Providing Context: Medical Device Litigation and Inferior Vena Cava Filters

Abstract

Over the last few years, an increasing number of lawsuits have been filed involving inferior vena cava filters. This has prompted the U.S. Judicial Panel on Multidistrict Litigation to centralize these lawsuits into two multidistrict litigations: one for Cook's filters and one for Bard's. Both sets of cases share similar questions of facts, in particular whether these filters' design and manufacturing practices made them unreasonably prone to serious complications. The resolution of these cases will add to a larger legal debate concerning how much legal protection the 1976 Medical Device Amendments should offer firms from tort liability. As a specialty that often relies on medical devices, it is not only important for interventional radiologists to have a general understanding of medical device litigation but also to reflect upon the approaches to informed consent regarding these devices.

Objectives: Upon completion of this article, the reader will be able to describe the current status of legal claims regarding inferior vena cava filters, medical device liability, and related ethical considerations.

Accreditation: This activity has been planned and implemented in accordance with the Essential Areas and Policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint providership of Tufts University School of Medicine (TUSM) and Thieme Medical Publishers, New York. TUSM is accredited by the ACCME to provide continuing medical education for physicians.

Credit: Tufts University School of Medicine designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.

In early 2008, a woman named Lessie Tillman awoke with leg pain following knee replacement surgeries. It was soon determined that this pain was due to a deep vein thrombosis and placement of an inferior vena cava (IVC) filter was ordered to, according to Ms. Tillman, “make sure that [the clot] didn't go all the way up.”¹ A G2 retrievable filter (C.R. Bard, Inc. Murry Hill, NJ) was placed by an interventional radiologist who later explained that he always places a retrievable filter unless he is absolutely convinced that a permanent filter is needed; this is done with the understanding that a retrievable filter can become permanent if necessary.

Ms. Tillman's venous thrombosis ultimately resolved, and upon seeing a primary care practitioner later that June, she was advised to speak to a surgeon about filter retrieval. However, it was not until February 2009 that Ms. Tillman was referred to a vascular surgeon. Although retrieval was attempted, the surgeon was unsuccessful, noting the filter had tilted and become incorporated into the left renal vein due to the length of placement. Three days later, Ms. Tillman returned to the hospital, stating she felt “sick and couldn't stand up.”¹ She alleges that “they” told her she had a clot in her kidney from the attempted filter removal. Subsequent imaging revealed filter struts extending through the IVC wall, and Ms. Tillman was told that the risk of open surgery now required to remove the filter likely outweighed the risks of leaving the filter in place.

In March 2013, Ms. Tillman filed a suit against C.R. Bard, Inc. and Bard Peripheral Vascular, Inc. (Tempe, AZ) (“Bard”), asserting negligence under state law as well as liability for the filter failing to be retrieved. As she explained, “I'm not saying I have anything that's caused by the filter. I'm just saying that the filter's there, and I don't know what's going to happen down the line….” She argued that she would not have had it put in if she knew it could not be removed.¹

This case is one of an increasing number of suits questioning whether certain IVC filters possess defects that make them more likely to cause serious injury, particularly Cook's (Bloomington, IN) Celect and Günther Tulip and Bard's Recovery and G2 filters. In fact, a recent article² named IVC filters among the top 5 current mass tort cases for plaintiff lawyers to pursue. Due to the large number of similar cases, the U.S. Judicial Panel on Multidistrict Litigation decided to centralize these lawsuits into two multidistrict litigations (MDLs): 27 suits concerning Cook filters were centralized to the Southern District of Indiana in October 2014, and 15 actions plus 5 potential tag-alongs concerning Bard filters were centralized to the District of Arizona in August 2015. Despite opposition from the two companies, the council felt there were common factual questions related to complex devices, allowing resources to be conserved.^{3 4} Plaintiffs will still have to make their individual cases, but subsequent related cases are grouped under these two umbrellas. For example, the Cook MDL has grown to 156 individual cases as of November 2, 2015. The next major step is for a few representative cases from each MDL to be brought to trial, termed “bellwether trials,” to gauge the strengths and weaknesses of different claims. This has been scheduled for August 2016 for the Cook MDL but will likely not be scheduled until far into 2016 for Bard.

