degree of device benefit, cost of posttrial access, etc.).Footnote 57
These additional responsibilities should be more than an ethical duty – they should also be legally required. But unlike the current norm of leaving matters of posttrial access to private agreements, we argue that posttrial access decisions should be subject to regulatory oversight by the FDA, which can ensure that investigators and sponsors are fulfilling their ethical duties to study participants after balancing the competing interests, if any, of the parties. Additionally, both aspects of this bifurcated set of responsibilities should be approved by an Institutional Review Board prior to the beginning of the clinical trial.Footnote 58
20.1 Introduction
There is growing concern over the FDA’s persistent failure to prevent the marketing of medical devices that subject patients to previously undetected risks of death, disability, and other serious injuries.Footnote 1 Departing from the dominant approach to reform calling for expanding FDA authority and resources, state tort law, and other modes of government oversight, we consider harnessing the regulatory power of market forces, particularly those uniquely exerted by health care insurers (“insurers”).Footnote 2 Essentially, insurers’ regulatory power derives from their market-gatekeeping coverage and purchase decisions that determine the economic fate of all FDA-approved devices; capacity to constantly and comprehensively monitor the market for product-related accidents, including manifestations of new and increased risks; and exposure to paying the medical and other expenses of injured insured patients.
Insurers thus can surpass other nongovernmental as well as governmental forms of oversight (for example, academic researchers, physicians, manufacturers, tort lawyers) in enhancing FDA efforts to protect patients from unreasonably risky product designs, warnings, and usage. They can draw on the continuous inflow of insured-patient requests for payment of medical and other expenses resulting from product-related injuries to supply the FDA with both a superior source of reliable postmarket data on product risk and efficacy and virtually instantaneous notice of emerging signs of new or increased risks.Footnote 3
Monitoring for emerging device risk is of vital importance to insurers because they will pay the medical and other product-related accident costs incurred by their insured injured patients. Exposure to bearing product-related accident costs drives insurers to include the implicit price of accidents in coverage and purchase decisions. Therefore, insurers operating in the normal course of business select for safer and more efficacious medical devices and uses.Footnote 4
Insurers’ risk-rated coverage and purchase decisions can serve as an unmatched means of fortifying manufacturers’ incentives to exercise reasonable precautions in developing, testing, and marketing their products. They also supplement FDA-prescribed warnings and informed physician judgments by curtailing overuse of medical devices, evaluating the comparative safety and efficacy of products and other treatments, and better-fitting product benefits to patients’ medical needs. Because insurers bear the accident costs of false positives – that is, of curtailing patient access to a device based on an erroneous finding of undue risk – as well as false negatives, they have incentives to make measured and reliable decisions.
Yet, market impediments may prevent insurers from exercising their regulatory powers for maximum social benefit. Depending on market and other factors, sharp competition can be part of the problem. An insurer might delay or refrain from publicly reporting the discovery of an emergent risk to the FDA for fear of competitors freely capitalizing on proprietary information concerning adjustment of its coverage and purchase decisions. While transmission of product-related risk in insured-patient payment requests may not involve great expense, translating that information into risk-rated coverage and purchase decisions is another matter. Deriving reliable implicit risk-rated prices to incorporate into coverage and purchase decisions can involve expensive AI systems and other private sources and methods of aggregating and analyzing data to discern patterns or even signs of new or increased risks; determine causal mechanisms and associations in the various contexts, practices and behaviors that frequently characterize the heterogeneity of health care provider and patient use of the product; and estimate accident prevalence and costs among the patient population.
