Risk Management 360: Managing Risk Through Product Recall Preparation
In their years working with life sciences companies – large and small – Medmarc’s risk managers have evaluated numerous medical devices and pharmaceuticals as well as their manufacturing and distribution operations and FDA compliance histories. From these experiences, our team of risk managers has developed a unique perspective on this innovative and growing industry. Life sciences companies operate in a heavily regulated and litigated environment. Through studying these companies’ successes, and their missteps, the authors of this series have discerned common trends that can influence a company’s products liability exposure. This series will discuss these trends and what companies can do to manage the products liability risks they create.
Managing Risk Through Product Recall Preparation
FDA’s Recall Authority
While the life sciences industry has long been in the practice of recalling products from the market after post-market discoveries of product defectsi, FDA’s authority to compel the recall of prescription products is comparatively recent. Neither the original 1906 Pure Food and Drug Act, nor the 1938 Food, Drug, and Cosmetic Act, contained a specific recall authority for the agency. Rather, through the administrative rulemaking process, FDA issued a Final Rule in June of 1978 that was intended to set forth FDA’s policy on recalls—in short, that they are the best method for removing commercially-distributed product from the market. The Final Rule was rooted in the agency’s authority to set forth recall obligations and procedures as well as to seize adulterated product and prosecute the responsible parties, but it also recognizes that the agency did not have the authority to compel a recall should a manufacturer choose not to undertake one voluntarily.ii
This gap was closed for medical devices with the passage of the 1990 Safe Medical Devices Act. Under the original 1976 Medical Device Amendments, FDA’s authority with regard to post-market device defects read more like a consumer protection statute, permitting the agency to direct a repair, replacement, or refund if a device were found to be defective. However, Congress recognized the need for an administrative mechanism that would allow FDA to direct manufacturers to protect the public health rather than merely ensure patients were not paying for defective products. The result was an expansion of the repair/replace/refund provision to include an express authority to order a manufacturer or other responsible party to cease distributing a defective device and to notify healthcare providers and patients to stop using it. Further, should the risks not be sufficiently mitigated through notice, FDA can expand upon this order to require the responsible party(ies) to recall the defective device.iii The Final Rule detailing FDA’s implementation of this authority in its regulations (see 21 CFR §810, et seq.) was published on November, 20, 1996.iv
Recall Requirements: FDA Policy
FDA’s recall regulations, set forth at 21 CFR §7 and applying to all products regulated by FDA, contain a provision titled, “General industry guidance.”v In this provision, FDA provides suggestions for steps a “prudent firm” can take to minimize the business interruption should a recall become necessary—much like a professor providing “suggestions” for how one might pass his or her class.
Though presented as guidance, these steps are critical crisis management measures companies can implement to mitigate not only their own business interruption but also to ensure that they eliminate any potential lag time in responding to a public health risk of their own creation. Firms should consider:
- Develop and implement a written contingency plan for initiating and carrying out a recall, should the need arise;
- Incorporate adequate lot identification practices into manufacturing procedures in order to accurately identify all impacted product in the event of a recall; and
- Incorporate an identification method into all purchasing, manufacturing, and distribution practices in order to ensure that they can contact downstream customers in the event of an upstream component recall and maintain the necessary records beyond the shelf life of the product.vi
Each of these measures serves to mitigate specific vulnerabilities companies face when recalling a product. Consider the case study addressed above. What if ABC Company had implemented these measures? Could a written recall plan—including templates for communication with customers and healthcare providers—have allowed ABC Company to make contact before the plaintiff was injured? Might tighter lot identification practices have allowed ABC to more accurately pinpoint the root cause, allowing them to recall fewer units?
Product recalls are not an uncommon phenomenon. According to an industry service provider that specializes in recall facilitation, there were 243 medical device recalls in the third quarter of 2019, covering nearly 219 MILLION units, a more than 1000% increase.vii And, as in our case study, software was the primary reason for medical device recalls, as it has been for 14 consecutive quarters.viii Recalls are not "something that only happens to other companies," and a robust risk management program incorporates proactive measures to ensure you are able to respond swiftly and thoroughly should the need arise.
