Review of approvals and recalls of US specific medical devices in general and plastic surgery

Abstract

Background

Medical devices in the United States can be FDA approved either by the premarket notification (510(k)) or premarket approval process (PMA). Applications for PMA are more intensive than 510(k). This study aims to compare the recall rates of 510(k) and PMA General and Plastic Surgery (Surgery) devices and to educate physicians about potential flaws in the approval process.

Methods

This is a retrospective case study utilizing the public FDA Medical Device Recalls database. Inclusion criteria for this study includes devices approved by the Surgery Devices Panel via 510(k) or PMA in the United States between 11/2/2002- 11/9/2021.

Results

Out of 8,985 devices, 81% of surgery devices were approved by 510(k) while 19% were approved by PMA. The recall rate for devices approved by PMA and 510(k) was 2.3% and 11.6% respectively (p < .01). 510(k) device recalls are 5.32 times more likely. The differences in severity of recall is also significant (p < .01). From 2002 to 2021, the percent of devices that were approved by 501(k) decreased (p < .01) but rate of recalls was consistent.

Discussion and Conclusion

Surgical devices were overwhelmingly approved with 510(k) versus PMA. 510(k) applications are cheaper, faster, and less stringent. These factors may contribute to the disproportionate 510(k) approvals and the discrepancy in recall rates. Cardiology, Orthopedics, and Obstetrics and Gynecology have similar trends for devices that go through the 510(k) device approval pathway. Though the rate of 510(k) approvals in Surgery is decreasing, more must be done to ensure that the 510(k) process sufficiently minimizes potential patient risk.

Introduction

Medical device innovation plays a vital role in the advancement of healthcare and the speed at which devices are approved by the Food and Drug Administration (FDA) is extremely important for patient care. In order to market a device for use, a device must be approved, cleared, or registered [1]. While rigorous approval processes and regulation of devices are critical for patient safety, the ability to quickly approve devices is also essential in guaranteeing physician access to cutting edge resources.

In 1976, the FDA Medical Device Amendments was created as a consequence of adverse events encountered by patients using the Dalkon Shield, an intrauterine contraceptive device implanted in over 2.2 million women. The Dalkon Shield's suboptimal contraceptive effect was demonstrated with a pregnancy rate of 4.7% and an expulsion rate of 6.3%, unacceptable even in the early years of long-term reversible contraceptive use [2]. More importantly, the Dalkon shield was noted to cause potentially life-threatening complications including pelvic infection, infertility, septic pregnancy and death [3]. Recall of the device was slow and over 200,000 women were harmed by this product.

Recent recalls, such as those for the DeRoyal® procedure pack and the Allergan breast implants, highlight the importance of physician awareness regarding device recalls. The DeRoyal® procedure pack with a warming irrigation set was used by physicians to rewarm patients after surgery, irrigate the bladder, and provide more blood for hemorrhagic patients. The device was cleared by the 510(k) pathway for surgical procedures in 2008; however, in June of 2021 it was found to leak aluminum ions. This was a major defect. The symptoms of aluminum poisoning are difficult to recognize and can put patients at risk for serious adverse effects and toxicity; thus the device was recalled [4]. DeRoyal Industries sent a recall letter to customers requesting that the affected devices be returned. Similarly, the Allergan breast implant is a textured breast implant. It underwent a worldwide recall as studies found that of the 573 unique cases of breast implant-associated anaplastic large cell lymphoma in 2019, 481 were attributed to Allergan implants. The device was approved through the PMA pathway in 2006 and once approved, underwent a 10 year post approval study period until recall in 2019. Several models of the breast implants were therefore removed from the market [5]. Both the DeRoyal and Allergan recall were of the most severe class of recalls.

As a result of the FDA Medical Devices Amendments act, a risk-based regulatory process was established. Devices were categorized as Class I, II, or III. Class I devices like elastic bandages have the lowest risk potential; Class II devices, such as wheelchairs and infusion pumps, possess moderate risk; Class III devices support or sustain life and carry significant risk in the case of a malfunction. Examples of Class III devices include pacemakers, drug eluding stents, and inflatable silicone breast prosthesis [6].