Most importantly, insurers lack sufficient financial incentives to exercise their regulatory powers for maximum social benefit because two structural features of all insurance systems, private and public, shield them from bearing the total costs of product-related accidents. First, because risk-averse individuals are unwilling to pay premiums or taxes for nonmonetary losses (harm that money cannot remedy, such as death), insurance does not provide coverage for it.Footnote 5 Nonmonetary loss, however, represents a real and major diminution of individual welfare that must be included in the total amount of product-related accident costs when determining the reasonable amount of resources to expend in securing maximum social benefit from safety precautions. The second structural constraint is insurance subrogation, the contractually or legally created means by which insurers recoup from insured-patient tort recoveries the amount they have paid out in covering medical and other injury-related expenses.Footnote 6 In calculating the product-related accident costs they anticipate bearing – to determine coverage and purchase decisions – insurers rationally discount that burden by the amount they expect to be reimbursed from tort recoveries through subrogation. Cumulatively, these structural constraints relieve insurers of well over half of the total product-related accident costs.Footnote 7
We propose two simple and virtually costless federal statutory reforms to correct these market defects. Pursuant to the first, Congress would require insurers to report medical device accidents to the FDA. This would overcome any market competition constraints on insurers’ willingness to publicly disclose proprietary information. The second would have Congress establish a federal rule of manufacturer strict tort liability that is predicated on proof of causation alone and pays damages directly and fully to the US Treasury.Footnote 8 For the purposes of removing the structural constraints on insurers’ financial incentives to reduce risk, the principal virtue of manufacturer strict liability is that tort damages account for both monetary and nonmonetary losses and – because manufacturers will reflect total expected tort damages in their product prices – thus lead insurers to consider the total costs of product-related accidents in monitoring the market for risk and risk-rating their coverage and purchase decisions. Paying recoveries into the US Treasury eliminates the other market defect of subrogation reimbursement.Footnote 9 Initiation of strict manufacturer liability actions would first require FDA validation of the causal connection between product and injury, and then the decision by the Civil Division of the Department of Justice to litigate claims directly or by auctioning them to private attorneys.
We examine the mandatory reporting proposal in Section 20.2 and the manufacturer strict liability proposal and system for enforcing it in Section 20.3. Section 20.4 concludes.Footnote 10
20.2 Mandatory Reporting
Currently, Congress requires only manufacturers (or importers) and device user facilities (end-users) such as hospitals to report medical device accidents. In this section, we address whether insurers should be included.
Driven by financial self-interest and informed by the constant flow of insured-patient payment requests, insurers have the unrivaled capacity to monitor the market continuously and comprehensively for incidents of product-related accidents generally, and signs of emergent danger especially. Insurers are thus uniquely equipped to serve FDA market surveillance objectives, particularly as early warning “watchdogs.”
Undoubtedly insurers are motivated to voluntarily report newly detected risks to the FDA. Insurers, like other participants in the health care system, embrace the ethos of “doing no harm.” Further, in accelerating FDA investigation and intervention, insurers’ reporting will reduce accident risk and hence their outlays for medical and other expenses, and relatedly their costs to analyze risk data and adjust coverage and purchase decisions accordingly. Expedited FDA intervention has the further beneficial effect of preventing insurers from perversely competing for market share by delaying or otherwise manipulating coverage and price responses to newly discovered risks.
Competition gives rise to concern that insurers may lack optimal reporting incentives. Despite benefiting from accelerated FDA intervention, insurers may hesitate to report newly discovered risks in some cases. Doubtless, insurers will not think twice about reporting accidents that directly implicate readily determinable defects or risky features of a widely sold or frequently used product. In such cases, no competitive advantage is likely to accrue from delayed reporting, as other insurers probably would be experiencing similar accidents and making corresponding adjustments in coverage and purchase decisions. A different case arises when accidents are sporadic or the insurer incurs substantial expense in generating proprietary information to discover the risk, determine its nature, estimate product-wide accident incidence and costs, and based on that analysis, make risk-rated coverage and purchase decisions. The prospect of competitors free riding on this investment may dull the insurer’s incentive to immediately notify the FDA.
