Risk of Recall Among Medical Devices Undergoing US Food and Drug Administration 510(k) Clearance and Premarket Approval, 2008-2017

Cardiac device creativity and innovation under constraints: Exploring trends from the food and drug administration's device clearances and recalls

BACKGROUND

Medical device expenditures have increased in the 21st century, with cardiac devices comprising an outsized portion of the market. Meanwhile, the disproportionate share of FDA recalls of cardiac devices is often overshadowed. Using the FDA 510(k) premarket notification pathway and FDA recalls issued from 2000 to 2020, this project seeks to engage our understanding of innovation and recalls in the cardiac device space.

METHODS

510(k) premarket notification submission dates, outcomes, and recalls from 1/1/2000 to 12/31/2019 were obtained from publicly available FDA data as a function of cardiac device innovation. We compared the annual number of 510(k) premarket clearances and FDA recalls from 1/1/2000 to 12/31/2009 to 1/1/2010 to 12/31/2019.

RESULTS

We found 343 moderate risk cardiac medical devices cleared for sale between the years 2000 and 2020. Comparing the last 10 years of the study period to the first, the yearly number of cleared devices decreased 39.7 %, from 21.4 to just 12.9 (p = 0.0019), defying positive trends in U.S. GDP and healthcare expenditures. Meanwhile, the number of FDA recalls issued for these devices increased 94.5 % from 7.3 to 14.2 recalls per year (p = 0.031). 215 device recalls were issued; 78 % Class II and 16 % Class I which constitute serious, potentially fatal recalls.

CONCLUSIONS

While United States healthcare spending continues to trend upward, there was a distinct decrease in the number of new and updated cardiac devices entering the market between 2000 and 2020. Meanwhile, recalls of these devices have uncomfortably increased. Together, these trends suggest cardiac device innovation has become risk averse.

Modernizing Medical Device Regulation: Challenges and Opportunities for the 510(k) Clearance Process

Nearly all medical devices reviewed by the U.S. Food and Drug Administration (FDA) are authorized via the 510(k) clearance process. Established in 1976, this review pathway bases authorizations on the comparability of new devices to previously authorized devices ("predicates"). This evaluation usually does not require clinical evidence of safety and effectiveness. Advocates of the 510(k) clearance process tout its support for device innovation and rapid market access, and critics of the 510(k) clearance process express that it may inadequately protect patient safety. In September 2023, the FDA issued 3 guidance documents that, if finalized, would significantly change medical device regulation. This article provides clinical and regulatory context for the proposed guidance documents, which focus on predicate selection, clinical testing requirements, and implantable devices, and identifies opportunities for further reforms that promote transparency and patient safety.

Challenges and Potential of Artificial Intelligence in Neuroradiology

PURPOSE

Artificial intelligence (AI) has emerged as a transformative force in medical research and is garnering increased attention in the public consciousness. This represents a critical time period in which medical researchers, healthcare providers, insurers, regulatory agencies, and patients are all developing and shaping their beliefs and policies regarding the use of AI in the healthcare sector. The successful deployment of AI will require support from all these groups. This commentary proposes that widespread support for medical AI must be driven by clear and transparent scientific reporting, beginning at the earliest stages of scientific research.

METHODS

A review of relevant guidelines and literature describing how scientific reporting plays a central role at key stages in the life cycle of an AI software product was conducted. To contextualize this principle within a specific medical domain, we discuss the current state of predictive tissue outcome modeling in acute ischemic stroke and the unique challenges presented therein.

RESULTS AND CONCLUSION

Translating AI methods from the research to the clinical domain is complicated by challenges related to model design and validation studies, medical product regulations, and healthcare providers' reservations regarding AI's efficacy and affordability. However, each of these limitations is also an opportunity for high-impact research that will help to accelerate the clinical adoption of state-of-the-art medical AI. In all cases, establishing and adhering to appropriate reporting standards is an important responsibility that is shared by all of the parties involved in the life cycle of a prospective AI software product.