In light of these developments, the authors briefly discuss three related topics which are felt to be salient for interventional radiology: (1) a representative pre-trial court opinion as an example of how this topic may be analyzed by courts; (2) the larger context of medical device litigation; and (3) ethical considerations regarding informed consent and serving as an expert witness.

The Tillman Case Opinion

Only relatively few IVC filter cases have reached agreements in undisclosed settlements or had pretrial court opinions made publically available. Primarily, these opinions have concerned filter companies' motions to exclude plaintiffs' expert testimonies and motions from both parties for summary judgment, determining whether or not the case would need to proceed to trial. In March 2015, the court of the central district of Florida issued one such opinion regarding Lessie Tillman's case against Bard. This opinion provides a window into understanding the current legal environment that will shape the multidistrict cases as well as providing a starting point for a larger discussion.¹

Assessing the appropriateness of an expert testimony is often based on Daubert v. Merrell Dow Pharm, Inc. In this case, the U.S. Supreme Court explained that trial courts should act as gatekeepers, ensuring expert testimonies are both relevant and reliable. It holds that experts must be qualified on the topic, use reliable means of reaching conclusions, and their testimony must assist in determining the facts.⁵ As such, there is ample opportunity for the court to permit an expert's opinion in some areas while excluding his/her comments in others.

For example, Dr. Robert Ritchie, a distinguished professor of materials science and mechanical engineering, claimed to have examined several Bard filters. He testified that flaws in Bard's filter designs and omission of certain manufacturing steps make them more likely to fracture as well as tilt, migrate, and perforate the IVC. Although Bard agreed Ritchie was qualified, the company contended that his opinions on filter fracture did not fit the facts of the case. The court ultimately found Ritchie's opinion on design flaws to be qualified and relevant; however, it felt Ritchie's opinion that these flaws also cause tilt, migration, and perforation was an intellectual leap without sufficient evidence and excluded this portion of his opinion.

The other half of this court opinion involved deciding whether or not a “summary judgment” could be made. “Summary judgment” is a means by which a case can be decided without proceeding to trial. It requires there to be no question of facts (juries determine questions of fact), so the case is decided on interpretation of the law alone. Tillman had sought damages under what is called “strict liability,” which is negligence as a matter of law (or “negligence per se”). This relieves the plaintiff of proving a specific act of negligence and instead contends the consumer product possessed a defect making it unreasonably dangerous. This concept holds a firm strictly liable only for the period of time the product is under the firm's control. Thus, a defect can exist by virtue of design, manufacturing, or marketing (Tillman charged Bard with all three). Both filter companies' primary arguments have been what are called the regulatory compliance defense and learned intermediary doctrine. In short, the Food and Drug Administration (FDA) had approved the filter designs, and per the 1976 Medical Device Amendments (MDA), the companies should not be liable for aspects of their design and manufacturing that the FDA approved. Additionally, their duty to warn of complications only requires communication of such risks to the learned intermediaries (physicians) who use the device.

Design Defect

The Günther Tulip, G2, and Celect filter designs all received FDA approval during the last decade via an accelerated approval process (the “510(k) process”).⁶ This process allows medical devices that are “substantially equivalent” to devices on the market prior to enactment of the 1976 MDAs to gain FDA approval and be introduced to the market more rapidly—an estimated difference of 1,180 hours of work.⁷

This and the topics below relate to a debated area of law regarding negligence per se and the regulation compliance defense. The 1976 MDAs do protect medical device manufacturers from some liability. The idea is that if a firm has complied with federal regulation (e.g., those of the FDA), they should be protected from state litigation questioning what the regulation was meant to regulate. However, case law in most jurisdictions has set a precedent that regulations are a minimum, and thus the firm is not immune from tort liability if the plaintiff can prove a reasonable manufacturer would have done more.⁸

In the Tillman case, the court considered two tests to make their decision. In some areas, courts assess a design defect based on what a reasonable consumer would expect from the product; however, this test was rejected based on IVC filters being too complex for an average lay consumer to reasonably understand and form such expectations. The court then turned to a risk-utility assessment of the G2 filter, finding a convincing question of fact. Although tilt, migration, and perforation are acknowledged filter risks, Tillman's experts presented sufficient evidence of an increased tendency to tilt and migrate as well as fracture due to its geometry and lack of chamfer on the rim of the filter cap, which would have been reasonable for Bard to address. Both sides presented opposing clinical opinions on whether these risks were outweighed by the benefits, requiring a trial to resolve these differences in reasonable opinions.