Congress can address this problem simply by subjecting insurers at minimum to equivalent investigation and reporting requirements as those presently applied to manufacturers and end-users. That mandate casts a broad discovery net for any information the reporter may have or can reasonably obtain that suggests that use of, or exposure to, a medical device caused, contributed to, or had been a factor in causing or contributing to the injury of a patient (or health care employee, or another person). The source of the risk is also defined capaciously to include product malfunction, failure, manufacturing or labeling defects, or user error. The mandate applies to major product accidents involving death or other serious injuries – defined as posing a threat to patient life, danger of permanent impairment of body function or structure, or need for medical intervention to prevent such fatality or impairment.Footnote 11
Generally, in choosing between voluntary and mandatory reporting of adverse information, the system designer considers the relative social value of the former motivating discovery of more information for private use and the latter motivating discovery of less information for public use.Footnote 12 Regarding insurers, both factors point unambiguously in favor of mandatory insurer reporting.
The key variable affecting the quantity of reported information is whether insurers’ concerns about adverse effects from public disclosure might lead them to ignore or underinvest in discovering product-related risks. Normally, such perverse incentives arise when the adverse information triggers administrative agency and tort liability sanctions. Contrary to manufacturers and end-users, insurers face no such adverse consequences from reporting product-related risks to the FDA. The only potential cost is competitor free riding, which affects manufacturers to a far greater extent. Moreover, insurers’ marginal cost, if any, will likely be negated by the benefits of FDA intervention and the fact that the entire industry is subject to the mandatory disclosure rule. Regardless, insurers will hardly find wilfully reducing monitoring efforts worthwhile, as this increases the chance of paying large, unexpected accident costs and only prevents the possibility of a temporary and usually small competitive disadvantage.
Regarding the second factor, the question is whether greater regulatory benefits accrue from private party use of more discovered information than from public regulator use of less disclosed information. Greater discovery efforts under a voluntary disclosure regime might result in manufacturers detecting product-related risks that they can remedy, for example, by recalling the device before anyone else recognizes the problem. Yet, as Congress apparently decided in mandating manufacturer disclosure, the prospect of voluntary recall – which might well be small given the high costs exacted by competitive market forces, including bankrupting a firm with few revenue-generating products – was outweighed by the regulatory benefits from disclosure, including spillover gains in agency knowledge and experience for overseeing similar products and benefits from its independently remedying the problem with the product in question. Extending the mandate to insurers is not a close call, as there is no conflict of interest in the public and private use of product-related risk information. Quite the contrary, their complementary use of the information synergistically enhances joint regulatory benefit.
20.3 Strict Liability
This section explains the purpose and evaluates the cost-effectiveness of the proposed rule of strict liability and the system for enforcing it.
The regulatory function of civil liability, like that of the FDA and other government and nongovernment modes of controlling medical device risks, is motivating risk-controllers (manufacturers) to invest in reasonable precautions. Reasonable precautions result from manufacturers optimally adjusting two principal interrelated risk-control factors: level of care (for example, improving product design to facilitate sterilization) and level of risky activity (for example, reducing resorts to CT scans). In threatening manufacturers with paying for a patient’s total product-related monetary and nonmonetary losses – to the extent measured and monetized in tort – civil liability induces the manufacturer to take reasonable precautions by adjusting the interrelated care and risky activity levels to avoid creating and marketing an unreasonably dangerous product.Footnote 13
Because the straightforward threat of bearing total accident costs motivates manufacturers to exercise reasonable precautions, strict liability achieves this regulatory objective without entangling courts and litigants in a misbegotten fact-finding process of determining what interrelated levels of care and risky activity constitutes reasonable precautions and whether the manufacturer took such precautions in fact. Manufacturers will consider all relevant dimensions of care – from the salient matters of product design to the many less conspicuous but no less critical choices in the scope of research, including the performance of nonmedical devices; methods, types, setup, and management of safety studies; qualifications, training, and compensation of researchers and managers; extent of premarket tests and other efforts to discover the potential for latent risks; and investigation of countless scenarios of how, when, and where the product will be used, including consideration of differences in end-user abilities and behavior. Similarly, the manufacturer will make the socially appropriate investment in moderating its risky activity level. For example, it may reduce excessive sales – exposing a sub-group of patients to a risky device for little or no offsetting gain in medical benefit – by toning down advertising and refraining from engaging in problematic promotional tactics. Beyond that, strict liability has the singular activity level-reduction benefit of compelling manufacturers to internalize expected damages and incorporate the anticipated total accident cost in their product prices. Thus, strict liability engenders a “price-signaling” effect that lowers demand, reducing unnecessary use of a risky product, and, ultimately, the incidence of injury.