Recall Requirements: Industry Practice
Implementing a recall contingency plan does not look the same for every entity. Much like the Quality System Regulations (QSR), which provide a framework for crafting appropriate procedures based upon a firm’s operations, a recall plan is not a one-size-fits-all endeavor. Consider two manufacturers – one, a global manufacturer of Class I consumer dental devices, like toothbrushes and dental floss, and the other, a small business that specializes in manufacturing Class III pediatric pacemakers. The recall obligations for these two entities, and thus the necessary proactive planning, could not be more different.
The dental firm, as a manufacturer of high-volume, low-risk devices, may be far less likely to conduct a recall, and if they do, it is far less likely to be for a serious public health risk. Their focus is perhaps best placed on lot identification in order to better identify impacted product and thus minimize the number of units recalled. Likewise, where the dental company will need to recall to the consumer level on a large scale, having an information dissemination plan as well as communication templates written in local languages is likely to benefit them in the event of a recall.
On the other hand, the low-volume but high-risk pediatric pacemaker manufacturer may need to focus on expediency, drafting measures that will allow them to directly contact the healthcare providers using their products and provide them with specific steps to ensure patient safety. Product identification is likely an important consideration for this firm but for different reasons than the dental firm. Rather than whittling down the recall to as few units as possible, the high-risk nature of the pacemaker requires that the manufacturer ensure all patients with potentially defective products are identified.
Recall plans are not just the jurisdiction of manufacturers. Distributors down to the clinic and retail levels need to consider both how they will comply with recalls initiated upstream from them as well as the less likely, but still possible, scenario in which the distributor’s own actions cause a product defect (e.g., contamination, temperature deviation rendering the product adulterated and/or misbranded, etc.).
Is everything a recall?
There are two types of product recalls: corrections and removals.ix Both a correction and a removal are intended to amend some violation of FDA regulations rendering the product a public health risk. The terminology differs based upon where the product remediation takes place. The recall as most people consider it, wherein product already on the market is returned to the manufacturer, is a removal. The product is physically removed from the marketplace in order to be repaired, modified, relabeled, destroyed, or whatever corrective action has been identified by the manufacturer and approved by FDA. This differs from a correction, in which the approved product remediation can be accomplished without physically relocating the product. An example of a recall that would be considered a correction is one in which a software update is sufficient and can be pushed out wirelessly to the recalled devices in their current location.
It is also possible to conduct a removal or correction that does not reach the level of a recall. This is known as a market withdrawal and can be initiated for minor violations that would not be subject to legal action by FDA, or for ordinary business purposes like stock rotation or maintenance/repair work.x Similarly, a stock recovery is a removal or correction not reaching the level of a recall because it has not yet been released for sale or left the manufacturer’s control.xi The distinction between removals and corrections reaching the level of a recall and those that are merely business practices is critical, as recalls must be reported to FDA immediately, and business practices merely need to be logged by the manufacturer or other responsible party.
|The Product||ABC Company’s extracorporeal membrane oxygenation (ECMO) device for the neonatal market, known as the BabyVent EC.
Baby Doe, a female delivered prematurely at 28 weeks, 5 days gestation, was born on March 17, 2017, weighing only 2 lbs., 12 oz. The hospital where Baby Doe was delivered did not have sufficient resources to care for very premature babies, so upon stabilization, Baby Doe was transferred to a regional hospital with a Level IV neonatal intensive care unit (NICU). Due to her prematurity, Baby Doe’s lungs were too immature to allow her to breathe adequately on her own, so she was placed on a ventilator and surfactant therapy—premature lungs do not yet produce enough of the normal secretions that allow the lungs to function, so an artificial replacement is administered—was initiated.