Manufacturing Defect

Bard argued that Tillman had failed to provide evidence that her filter was different from any other G2 filter. To have a manufacturing defect, the product in question must (1) not conform to the original design and (2) fail to perform as safely because of this. The court agreed with Bard that Tillman failed to present evidence that the filter, as designed, should not have tilted or migrated as these are known complications. However, the court was convinced by Dr. Ritchie's observations that some G2 filters possess defects that could be easily removed with electropolishing or other quality controls. As such, the court found sufficient evidence to question whether a reasonable filter manufacturer would have done more to prevent such defects.

Marketing Defect (a.k.a. Inadequate Warning)

As a producer of a prescribed device, Bard has a duty only to sufficiently warn the learned intermediaries (physicians) of its filters' risks. The responsibility then falls on physicians to sufficiently warn patients. Furthermore, if injury occurs and the physician knew of the risk, a firm's failure to warn this physician of that risk is not the cause of injury.⁹

In this case, Bard contended it adequately warned the interventional radiologist, who supported this position, testifying that the instructions for use (IFU) accompanying the G2 filter warns of the exact scenario Tillman experienced and he always reviews such material and discloses this to patients. Additionally, Tillman had signed a consent form claiming she understood the risks of filter placement. Tillman instead argued that the IFU's language is misleading, minimizes the actual risks, and should have warned of worse performance compared with other filters. Although the court agreed that the language was misleading, it granted summary judgment in Bard's favor—the adequacy of the filter's warning would not be obvious to a lay person, and Tillman failed to produce an expert witness to argue this point.

Causation of Harm

Bard argued that there is insufficient evidence that Tillman experienced any harm as a result of her filter. The court agreed as there was no evidence that the kidney thrombosis was caused by the filter, and Tillman failed to elaborate who “they” were who told her this. However, the court was convinced that the lack of removal potentially caused by the filter design and manufacturing may expose Tillman to greater risk of subsequent harm and the need to undergo serial imaging to monitor this risk.

Larger Context of Medical Device Litigation

Although the aforementioned case provides an example of how these cases may be analyzed, it is important to realize that this case exists in an ever-changing legal environment. Much like the field of medicine, legal standards continue to evolve and rely on a degree of interpretation. To better understand this, it is worth briefly considering the larger context of medical device litigation.

In 1976, Congress enacted the MDAs partly to encourage the rapid development and adoption of lifesaving medical devices. This allowed medical device manufacturers to circumvent the onerous FDA premarket approval process by instead submitting a premarket notification to the FDA for approval. The 1976 MDAs were also meant to prevent state counts from undermining federal approval of a device, but case law has challenged this suggesting that compliance with regulations does not necessarily prove due care was taken by a firm.^{10 11}

This legal environment set the stage for the subsequent complexity of and debates surrounding medical device litigation. Consider, for instance, the Medical Device Safety Act (MDSA) of 2009 that sought to weaken the liability protection offered by the 1976 MDAs. This initiative was prompted by two U.S. Supreme Court decisions. In Riegel v. Medtronic, Inc., a physician inflated a Medtronic balloon catheter beyond its recommended pressure, causing it to rupture in one of Riegel's coronary arteries. Riegel sued based on New York common law, but the Supreme Court held in 2008 that the 1976 MDAs protected Medtronic as they had obtained FDA approval and provided adequate warning to Riegel's physician.¹² In Wyeth v. Levine, a physician injected Wyeth's antinausea drug Phenergan directly into Levine's vein, leading to subsequent necrosis, amputation, and loss of Levine's career as a musician. In 2009, the court held that Wyeth was not protected by previous approval of their label, as sufficient evidence had since arose to claim that a reasonable firm would have provided more adequate warning.¹³ Those supporting the MDSA of 2009 saw a contradiction here and felt that federal regulation and state tort law should be complementary in holding medical device firms accountable.¹⁴ Those supporting the original intensions of the 1976 MDAs instead noted how the 2008 decision only related to devices undergoing the most rigorous FDA approval process and/or emphasized how the legislation encourages medical innovation and attempts to prevent lay judges from undermining the FDA. The MDSA of 2009 ultimately died in Congress.