Health care insurers, functioning as expert buyers with gatekeeping market power, make the medical device market ideal for the use of strict liability to regulate product risks. Strongly motivated to monitor for, and incorporate, the implicit cost of expected product-related accidents into their coverage and purchasing decisions, insurers will be highly attuned to the price signals from strict liability.Footnote 14 Far from price takers, insurers would respond to those signals with speedy and deliberate adjustments of coverage and purchase decisions, effectively reducing risky product sales, use, and hence patient injury.Footnote 15
Strict liability and health care insurers mutually reinforce their power to regulate medical device risk. Insurers improve the regulatory coherence of strict liability pricing signals. Risk-neutral, rational, and expert insurers will be free of the risk misperception and demand elasticity problems that may distort the effects of price signals on ordinary consumers.
Regarding insurers, strict liability per se, through its price-signaling effect, closes the major gap in their accident-cost exposure for product-related injuries, in addition to saving them the cost of calculating the implicit price of accident risk. Threatening liability for total expected accident costs, strict liability leads insurers to internalize nonmonetary as well as monetary losses, and to adjust their coverage and purchase decisions accordingly, thereby maximizing the social benefit of their regulatory power.
However, the proposed strict liability rule is needed to fully correct structural market defects. This is because, under conventional strict liability, insurers retain subrogated reimbursement for outlays to cover monetary losses. Subrogation reimbursement substantially reduces insurers’ financial incentives to maximize regulatory benefits both by offsetting their coverage exposure and by diluting strict liability’s price-signaling effects for monetary losses. The proposed rule corrects this market defect, effectively eliminating subrogated reimbursements, by requiring payment of all recoveries directly and in full to the US Treasury. Ending the prospect of subrogated recoupment will spur insurers to take full account of total expected accident costs – nonmonetary and monetary – when determining the implicit risk-rated price of a device they are considering covering and purchasing.
The system we envision for enforcing the proposed rule of strict manufacturer liability should assure its reliable, measured, and socially appropriate use. Prospective claims would proceed through two stages of merits screening. First, the FDA would, in the normal course of investigating and considering its regulatory response to reported incidents of serious device-related accident, verify the nature, extent, and harmful consequences of the causal connection between product use and patient injury.Footnote 16 The manufacturer probably would be notified that the investigation is ongoing and, when needed, required to disclose relevant information and otherwise participate and cooperate fully in the investigatory process. Only positive determinations of causation and harm would send the case to the next stage. At any point in this process, the FDA can exercise its normal regulatory power to control the product risk, including order recalls, curtail marketing, and require new or amplified warnings.
The Civil Division of the Department of Justice would conduct the second stage of merits screening. Division lawyers will formulate and review the merits of the strict liability claim and appraise its expected recovery value net of litigation cost. To avoid wasting government, manufacturer, and court resources, the claim would be dropped (or converted into a fixed fine) unless its expected net recovery value exceeds some minimum threshold amount, best set by Congress. Before litigation commences, the manufacturer may present contradictory or mitigatory evidence and seek settlement.