Are Surgeons Going to Be Left Holding the Bag? Incisional Hernia Repair and Intra-Peritoneal Non-Absorbable Mesh Implant Complications

Ventral incisional hernias are common indications for elective repair and frequently complicated by recurrence. Surgical meshes, which may be synthetic, bio-synthetic, or biological, decrease recurrence and, resultingly, their use has become standard. While most patients are greatly benefited, mesh represents a permanently implanted foreign body. Mesh may be implanted within the intra-peritoneal, preperitoneal, retrorectus, inlay, or onlay anatomic positions. Meshes may be associated with complications that may be early or late and range from minor to severe. Long-term complications with intra-peritoneal synthetic mesh (IPSM) in apposition to the viscera are particularly at risk for adhesions and potential enteric fistula formation. The overall rate of such complications is difficult to appreciate due to poor long-term follow-up data, although it behooves surgeons to understand these risks as they are the ones who implant these devices. All surgeons need to be aware that meshes are commercial devices that are delivered into their operating room without scientific evidence of efficacy or even safety due to the unique regulatory practices that distinguish medical devices from medications. Thus, surgeons must continue to advocate for more stringent oversight and improved scientific evaluation to serve our patients properly and protect the patient-surgeon relationship as the only rationale long-term strategy to avoid ongoing complications.

FDA-cleared artificial intelligence and machine learning-based medical devices and their 510(k) predicate networks

The US Food and Drug Administration is clearing an increasing number of artificial intelligence and machine learning (AI/ML)-based medical devices through the 510(k) pathway. This pathway allows clearance if the device is substantially equivalent to a former cleared device (ie, predicate). We analysed the predicate networks of cleared AI/ML-based medical devices (cleared between 2019 and 2021), their underlying tasks, and recalls. More than a third of cleared AI/ML-based medical devices originated from non-AI/ML-based medical devices in the first generation. Devices with the longest time since the last predicate device with an AI/ML component were haematology (2001), radiology (2001), and cardiovascular devices (2008). Especially for devices in radiology, the AI/ML tasks changed frequently along the device's predicate network, raising safety concerns. To date, only a few recalls might have affected the AI/ML components. To improve patient care, a stronger focus should be placed on the distinctive characteristics of AI/ML when defining substantial equivalence between a new AI/ML-based medical device and predicate devices.

The 510(k) Third Party Review Program: Promise and Potential

Every year, the Food and Drug Administration (FDA) clears approximately 3,000 medical devices for marketing via the 510(k) pathway. These constitute 99% of all devices approved for human use and includes the premarket review of many devices incorporating newer technology such as artificial intelligence (AI), machine learning (ML), and other software. As the complexity of these novel technologies and the number of applications is expected to increase in the coming years, statutory changes such as the 2016 21st Century Cures Act, regulations, and guidance documents have increased both the volume and complexity of device review. Thus, the ability to streamline the review of less complex, low-to-moderate risk devices through the 510(k) pathway will maximize the FDA's capability to address other important, future-oriented regulatory questions. For over twenty five years, third party review organizations have served a defined function to assist with the review of 510(k) applications for a set of enumerated device classes. This paper reviews the history of FDA device regulation, the evolution of the 510(k) review pathway, and the recent history of the 510(k) third party review program. Finally, the paper addresses policy concerns from all stakeholders - including the FDA - along with policy suggestions to improve the third party review program and FDA device regulation writ large.

Characterization of US Food and Drug Administration Class I Recalls from 2018 to 2022 for Moderate- and High-Risk Medical Devices: A Cross-Sectional Study

Background

Medical device recalls are initiated in response to safety concerns. Class I (highest severity) recalls imply a reasonable likelihood of serious adverse events or death associated with device use. Recalled devices must be identified, assessed, and corrected or removed, upon which a recall can be terminated.

Objective

To characterize Class I medical device recalls and corresponding recalled devices.

Methods

This was a cross-sectional study of Class I recalls posted on the Food and Drug Administration's annual log from January 1, 2018 to June 30, 2022 for moderate-risk and high-risk medical devices. Devices were categorized by therapeutic use, need for implantation, and life-sustaining designation; recalls were categorized by reason, status, and time elapsed.