Baby Doe was determined to be an otherwise healthy baby, with no congenital anomalies. She progressed as expected for the first several weeks of her life, moving from the mechanical ventilator to a continuous positive airway pressure (CPAP) device as her body grew stronger. However, as her doctors attempted to wean her off of the CPAP to room air, she began struggling to maintain adequate oxygen saturation levels on her own, even after she was returned to supplemental oxygen, and her condition began to worsen. Fearing Respiratory Distress Syndrome (RDS) due to damage from the high-pressure ventilation on her premature lungs, her doctors attempted several aggressive medical strategies to reduce the stress on her lungs and improve her oxygenation levels. After these approaches failed to improve Baby Doe’s status and her organs began to show signs of deteriorating function, her medical team decided to try venovenous ECMO and, on May 7th, ordered that she be placed on the BabyVent EC.
Venovenous ECMO works by surgically placing a cannula—thin tubing—into two major veins. ECMO imitates the gas exchange process that normally takes place in the lungs by removing blood from the body through one cannula, artificially filtering out carbon dioxide and oxygenating the red blood cells, then returning the oxygenated blood to the body through the other cannula. The freshly oxygenated blood is then circulated to the organs. Artificially performing the work of the lungs outside of the body can allow the lungs time to recover from trauma. However, ECMO is by no means a standard treatment. It is an invasive, advanced life support technique and is associated with very serious potential side effects when used in preemies. These side effects include stroke and bleeding in the brain.
In Baby Doe’s case, ECMO appeared to be working. Her oxygen levels stabilized and organ function appeared to improve. However, on her fourth day of ECMO therapy, Baby Doe’s primary nurse responded to alarms signaling that Baby Doe’s vitals were plummeting, which was confusing since she was still on the BabyVent EC device. The nurse called for the NICU’s Rapid Response Team, which immediately began attempts to resuscitate Baby Doe. The Rapid Response Team was able to get Baby Doe’s heart to restart, but after she was stabilized, they discovered that the ECMO unit had completely ceased functioning. It appeared that the battery had drained in a fraction of the time it should have been able to power the unit, and rather than triggering an alarm when the unit reached low power levels, it simply shut down without notice, leaving Baby Doe without oxygen.
The hospital reported the incident to ABC Company and to FDA. Later that week, ABC Company initiated a recall of nearly 2,500 BabyVent EC units, which FDA identified as a Class I recall—the highest risk level. It was determined that the BabyVent EC had a software defect that caused it to drain the battery rapidly and shut down without notice. ABC Company prepared and deployed a software update to correct the battery issue.
In August of 2017, John and Jane Doe, on behalf of their daughter, Baby Doe, filed suit against ABC Company. Though the medical team was able to restart Baby Doe’s heart, it was later discovered that she had suffered irreversible brain damage due to the lack of oxygen when her EMCO unit failed. As a result, Baby Doe developed significant cognitive and motor deficits and would require lifelong care and therapy as well as adaptive aids and equipment.
The Does’ suit alleged that ABC Company’s defective design caused Baby Doe’s injuries, for which ABC Company is strictly liable, and demanded $25,000,000 for future medical costs, lost earning capacity, non-economic damages, and pain and suffering.
Recalls and Products Liability: Is compliance enough?
Medical devices will never be perfect. There will always be failures, and sometimes those failures will cause harm. However, the actions a company has taken prior to an incidence of product failure that causes harm can make or break the company’s defensibility in the event of a claim. In considering the facts above, we can consider various approaches companies might take in facing a scenario like that faced by ABC Company and evaluate the likely outcomes.
Ideally, ABC Company would be able to demonstrate that they had taken all possible steps to prevent injuries like Baby Doe’s from happening. They would be able to demonstrate that they are in complete compliance with FDA regulations, as evidenced by an inspection history without the issuance of Form 483 reports identifying compliance deficiencies observed by FDA investigators. If they had received 483s, then they should have thorough records of the corrective actions implemented to bring the company into compliance. Their device development records should be thorough, demonstrating that the product was sufficiently tested to allow the company to design out serious risks and to craft warnings and instructions for the appropriate use of the product. They should have procedures in place to guide their evaluation of critical suppliers and documentation to demonstrate that the procedures were followed. The interactions between ABC Company representatives and hospitals via ABC’s sales representatives and medical staff should be controlled, with the company training the representatives regarding clear parameters for their discussions about the products.