When the case goes to court, the government could sue directly or auction the claim to private attorneys. If the claim is auctioned, the winning bidder will pay the bid amount to the US Treasury and retain any recovery from successfully litigating the case. To reduce the complications and costs of litigation, Congress could give FDA findings of a causal connection the evidentiary force of a rebuttable presumption establishing a prima facie case of liability on the causation element and promulgate a schedule of damages that would replace ad hoc and disputed case-by-case litigation and recoveries.Footnote 17
In sum, the combined effect of strict liability price-signals and, with the elimination of subrogation reimbursements, exposure to paying insured-patient economic losses will lead insurers to optimally risk-rate coverage and purchase decisions. This, in turn, will reinforce manufacturers’ incentives to take reasonable precautions in developing, testing, and marketing medical device products. The inflow of insured-patient bills will also enable insurers to inform the FDA of product-related accidents, including those indicating emergence of increased and new risks. Based on their current and comprehensive knowledge and estimates of the therapeutic and accident experience of products on the market, insurers’ coverage and purchase decisions can disaggregate the generalizations of FDA warnings and statistical models of academic researchers to supplement physician judgments in fine-tuning the fit between comparative product benefits and patients’ medical needs.Footnote 18
Two questions about the cost-effectiveness of the proposed strict liability rule and its enforcement system warrant attention: first, as with any reform proposal, whether expected social benefits exceed administrative and substantive law enforcement costs; and second, more specifically, whether the strict liability rule would better promote social welfare by paying damages as compensation to injured patients, rather than to the government.Footnote 19
20.3.1 Administrative Costs
The dispositive answer to this question is that the administrative-cost footprint of our proposal is virtually nil. Enforcing the proposed strict liability rule generally entails no complicated legal and factual issues. All courts, and hence the government and manufacturers, need to know is the causal connection between the patients’ product use or exposure and the resulting accident losses. These are straightforward matters in most cases.
This no-cost assessment holds even though our proposal extends civil liability to medical devices that Supreme Court preemption rulings currently shield from state tort law, and could give rise to disputes over causation and nonmonetary loss in some cases.Footnote 20 The reason is that the litigation of all strict liability claims hinges on FDA findings of causation, and the FDA (with manufacturers typically participating) will continue to investigate and determine that question exactly as it currently does in carrying out its regulatory function in every case of serious product-related injury for all classes of device.
Disputes will be especially likely to settle quickly and inexpensively in the proposed system. Expecting FDA causation findings to strongly influence the outcome of adjudicated claims and leery of chancing juries awarding high nonmonetary damages, manufacturers will almost surely forgo follow-on litigation in favor of settlement. Moreover, because strict liability damages will be levied and distributed solely for deterrence purposes, and therefore can be assessed on average rather than for individual patients, courts could readily employ collectivized modes of adjudication, such as class actions and sampling, to resolve any causation and nonmonetary loss disputes.Footnote 21 Congress could further reduce administrative costs, as noted above, by giving FDA causation findings the force of a rebuttable presumption establishing a prima facie case for strict liability and promulgating a schedule of damages.
Some might think, mistakenly, that strict liability damages will inflate manufacturers’ costs of doing business and inhibit their investment in device innovation. The proposed rule merely shifts the burden of bearing accident costs from patients to manufacturers, who would otherwise have borne them but for the defective medical device market. Indeed, manufacturers could never successfully dump such accident costs on well-informed patients purchasing medical devices in a well-functioning market. In correcting the defective medical device market, the proposed strict liability rule thus revokes a subsidy that perversely increases manufacturers’ profit margin at the expense of patients’ safety.Footnote 22
20.3.2 Substantive Costs
It would also be a mistake to think the negligence rule is more cost-effective than strict liability. The negligence rule suffers from long-recognized and well-documented fundamental regulatory failings.Footnote 23 In requiring courts to determine whether a defendant manufacturer exercised reasonable precautions, the negligence rule entails an enormously expensive regulatory inquiry, one that is inevitably misguided and socially wasteful. Primarily, high-cost barriers prevent courts from obtaining and analyzing evidence of critical relevance regarding multiple dimensions of care and risky activity. Deprived of this evidence, courts cannot reliably make the complicated factual findings on which the basic questions of negligence liability must turn: first, establishing the optimal, interrelated adjustment of levels of care and risky activity that defines the standard of reasonable precautions governing the case; and second, determining whether the manufacturer’s actual precautions satisfied the standard. Consequently, enforcement of the negligence rule systematically fails to confront manufacturers with sufficient sanctions – that is, with a threat of liability for damages equaling total accident costs – and hence fails to create optimal legal incentives for them to take all reasonable safety precautions in developing, testing, and marketing their products.Footnote 24