Results

There were 189 unique Class I medical device recalls, including 151 (79.9%) for moderate-risk and 34 (18.0%) for high-risk devices. Sixty-five (34.4%) recalls were for cardiovascular devices, 36 (19.0%) for implanted devices, and 37 (19.6%) for life-sustaining devices. The median number of device units recalled in the US per recall notice was 4620 (interquartile range [IQR], 578-42,591), with 11 (5.8%) recalls associated with more than 1 million device units. Overall, 125 (66.1%) devices had multiple recalls, with a median of 4 (IQR, 3-11) recalls issued per recalled device. As of September 15, 2022, 50 (26.5%) recalls were terminated, with a median of 24 (IQR, 17.3-30.8) months elapsed between recall initiation and termination. Recalls were terminated more commonly among devices recalled once compared to those recalled multiple times (36.2% vs 19.2%; p=0.02) and for recalls that recommended discontinuing further use of affected devices compared to those that recommended device assessment and/or education of affected population (31.8% vs 18.2%; p=0.04).

Conclusion

High-severity medical device recalls are common and affect millions of device units annually in the US. Recall termination takes a significant amount of time, putting patients at risk for serious safety concerns.

A Review on Electronic Health Record Text-Mining for Biomedical Name Entity Recognition in Healthcare Domain

Biomedical-named entity recognition (bNER) is critical in biomedical informatics. It identifies biomedical entities with special meanings, such as people, places, and organizations, as predefined semantic types in electronic health records (EHR). bNER is essential for discovering novel knowledge using computational methods and Information Technology. Early bNER systems were configured manually to include domain-specific features and rules. However, these systems were limited in handling the complexity of the biomedical text. Recent advances in deep learning (DL) have led to the development of more powerful bNER systems. DL-based bNER systems can learn the patterns of biomedical text automatically, making them more robust and efficient than traditional rule-based systems. This paper reviews the healthcare domain of bNER, using DL techniques and artificial intelligence in clinical records, for mining treatment prediction. bNER-based tools are categorized systematically and represent the distribution of input, context, and tag (encoder/decoder). Furthermore, to create a labeled dataset for our machine learning sentiment analyzer to analyze the sentiment of a set of tweets, we used a manual coding approach and the multi-task learning method to bias the training signals with domain knowledge inductively. To conclude, we discuss the challenges facing bNER systems and future directions in the healthcare field.

Risk of Recall for Total Joint Arthroplasty Devices Over 10 Years

BACKGROUND

Orthopaedic devices comprise nearly 20% of devices on the market and 12% to 20% of these devices undergo a recall within 10 years. More than 95% of these devices are approved without supporting clinical data through the Food and Drug Administration's 510(k) pathway. The risk of recall of orthopaedic arthroplasty devices approved through the 510(k) pathway has not been previously studied.

METHODS

The FDA 510(k) database was queried for orthopaedic devices approved between January 01, 2008 and December 31, 2018 and subsequently codified to hip and knee arthroplasty devices using product codes. The database included 904 arthroplasty devices during the study period, with hip and knee making up 53.7% (485) and 46.3% (419) of devices, respectively. Information regarding numbers, dates, and reasons for recall were recorded. Cumulative incidence function was conducted to compare the risk of recall between hip and knee arthroplasty.

RESULTS

In total, 94 (19.4%) hip and 85 (20.3%) knee devices were recalled. The hazard of recall by 10 years for hip and knee arthroplasty devices was approximately 24%, with no statistical differences between each region. The most common causes of recall were process control and device design, accounting for 29.6% and 26.3% of recalls, respectively, with no significant difference between study groups.

CONCLUSION

The risk of recall for arthroplasty devices is more than that previously understood. Improved postmarket surveillance strategies along with increased physician participation in detecting and reporting device safety issues are necessary to strengthen patient safety.

Use of Recalled Devices in New Device Authorizations Under the US Food and Drug Administration's 510(k) Pathway and Risk of Subsequent Recalls

IMPORTANCE

In the US, nearly all medical devices progress to market under the 510(k) pathway, which uses previously authorized devices (predicates) to support new authorizations. Current regulations permit manufacturers to use devices subject to a Class I recall-the FDA's most serious designation indicating a high probability of adverse health consequences or death-as predicates for new devices. The consequences for patient safety are not known.

OBJECTIVE

To determine the risk of a future Class I recall associated with using a recalled device as a predicate device in the 510(k) pathway.