In an ideal world, ABC Company would be able to check all of these boxes. However, there are very few ideal companies in the real world. How would ABC Company’s defensibility be impacted in light of particular deficiencies in regulatory compliance and/or risk management?
- …ABC Company had received a Warning Letter from FDA regarding deficiencies in its design procedures. This document is public record, published on FDA’s website. While a Warning Letter would not necessarily prove dispositive with regard to a design defect in a specific device, it is easily deployed by plaintiff’s counsel to bolster their case.
- …ABC Company had received several complaints of similar battery issues in ECMO devices and failed to recognize their critical nature. ABC Company might have treated the complaints like an ordinary adverse event, reporting it to FDA in 30 days rather than reporting it within five days—the requirement for adverse events that could prove to be life threatening if they were to recur. Injuries occurring during that extended reporting time could be attributed to ABC Company’s failure to initiate proper warnings of a known product problem.
- …ABC Company purchased the defective software from a supplier but failed to properly qualify the supplier according to their own procedures. A savvy plaintiff’s attorney would understand the obligations life sciences companies have under FDA’s regulations, which include supplier qualification, and the attorney would be able to use this knowledge to identify a particular weakness in ABC Company’s operations.
- …ABC Company purchased the defective software from a properly qualified supplier but chose to operate on a purchase order basis rather than entering into a supply agreement with a requirement that the supplier indemnify ABC Company in the event of an injury caused by the supplier’s component. In this case, ABC Company would rightfully want to assign liability to the software vendor but may be unable to do so. However, if they had entered into a supply agreement, the indemnification provision would provide an obligation for the supplier and a clear procedure for transferring liability to the responsible party.
- …ABC Company failed to take action in initiating a recall in the face of multiple adverse event reports. If ABC had knowledge of a repetitive product problem with life-threatening implications, this information would provide plaintiffs with clear evidence of design and/or manufacturing defects and a failure to warn.
By virtue of the type of products they manufacture, life sciences companies are at a heightened risk of products liability vulnerability. Their products impact the health and lives of patients every day and they are critical to our wellbeing. However, when a life sciences product fails—and they will fail—the risk of a patient being negatively impacted is significant. It should be every life sciences firm’s goal to ensure that the products they place on the market are of the highest possible quality. It is equally important that companies prepare for the worst case scenario so that they can act quickly in the case of adverse events and product problems, thereby mitigating any potential health risks to patients.
iSee November 1959 recall of the U.S. cranberry crop due to a risk of contamination with carcinogenic weed killer. https://www.fda.gov/AboutFDA/History/FOrgsHistory/EvolvingPowers/ucm2007256.htm
ii Wood, James M., et al. Recalls of Prescription Products: A Proactive Primer. Washington, D.C.: FDLI, 2004, p. 3-4.
iii 21 U.S.C. §360h(e).
iv 61 Fed. Reg. 59,004 (Nov. 20, 1996).
v 21 CFR §7.59.
vii The number of units recalled in this quarter, and consequently the massive percentage increase over the previous quarter, is an anomaly driven by one recall that accounted for 83% of all units recalled in Q3 2019. See https://www.stericycleexpertsolutions.com/wp-content/uploads/2019/11/ExpertSolutions-RecallIndex-Q32019-LQ-WEB.pdf at p. 19.
viii Stericycle Expert Solutions (2019). Recall Index, Q3 2019. Available at: https://www.stericycleexpertsolutions.com/wp-content/uploads/2019/11/ExpertSolutions-RecallIndex-Q32019-LQ-WEB.pdf.
ix 21 CFR §7.3(g, h).
x 21 CFR §7.3(j).
xi 21 CFR §7.3(k).
For additional resources contact the Marketing department
Medmarc is a part of ProAssurance Group, a family of specialty liability insurance companies. The product material is for informational purposes only. In the event any of the information presented conflicts with the terms and conditions of any policy of insurance offered from ProAssurance, its subsidiaries, and its affiliates, the terms and conditions of the actual policy will apply.
Copyright © 2023 - Medmarc