In addition to device classification, the FDA established two primary regulatory pathways for devices seeking approval: the more rigorous premarket approval (PMA) process and the less involved premarket notification (510(k)). The highest risk devices, Class III, require the more rigorous PMA pathway. The PMA is the only process which technically approves a device for market. Applications for PMA device approval include submission of detailed technical information, a non-clinical laboratory studies section, and clinical investigations section. Once the FDA accepts the application for substantive review, it may be sent to a medical specialty advisory committee with knowledge of the devices currently and the existing clinical need. FDA field personnel performs a quality safety inspection of the facility, an audit of clinical study data, and a substantive review of both in-vivo and in-vitro studies [7].

Applications for 510(k) pathway are far less stringent and require only that the device is similar to a currently approved device (predicate). The FDA determines if the new device is substantially equivalent to the predicate by reviewing performance data of the new device, which can include scientific studies about the engineering, biocompatibility, and sterility [8]. The 510(k) pathway seeks to clear devices for market as opposed to approving it. Once a device manufacturer receives an FDA approval letter they are able to immediately market their product to the public.

Additionally, with the FDA Medical Devices Amendment, the FDA centralized the recall process, mandating post market adverse event reporting and surveillance. As a result, in 1993 the Manufacturer and User Facility Device Experience (MAUDE) database was created [9,10]. Postmarket surveillance is done by means of reporting adverse events via tracking systems, and primarily through postmarket studies performed by the manufacturer. Postmarket studies are used to gather additional safety information regarding a device's use with patients [11]. If the device is found to cause potential harm, a recall can be ordered. Recall, in the FDA's specific definition, means “removal or correction of a marketed product the FDA considers to be in violation of the laws it administers”[11]. Recalls fall into three classes, I, II, and III, with Class I being the most urgent and associated with greatest risk [12]. Table 1 details the different device and recall classes.

Table 1.

Descriptions and examples of the classification of devices and recalls. Device in Class III have the greatest risk to patient health while recalls in Class I are identified as the most severe recalls [5,6,12,13].

	Description	Examples in General and Plastic Surgery
Device Class
I	Devices with minimal potential harm to the patient; most are 510(k) exempt (95%) but some go through the 510(k) application process	Shoe covers, surgeon gloves, speculums
II	Mid-risk devices; most need to be approved by 510(k)	Implantable staples, N95 respirators
III	Devices with greatest risk to patients which usually support life, are implanted, or can cause severe injury if defect; must use PMA	Internal tissue adhesives, absorbable hemostatic agents
Recall Class
I	Recalls where the adverse event could cause death or severe health consequences	Fluid leaks and device breakage inside patient, dangerously incorrect chemical concentrations
II	Recalls where the adverse event could cause temporary or reversible health consequences	Degradation of product over time, missized devices
III	Recalls where the adverse event is not likely to cause health consequences	Mislabeled products, software error

While modifications have been made since 1976, the process remains far from optimal. Devices in orthopedics, cardiology, and obstetrics and gynecology demonstrate that the 510(k) and PMA pathways are not equal. For example, studies indicate that the recall rate for obstetrics and gynecology devices is 13.6 times higher for those approved via the 510(k) pathway versus PMA [13]. High risk orthopedic devices are mistakenly approved by the 510(k) pathway instead of the PMA pathway [14]. And, in cardiology, PMA devices spent, on average, 94 months on the market until recall compared to 34 months for 510(k) devices [6]. There have been reviews on recalls of medical devices in different specialties but, to date, no review on medical device recall in general and plastic surgery has been done [3,6,13,15]. This study seeks to analyze recall rates of devices that have approved/cleared through the 510(k) or PMA pathway in General and Plastic Surgery. We hypothesize that the 510(k) pathway is associated with more device recalls and patient safety could be improved through greater utilization of the PMA pathway.