Compared to the negligence rule, strict liability produces superior regulatory results because courts can enforce it without undertaking the costly task of establishing and applying a standard of reasonable precautions and making the resultant complicated factual findings. As Holmes observed in explaining the policy supporting use of strict liability rather than negligence, “as there is a limit to the nicety of inquiry which is possible in a trial, it may be considered that the safest way to secure care is to throw the risk upon the person who decides what precautions shall be taken.”Footnote 25 The same advantage of strict liability applies with added force to avoiding even greater cost barriers to determining the far more complex questions regarding the reasonable level of risky activity, and ultimately, the reasonable combination of care and risky activity levels.Footnote 26
Many think the negligence rule has a possible litigation-cost advantage because only claims evincing both negligence and causation will be filed compared to strict liability allowing suit on causation alone. The plausibility of this conjecture, however, is undermined because it never accounts for the costs of plaintiff-lawyers necessarily investigating the entire pool of plaintiff device-caused injuries to determine which among them involve sufficient evidence of negligence, while the proposed strict liability rule entails no such need and cost. It also fails to account for strict liability’s superior deterrent effects that reduce the number of injuries and hence resulting claims. Even assuming some marginal filing-cost advantage of the negligence rule, it is doubtful that the savings would come close to negating the rule’s disadvantages of great trial and settlement expense, and, most importantly, of regulatory deficiencies and resulting unpoliced device risk.Footnote 27
20.3.3 Compensation for Injured Patients
Regarding payment of damages to injured patients rather than the government, the question, essentially, is whether patients would be better off under the conventional tort system of compensation than the proposed strict liability rule. The short answer is that under the conventional tort system the costs of the increased risk of harm far exceed the benefits of possible compensation. This would be so even if the tort system employed strict liability. Paying damages to patients would preserve subrogation reimbursement, shielding insurers from bearing total accident costs and resulting in their insureds incurring otherwise avoidable unreasonable risk of product-related accident, as well as higher insurance premiums to cover it.
Moreover, patients who suffer medical device injuries are already insured for their medical and other monetary losses from product-related injuries. Even if some patients need supplemental coverage, they surely would not willingly, let alone rationally, turn to tort liability to supply it. “Tort insurance” imposes exorbitant overhead costs – far greater than the cost for comparable coverage from public or private insurers – amounting to a dollar or more charge on every dollar recovered (before subrogation deduction).Footnote 28 Nor would risk-averse individuals, in need of insurance, willingly pay for taking the wildly variable chance of winning a lawsuit to cover pressing medical needs (for example, ICU stays for COVID-19 patients), with recovery depending not only on the fact of medical and other monetary loss (which alone suffices for true insurance) but also predominantly on the lucky alignment of such unlikely litigation contingencies as tortiously (as opposed to non-tortiously) caused injury, solvent tortfeasor, and net expected damages high enough for a competent plaintiff-lawyer to profit from taking the case.Footnote 29 Any suggestion that patients might willingly buy tort insurance coverage of nonmonetary loss is refuted by evidence showing that despite the annual expenditure of trillions on premiums and taxes worldwide for public and private insurance, no insurer provides such coverage. The reason is simple: no one is willing to pay for it.Footnote 30 On top of all of that, tort liability imposes a grossly regressive “premium” tax for coverage of risk in the price of standardized products such as medical devices. While all patients (and other consumers) pay the same premium charge in the product price, tort recoveries greatly vary according to plaintiffs’ relative wealth. This alone is sufficient to justify characterizing “tort insurance” “insurance fraud.”
20.4 Conclusion
In closing, we note several possible refinements of the proposed system for correcting the market to further strengthen insurers’ regulatory power. First, to increase operating efficiency, the system might make use of non-judicial administrative tariffs rather than judicially enforced strict manufacturer liability damage awards. Earmarking recoveries (or tariff levies) for deposit in Social Security rather than the Treasury might provide true insurance value without compromising the objective of eliminating subrogation and exposing insurers to the total monetary costs of product accidents. Finally, the proposed system could well be employed for all FDA-approved medical goods, pharmaceuticals as well as devices.
Open access