DESIGN AND SETTING

In this cross-sectional study, all 510(k) devices subject to Class I recalls from January 2017 through December 2021 (index devices) were identified from the FDA's annual recall listings. Information about predicate devices was extracted from the Devices@FDA database. Devices authorized using index devices as predicates (descendants) were identified using a regulatory intelligence platform. A matched cohort of predicates was constructed to assess the future recall risk from using a predicate device with a Class I recall.

MAIN OUTCOMES AND MEASURES

Devices were characterized by their regulatory history and recall history. Risk ratios (RRs) were calculated to compare the risk of future Class I recalls between devices descended from predicates with matched controls.

RESULTS

Of 156 index devices subject to Class I recall from 2017 through 2021, 44 (28.2%) had prior Class I recalls. Predicates were identified for 127 index devices, with 56 (44.1%) using predicates with a Class I recall. One hundred four index devices were also used as predicates to support the authorization of 265 descendant devices, with 50 index devices (48.1%) authorizing a descendant with a Class I recall. Compared with matched controls, devices authorized using predicates with Class I recalls had a higher risk of subsequent Class I recall (6.40 [95% CI, 3.59-11.40]; P<.001).

CONCLUSIONS AND RELEVANCE

Many 510(k) devices subjected to Class I recalls in the US use predicates with a known history of Class I recalls. These devices have substantially higher risk of a subsequent Class I recall. Safeguards for the 510(k) pathway are needed to prevent problematic predicate selection and ensure patient safety.

Association Between Regulatory Submission Characteristics and Recalls of Medical Devices Receiving 510(k) Clearance

IMPORTANCE

Most regulated medical devices enter the US market via the 510(k) regulatory submission pathway, wherein manufacturers demonstrate that applicant devices are "substantially equivalent" to 1 or more "predicate" devices (legally marketed medical devices with similar intended use). Most recalled medical devices are 510(k) devices.

OBJECTIVE

To examine the association between characteristics of predicate medical devices and recall probability for 510(k) devices.

DESIGN, SETTING, AND PARTICIPANTS

In this exploratory cross-sectional analysis of medical devices cleared by the US Food and Drug Administration (FDA) between 2003 and 2018 via the 510(k) regulatory submission pathway, linear probability models were used to examine associations between a 510(k) device's recall status and characteristics of its predicate medical devices. Public documents for the 510(k) medical devices were collected using FDA databases. A text extraction algorithm was applied to identify predicate medical devices cited in 510(k) regulatory submissions. Algorithm-derived metadata were combined with 2003-2020 FDA recall data.

EXPOSURES

Citation of predicate medical devices with certain characteristics in 510(k) regulatory submissions, including the total number of predicate medical devices cited by the applicant device, the age of the predicate medical devices, the lack of similarity of the predicate medical devices to the applicant device, and the recall status of the predicate medical devices.

MAIN OUTCOMES AND MEASURES

Class I or class II recall of a 510(k) medical device between its FDA regulatory clearance date and December 31, 2020.

RESULTS

The sample included 35 176 medical devices, of which 4007 (11.4%) were recalled. The applicant devices cited a mean of 2.6 predicate medical devices, with mean ages of 3.6 years and 7.4 years for the newest and oldest, respectively, predicate medical devices. Of the applicant devices, 93.9% cited predicate medical devices with no ongoing recalls, 4.3% cited predicate medical devices with 1 ongoing class I or class II recall, 1.0% cited predicate medical devices with 2 ongoing recalls, and 0.8% cited predicate medical devices with 3 or more ongoing recalls. Applicant devices citing predicate medical devices with 3 or more ongoing recalls were significantly associated with a 9.31-percentage-point increase (95% CI, 2.84-15.77 percentage points) in recall probability compared with devices without ongoing recalls of predicate medical devices, or an 81.2% increase in recall probability relative to the mean recall probability. A 1-SD increase in the total number of predicate medical devices cited by the applicant device was significantly associated with a 1.25-percentage-point increase (95% CI, 0.62-1.87 percentage points) in recall probability, or an 11.0% increase in recall probability relative to the mean recall probability. A 1-SD increase in the newest age of a predicate medical device was significantly associated with a 0.78-percentage-point decrease (95% CI, 1.29-0.30 percentage points) in recall probability, or a 6.8% decrease in recall probability relative to the mean recall probability.