Methods

This retrospective study was IRB exempt and utilized the publicly accessible FDA Medical Device Recalls database [16]. All surgical devices approved from November 1, 2002 to November 9, 2021 were reviewed. The 510(k) Premarket Notification and the Premarket Approval (PMA) databases were searched chronologically to find the number of total devices approved in the time period. Note that this includes PMA supplements, not just original PMAs, because this is classified as a new device approval per the FDA. The product codes were found using the 510(k) Premarket Notification and the Premarket Approval (PMA) databases and then the total number of devices approved was found by searching each product code. There are 588 product codes in the General and Plastic Surgery Devices Panel. A product code is a three-digit alphabetical code which makes up the general category of a device [17]. The devices under one product code have similar characteristics and functions. Each product code was also entered into the Medical Device Recalls database to identify how many devices were recalled in that product code.

Devices with regulation numbers starting with 878 (ex: 878.3300) were included in our study population as this indicates the device was regulated and evaluated by the General and Plastic Surgery Devices Panel. Only devices which had a 510(k) or PMA evaluation and review were considered. Other FDA device approval pathways, such as the de Novo process, Emergency Use Authorization, and 510(k) exempt, were excluded [18].

Both recall and device class, plus the number of devices recalled, were collected for each product code. It is important to note that though two devices may share the same product code and have the same device class, recall classes can vary. Once the number of recalls was analyzed per pathway, it was compared to the overall recall rate. Trends with device class and recall class were identified. Chi square tests were used to calculate statistical significance and a p < .05 was considered significant.

Results

From November 1, 2002 to November 9, 2021 a total of 8985 devices were approved by the General and Plastic Surgery Devices Panel, using either the 510(k) pathway or PMA. Eighty-one percent were cleared through the 510(k) pathway and 19% utilized the PMA pathway. The percent of devices approved by the 501(k) pathway did decrease over time (Fig. 1).

Fig. 1.

Percent of approved general/plastic surgery devices approved per pathway over time. Surgical devices are decreasingly cleared by the 510(k) pathway while the percent of devices cleared by PMA is increasing, p < .01.

Out of the total 7270 devices approved by the 510(k) pathway, there were 842 recall events, representing an 11.6% recall rate. The product code with the greatest number of recalls was electrosurgical accessories (product code: GEI) with 191 device recalls; the most common reason for recall in this product code was device breakage. Other products with high recall rates include laser surgical instruments, implantable staplers and surgical mesh.

The largest number of 510(k) recalls were placed in Class II (94.3%), followed by Class III (3.7%), and then Class I recalls (2.0%). Class I recalls represent the most severe of recalls. Implantable staples had the most Class I recalls (4 occurrences or 23.5%). Reasons for recall included incomplete closure of the fastener, hemorrhage, hemorrhagic shock, additional surgical intervention, and death [19], [20], [21], [22]. Other examples of Class I recalls were surgical meshes secondary to the possibility of bowel perforation, surgical gowns and drapes for potentially compromised sterility, and hemostatic bone putty found to ignite when in contact with electrocautery systems [[23], [24]].

Comparatively, there were 40 recall events for 1715 PMA approved devices representing a 2.3% recall rate for this pathway. The product with the greatest number of recalls was inflatable silicone breast prosthesis (product code: FWM). This product encompassed Class I, II and III recalls. An injectable gel designed to minimize intranasal edema and bleeding is another example of a PMA Class I recall and was removed from the market because the concentration of hyaluronic acid listed was incorrect. This resulted in skin irritation and inflammation. Many of the Class III recalls were due to mislabeling; for example, a 4–0 Black Braided Silk suture mistakenly labeled as a 4–0 Black Mono Nylon suture. Overall, 70% of the PMA recalls were Class II; 22.5% were Class III and 7.5% were Class I. Table 2 displays the most recalled 510(k) and PMA devices.

Table 2.

Most recalled surgical devices in 510(k) and PMA pathways.

	Number of Devices Recalled	Percent of Recalls
510(k)
Electrosurgical, Cutting & Coagulation & Accessories	191	22.68%
Powered Laser Surgical Instrument	77	9.14%
Staple, Implantable	49	5.82%
PMA
Prosthesis, Breast, Inflatable, Internal, Saline	7	17.50%
Implant, Dermal, For Aesthetic Use	7	17.50%
Dressing, Wound And Burn, Interactive	6	15.00%

Using the odds ratio, we found that devices approved through the 510(k) pathway are 5.32 times more at risk of recall than those that went through PMA. The difference in recalls between 510(k) and PMA devices each year are seen in Fig. 2.