This tension was seen in the Tillman opinion and will likely shape the future IVC filter MDL court decisions. Interestingly, what makes matters more complicated for Bard is a recent FDA warning letter sent July 2015. The letter (which is available online) outlined several violations of federal regulations. These included not adequately receiving and evaluating complaints related to their G2 filter, and misrepresenting incidents as device malfunctions rather than death and serious injury as well as issues with several other products. This is significant as it strengthens arguments that Bard failed to take reasonable steps to address design or manufacturing defects as well as weakening their regulation compliance defense.

Ethical Considerations

Some familiarity with medical device litigation is especially salient for interventional radiology as a specialty, as it relies on an array of medical devices and is often involved in device innovation. However, it is also worth discussing some ethical considerations for interventional radiologists related to informed consent and serving as expert witnesses.

Informed consent is an ethical concept that has permeated much of medical culture but often remains surprisingly obscure and/or neglected. Its very familiarity is its downfall, and so “informed consent” can become a cursory signature or line in a progress note rather than a critical conversation or series of conversations that help establish boundaries and important context of one's therapeutic relationship—not an event but an approach. Christine Grady described this nicely in a recent article,¹⁵ noting the ongoing debate related to the scope and level of detail discussed, cultural or socioeconomic barriers, and other issues.

Legally this concept also continues to evolve. For example, a 1996 Wisconsin Supreme Court decision was one of the first to hold a physician accountable for not disclosing physician-specific characteristics. In Johnson v. Kokemoor, a relatively new physician reported a low risk (∼2%) of serious complications, comparing the operation he was about to perform to routine procedures while obtaining consent from Johnson. Unfortunately, the operation left Johnson with quadriplegia and eye/speech impairment. Johnson's expert witness stated that approximately 2% was a low estimate even for an experienced surgeon, and for Kokemoor, a more accurate risk would have been 20 to 30%. The court ultimately held that Kokemoor's experience was material information that would have caused a reasonable person to choose a different course of action, and thus ruled in favor of Johnson.¹⁶

Both clinically and legally, there is no bright red line between relevant and irrelevant information to ensure that informed consent is obtained. How should the recent concerns regarding Cook and Bard filters be discussed with patients? Should an interventional radiologist opt for an alternative filter for the time being? The authors spoke with a couple of plaintiff trial attorneys who had this advice: physicians are rarely named as defendants if they (1) stay relatively well informed, (2) carefully consider such information in their decision making, (3) clearly communicate these considerations to patients, and (4) documents their conversations. However, one attorney cautioned that the evidence brought against Cook and Bard is concerning and should be carefully considered by physicians who place these devices to ensure they are not exposing patients to undue risks.

Reasonable physicians will differ on their interpretation of some evidence, and so filter preferences will vary. Due to the professional nature of medicine, experience and expertise are given significant weight, coloring interpretation of available evidence (e.g., a physician cannot promise a patient technical success but can promise his/her expertise). That said, there are powerful virtues conveyed to a patient in saying, “this has been my experience with [device or procedure X]. These are my concerns… or experiences of my colleagues. Help me understand what you have heard about X; what are your concerns and goals?” or even “I tend to use [Firm A]'s filter and even do some consulting for them because I have had the best outcomes with their products.” If sincere and transparent, it conveys honesty and trustworthiness; the physician is willing to reflect upon and admit his/her limitations, incentives, and reasoning. It also begins to establish a prioritized loyalty—although the physician may value a relationship with a device manufacturer or their practice, they are committed first and foremost to helping that patient heal through their relationship with him/her, however brief.

The authors believe that serving as an expert witness requires similar virtues. Potential witnesses should ask themselves, “Has this been my honest experience? How have my relationships/environments affected my opinions? Am I arguing for a position which will best promote health and wellness for my community?” For example, the authors' department has recently preferred ALN Optional vena cava filters (Aln, Ghisonaccia, France) as our department has observed fewer problems with them. This should be something physicians are just as comfortable sharing in this article as with their patients or a jury.

Conclusion

As a specialty with close ties to medical devices such as IVC filters, interventional radiologists should have some familiarity with recent IVC filter litigation as well as the larger legal environment of medical device liability. Furthermore, reflecting upon conversations with patients or serving as an expert witness in light of these developments is likely worthwhile. Laws and policies often set an ethical minimum for clinicians and, as with medical device regulation, compliance does not necessarily equal appropriate behavior. Finally, it is worth emphasizing that whether one is interpreting legislation or counseling patients, context is important.