CONCLUSIONS AND RELEVANCE

This exploratory cross-sectional study of 510(k) medical devices cleared by the FDA between 2003 and 2018 demonstrated significant associations between 510(k) submission characteristics and recalls of medical devices. Further research is needed to understand the implications of these associations.

Orbital Implants Receiving Food and Drug Administration Premarket Notification

PURPOSE

This perspective explores the Food and Drug Administration (FDA) 510(k) program, occasionally referred to as premarket notification, which facilitates faster marketing of Class II medical devices, such as orbital implants by demonstrating "substantial equivalence" to previously approved devices. This allows for FDA clearance, rather than FDA approval of orbital implants via comparison to currently marketed implants rather than clinically proven safety standards.

METHODS

Utilizing the FDA's publicly available 510(k) Premarket Notification database, we conducted a thorough search of FDA-cleared orbital implants dating back to the inception of the 510(k) process in 1976.

RESULTS

We found that 29 orbital implants received 510(k) FDA clearance between 1987 and 2022. Four of the 29 implants were recalled. Only 9 of 29 implants had available data on their predicate or comparison devices; of these 9, 3 implants received clearance based on devices that were subsequently recalled.

CONCLUSIONS

This investigation into premarket approval of orbital implants identifies a shortcoming in the FDA 510(k) approval process. Long-term implant-associated morbidity is difficult to predict during premarket analysis but is further complicated for 510(k) cleared implants since devices approved based on substantial equivalence to recalled devices may not be automatically recalled. Clinicians should be aware of the approval process for the devices they select, and review of the 510(k) process, especially as it applies to substantial equivalence to devices subsequently recalled is warranted.

Estimated Cost of Developing a Therapeutic Complex Medical Device in the US

IMPORTANCE

The US medical device market is the world's largest, but estimates of the cost to bring a medical device to market are not available to help inform policy making and regulatory efforts to enhance device safety and innovation.

OBJECTIVE

To estimate the mean expected capitalized cost of developing a novel therapeutic complex medical device.

DESIGN, SETTING, AND PARTICIPANTS

In this economic evaluation, an analytical model of novel therapeutic complex medical device development using data from public and proprietary sources with coverage from 2000 through 2018 was used to estimate the cost, duration, and phase transition success probability associated with each stage of development. Data analysis was completed in September 2021.

EXPOSURES

Conduct of nonclinical and clinical studies; payment of FDA user fees for novel therapeutic complex medical devices.

MAIN OUTCOMES AND MEASURES

Mean development cost (in 2018 US dollars) incurred by developers for an FDA-approved novel therapeutic complex medical device, accounting for failures and cost of capital.

RESULTS

In this economic analysis, the mean development cost for a novel therapeutic complex medical device was $54 million (95% CI, $25 million-$200 million) excluding any postapproval studies that might be required. After accounting for the cost of failed studies and cost of capital, the mean capitalized cost of bringing a novel therapeutic complex medical device to the US market was $522 million (95% CI, $205 million-$3382 million). The key factors associated with this cost were the phase transition probabilities: 46.9% for nonclinical to feasibility study, 48.0% for feasibility to pivotal study, 75.7% pivotal study to FDA premarket approval submission, and 80.5% for FDA premarket approval submission to approval. The nonclinical development stage constituted the largest portion of overall cost at 85.0% with the FDA review stage with the highest phase transition probability accounting for only a small fraction at 0.5%.

CONCLUSIONS AND RELEVANCE

In this economic evaluation study, the cost of therapeutic complex medical device development from proof of concept through postapproval stages was assessed accounting for the cost of failures and the cost of capital. Existing estimates did not account for all stages of development, capitalization, or failure costs, which this study suggests were substantial.