Fig. 2.

Yearly breakdown of the number of 510(k) devices recalled compared to the number of PMA devices recalled. Devices using the 510(k) pathway for approval versus the premarket process are more likely to be recalled. Surgeons should be aware of the approval pathway of implantable devices before use.

Discussion

From 2002 to 2021, medical devices regulated by the General and Plastic Surgery Devices Panel were overwhelmingly approved via the 510(k) pathway (81% vs 19%, p < .001). However, the 510(k) pathway is less rigorous and informative, clinical data is not required, and manufacturing establishments are not inspected. This often leads to gaps in safety and efficacy information and poses a greater risk to patients [25].

Furthermore, some 510(k) approved devices use multiple predicates which can lead to substantial differences between the device under consideration and predecessors [26]. The 4.82 times increased recall rate in 510(k) devices supports the claim that they are less thoroughly vetted. Awareness of the differences between pathways will allow clinicians to accurately judge a device's potential benefits against its possible harm.

Pre-amendment devices, those approved pre-1976 before the classification of 510(k) and PMA was established, also play a role in the safety gaps of the 510(k) pathway, as many of those devices were grandfathered, and therefore do not have to undergo the approval process. Unfortunately, some pre-amendment devices have been put into the most severe Class I recalls. According to Hines et al., as of 2010 there were over 200 types of pre-amendment devices which had not been classified [[27], [28]]. Using the Product Classification database, there are currently 81 pre-amendment devices which have not been classified. Not only is the 510(k) lenient in its approval standards, but when reviewing pre-amendment devices, a predicate device was never formally approved through the rigorous PMA application [28]. This could lend itself to higher rates and more severe recalls for the pre-amendment devices and, more importantly, expose patients to unnecessary risk.

Increased recall rates for 510(k) approved devices are seen in other specialties as well; for example, in cardiology, high risk recall was twice as likely for a 510(k) device, and orthopedic devices approved by 510(k) were 11.5 times more likely to be recalled [6,14,15]. This further emphasizes the need for awareness in all specialties using medical devices.

Obstetrics and gynecology device recall classes were found to be independent of the approval process [13]. By contrast, we found a significant association between the approval pathways and recall class for General and Plastic Surgery devices, indicating these two variables (approval pathway and recall class) are not independent of each other. To analyze if type of approval pathway increased the likelihood of a recall in a specific class, the two-sample Z test of proportions was applied. Using the test, we found Class II recalls were more likely to have gone through the 510(k) pathway. There is a significant difference between the proportion of 510(k) devices in a Class II recall and the proportion of PMA devices in a Class II recall (p < .01). Though we were not able to test Class I or Class III due to small sample sizes, we hypothesize that the same trend follows.

We propose a few reasons for why devices continue to be approved via 510(k) rather than PMA. The relatively short turnaround time and less associated costs make it an attractive option for device companies. The FDA has set a goal to be able to review and evaluate 90% of 510(k) applications within 90 days of submission. Conversely, 295 days are allotted for decision on 90% of PMA applications [28].

The 510(k) pathway also has financial incentives for both the device company and the FDA. Currently, the standard application fee for a large company to file a 510(k) is $12,745 while the application fee for a PMA application is $374,858 [29]. The application fee covers a larger percent of the 510(k) cost than the application fee for the respective PMA cost, thus incentivizing the FDA to encourage companies to pursue the 510(k) pathway. Given the advantages, it is easy to see why there are disproportionally more 510(k) devices approved. Though we noted a decreasing trend in surgery devices approved by the 510(k) pathway over time (93% in 2002 vs 78% in 2020, p < .01); we believe there are still improvements to be made. There must be greater emphasis greater emphasis on evaluating pre-amendment devices through the PMA pathway and more discretion on which devices are appropriate for expedited approval, the recall rate can be decreased.

Though conducting clinical trials for every 510(k) device may be cost prohibitive, using tighter controls for devices which need to undergo more rigorous device approval would allow for a decrease in patient adverse events. The first transvaginal mesh (Boston Scientific's ProteGen mesh) was approved through the 510(k) pathway using abdominal/chest wall meshes as predicate devices. Though the indication and use for transvaginal meshes is substantially different, it was approved without clinical testing in under 90 days. This led patients to experience pelvic pain, bleeding, infection, etc. and though the ProteGen mesh was eventually recalled, many other devices had already been approved using the ProteGen as a predicate- leading to a downstream effect [30]. Requiring clinical trials for 510(k) devices that differ in indication/use is another way that the 510(k) device pathway can be strengthened. Another consideration would be how off-label use of devices should play a role in the device's approval or recall process as was seen in soft-tissue fillers [31].

While the 510(k) pathway allows for devices to be approved by a quicker and cheaper timeline, modifications to its application and the use of post-market studies can aid in making the process safer. According to industry manufacturers, current post-market studies are difficult to enroll in and suffer from incompleteness and underreporting [33]. With greater insight into safety of medical devices in real time use, the FDA can adjust the approval to be both quicker and safer. To address this, the FDA is pursuing the construction of the National Evaluation System of Health Technology (NEST), which is currently utilized at authorized test sites to monitor patient outcomes [34]. Additionally, implantation of NEST into electronic health records and databases is being studied [35,36].

The process of medical device approvals differs across the world, and the US FDA process is only one of many. The World Health Organization (WHO) has identified the need to harmonize approval processes and founded “The Global Harmonization Task Force”. Currently there are two major nomenclature systems which are used across the world to facilitate information of medical devices across different regional authorities [37]. The International Organization for Standardization was also developed in 1946 to regulate the safety of devices in the global market. Approximately 30% of countries have a developed regulation framework for medical devices [37]. Future studies of medical device recalls across different regulatory systems would enhance our understanding in identifying common trends of approval pathways which lead to device recalls.

Conclusions

In the US, General and Plastic Surgery devices are overwhelmingly approved using the 510(k) process, which is cheaper, faster, and less stringent. Furthermore, when classifying devices, the FDA first considers the least restrictive classification before placing the device into a stricter classification. These factors may contribute to the disproportionate number of devices being approved through the 510(k) pathway as well as increased recall rates which can lead to inappropriate classification and subsequent potential patient risk [32].

Currently, in general and plastic surgery, devices that are approved through the 510(k) pathway are undergoing more approvals and proportional recalls than their PMA counterparts. This indicates that the over utilization of the 510(k) pathway could be leading to compromised patient safety. With greater ease of post-market monitoring and the utilization of NEST, we expect device approval to evolve into a safer process. By increasing awareness of the trends in device approvals and recalls, clinicians can choose the device most appropriate for their patients.

In addition to improving the device approval process, improvements to the FDA Medical Device Recalls database would aid in more significant analysis of device recalls. A limitation of the database, and thus the study, includes the inability to filter by life cycle of the product. Thus, the analysis could include devices that were approved before 2021 but have since been recalled as well as devices that were recalled in our time period but were approved prior to 2002. Due to limitations of the database, it was difficult to decipher how devices with multiple recalls were handled. It was also difficult to adjust for total time of device on the market. Strengths of the study include a quantitative analysis of only devices in general and plastic surgery. Recall trends vary between specialties therefore specialty specific analyses allow for greater understanding of devices and their associated risks.

Call to action

We propose a call to action to address the flaws of the 510(k) device process. Though recalls are inevitable, a more rigorous device approval process, especially in 510(k) devices approved via pre-amendment devices, will decrease the possibility of device flaws. Furthermore, improving upon the FDA Device Recall database would allow for further studies of devices of their approval process. Lastly, we agree with Rosh J, et al.: "Surgeons who implant medical devices … should rely on their knowledge of the peer-reviewed literature to inform their decisions about the risks and benefits of using these products."³⁰ Educating surgeons about the device approval process will help them to understand the benefits and risks of the device and allow them to make an educated decision on which devices to use on patients.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.