The FDA and Ensuring Safety and Effectiveness of Devices, Biologics, and Technology

Orthopaedic devices account for nearly 20% of all devices on the market, with more than 600 novel orthopaedic devices cleared or approved by the FDA for marketing in the United States annually. Advances in technology and biologic therapies offer tremendous potential for patients with musculoskeletal ailments; however, it is important that the safety and effectiveness of these products be assessed to safeguard the public health. The FDA uses multiple different premarket pathways for devices, biologics, and combination products based on perceived risk of the novel product. More than 97% of orthopaedic devices go through the FDA's 510(k) pathway, which does not require clinical trials. The remaining high-risk devices must receive premarket approval and submit clinical trial data demonstrating safety and effectiveness. Similarly, high-risk biologics must obtain a biologics license application by submitting clinical trial data. Postmarketing surveillance strategies, including extended clinical trials or real-world evidence from registries, are increasingly being relied on by the FDA to expedite approval while also improving its capacity to identify problematic products.

Recalls of Moderate- and High-Risk Otolaryngologic Devices Approved by the US Food and Drug Administration, 2003-2019

OBJECTIVE

The US Food and Drug Administration (FDA) regulates the marketing of medical devices based on the premarket approval (PMA) or 510(k) pathway. We investigated the relationship among the regulatory pathway of otolaryngologic devices, the number of recalls, and the recall characteristics.

STUDY DESIGN

Retrospective cross-sectional analysis.

SETTING

Publicly available FDA databases.

METHODS

The FDA Ear, Nose, and Throat Devices Panel database was queried for 510(k) clearances and PMA approvals from 1976 to 2019. Device recalls from 2003 to 2019 were reviewed. Devices were then categorized by subspecialty, type, supporting evidence, and PMA supplement type. Logistic regression characterized the odds of recall for each device type and subspeciality.

RESULTS

A total of 1061 (57.8%) 510(k) and 778 (42.3%) PMA device applications and modifications were approved. There were 120 (11.3%) recall events associated with 42 unique otolaryngologic devices cleared via the 510(k) pathway, as compared with 25 (3.2%) recall events for 5 unique PMA devices. 510(k) device approvals were more likely to be recalled than PMA device approvals (odds ratio, 3.67; 95% CI, 2.38-5.88; P < .0001). 510(k) surgical devices (odds ratio, 2.1; 95% CI, 1.1-4.4; P = .03) were more likely to be recalled than diagnostic devices. Devices designated for laryngology (70.0%) and general otolaryngology (25.0%) composed the majority of recalls.

CONCLUSION

Otolaryngologic devices approved by the FDA via the 510(k) pathway exhibit a higher number of recalls than the PMA pathway. Given the balance between regulation and facilitating innovation, postmarket surveillance and ongoing regulatory improvements are critical to ensure optimal safety of medical devices.

Renewing the Call for Reforms to Medical Device Safety-The Case of Penumbra

IMPORTANCE

Strengthening premarket and postmarket surveillance of medical devices has long been an area of focus for health policy makers. The recent class I recall (the most serious of the US Food and Drug Administration [FDA] recalls) of reperfusion catheters manufactured by Penumbra, a US-based medical device company, illustrates issues of device safety and oversight that mandate attention.

OBJECTIVES

To review the regulatory history and clinical evidence of the Penumbra JET 7 Reperfusion Catheter with Xtra Flex Technology (JET 7) and use the device recall as a case study of the challenges associated with clinical evaluation, transparency, and oversight of medical devices in the US.

EVIDENCE

Regulatory history and clinical evidence for the Penumbra medical devices were analyzed through a qualitative review of decision letters in the Access FDA database for medical devices and medical device reports in the Manufacturer and User Facility Device Experience database and a review of market data (eg, earnings calls, company communications) and clinical literature.

FINDINGS

The JET 7 device was subjected to a class I recall following more than 200 adverse event reports, 14 of which involved patient deaths. Regulatory analysis indicated that each of the Penumbra reperfusion catheters was cleared under the 510(k) pathway (which allows devices to be authorized with limited to no clinical evidence), with limited submission of either new clinical or animal data. Clinical evidence for Penumbra devices was generated from nonrandomized, single-arm trials with small sample sizes. The regulatory issues raised by JET 7 are reflective of broader challenges for medical device regulation. Opportunities for reform include strengthening premarket evidence requirements, requiring safety reporting with unique device identifiers, and mandating active methods of postmarket surveillance.

CONCLUSIONS AND RELEVANCE

The case study of JET 7 highlights the long-standing gaps in medical device oversight and renews the impetus to build on the Institute of Medicine recommendations and reform FDA medical device regulation to protect public health.