Cooley v. C.R. Bard, Inc.

Opinion

LINDA COOLEY, Case No.: 3:22-cv-1754-MMA-KSC

Plaintiff, ORDER REGARDING DISCOVERY v. DISPUTE

C.R. BARD, INC. et al.,

Defendants.

I. Introduction On May 17, 2023, counsel for all parties called Chambers with a discovery dispute. See Doc. No. 23. Counsel spoke with the Court’s staff at great length, and the Court set the matter for hearing with directions to lodge the discovery requests at issue for the Court’s review. Id. The Court held a telephonic hearing with counsel on June 7, 2023. Doc. No. 24. Based on the information provided to the Court’s staff, the parties’ lodgments, and the arguments made during the hearing, the Court understands plaintiff seeks an order compelling further responses from defendants to plaintiff’s first Request for Production (“RFP”) numbers 7, 8, and 11. The Motion is granted in part and denied in part as explained in this Order. //// II. Factual Background This is a personal injury case, one of thousands of products liability cases filed against defendants C.R. Bard, Inc. and Bard Peripheral Vascular, Inc. (collectively “Bard”) based on Bard’s development and sale of retrievable inferior vena cava (“IVC”) filters. See generally Doc. No. 1 ¶¶ 9-59. Plaintiff’s doctor implanted Bard’s first generation IVC filter, the Recovery filter, in plaintiff in 2005. Id. ¶ 60. In the intervening years, Bard introduced five successive generations of its IVC filters into the market: the G2 Filter, the G2 Express Filter, the Eclipse Filter, the Meridian Filter, and, most recently, the Denali Filter. Only the Denali filter remains on the market. In November 2020, plaintiff discovered the filter had fractured. See id. At present, plaintiff alleges fragments of the filter have lodged themselves in her heart, lung, and spine. See id. The vast majority of the Bard IVC filter cases, which involved every generation of Bard filter, were consolidated for common discovery purposes in a Multi District Litigation (“MDL”) proceeding before Judge David G. Campbell in the District of Arizona. See In re Bard IVC Filters Products Liability Litigation, No. MDL 15-2461; see also Doc. No. 15-1. MDL discovery closed in 2017, and Judge Campbell issued a comprehensive remand order, the final version of which the parties to this case attached to their Joint Discovery Plan on April 26, 2021. See Doc. No. 15-1. Because this case was not filed until the MDL had closed and the final remand order issued, it was not eligible for inclusion in the MDL proceedings. Accordingly, this Court is not bound by the MDL orders. At the same time, the Court considers the MDL proceedings and orders entered therein highly persuasive because Judge Campbell, as the judge presiding over the MDL, had great familiarity with the facts surrounding the Bard IVC filter litigation and the applicable law. His orders reflect careful examination of the arguments advanced by all parties, resulting in sound determinations. The Court thus believes it prudent to avoid reinventing the wheel to the greatest extent practical here. Fortunately, in an admirable show of common sense, the parties to this matter generally concur, as they have agreed between themselves to make use of the MDL common discovery, and to limit local discovery in this case (as much as is practicable) to case specific issues. The parties, however, do not agree in all respects, which brings us to the narrow disputes now at issue which the Court must resolve. III. Analysis of the Parties’ Discovery Dispute Plaintiff moves to compel further responses to RFP Nos. 7, 8, and 11. A party seeking discovery may move the Court to issue an order compelling production. Fed. R. Civ. P. 37(a). This Court has broad discretion to permit or deny discovery. Hallett v. Morgan, 296 F.3d 732, 751 (9th Cir. 2002). Discovery must be “relevant to any party's claim or defense and proportional to the needs of the case.” See Fed. R. Civ. P. 26(b)(1). Ninth Circuit case law does not clearly answer the question of whether the party seeking discovery bears an initial burden of demonstrating the relevance of that discovery, or whether the party resisting discovery must make a showing of irrelevance to sustain an objection. See Fei Fei Fan v. Yan Yao Jiang, 2023 U.S. Dist. LEXIS 6544, at *5-6 (D. Nev. Jan. 13, 2023); V5 Techs v. Switch, Ltc., 334 F.R.D. 306, 309-10 (D. Nev. 2019). It is settled, however, that if the information sought is relevant, the party resisting discovery bears the ultimate burden of convincing the Court that the discovery sought should not be permitted. See V5 Techs, 334 F.R.D. at 309 (citing Blankenship v. Hearst Corp., 519 F.2d 418, 429 (9th Cir. 1975)). The essential nexus of the parties’ disagreement over the RFPs at issue is whether they are relevant and proportional to this matter, particularly in light of the vast swaths of common discovery available to the parties following the MDL. The Court will address each RFP in turn. (A) Request for Production No. 7 Plaintiff’s RFP No. 7 states:

Produce all DOCUMENTS AND ESI EVIDENCING or RELATING to any request made by the Food and Drug Administration for YOU to conduct post- market surveillance of the Recovery Filter, G2 Filter, G2 Express Filter, Eclipse Filter, Meridian Filter and/or the Denali Filter; and DOCUMENTS and ESI, including communications, plans, reports, or other information YOU submitted to the Food and Drug Administration in response.

Counsel for the parties explain this request concerns the Predicting the Safety and Effectiveness of Inferior Vena Cava Filters (“PRESERVE”) study, which observed the relative safety and effectiveness of multiple IVC filters from multiple manufacturers, including the Denali filter produced by Bard. The PRESERVE study was instigated, at least in part, at the insistence of the FDA, which plaintiff’s counsel describes as an “unusual” circumstance that demonstrates the dangerous nature of removable IVC filters generally, and Bard’s products in particular. The study itself has been published, and the results are available to the public.1 Plaintiff does not dispute that the results of the study are available to her, and she does not seek to compel Bard to produce the actual study. Rather, she contends Bard should produce all correspondence between Bard and the FDA, or between Bard and any other parties related to the study; and Bard should also produce documents created and circulated internally that pertain to the PRESERVE study. The main thrust of plaintiff’s argument is that subsequent generations of Bard’s IVC filters, including the Denali filter, can be offered into evidence as reasonable and feasible alternative designs to the Recovery filter implanted in plaintiff. As a threshold issue, Bard argues the study, and all ancillary documents, are irrelevant to plaintiff’s claims because plaintiff was implanted with the first-generation Recovery filter, which had been off the market for more than a decade prior to the introduction of the Denali filter that was the subject of the PRESERVE study. Bard suggests plaintiff has all the information she needs because she has access to the study, and her expert witnesses can use and refer to the PRESERVE study in their reports and when they testify. As argued, collecting, reviewing, and producing ancillary documents that

1 The Court was easily able to access the results of the study at relate to the PRESERVE study is disproportionate to the needs of the case given the low probative value of anything likely gleaned from any such documents. Bard argues that Judge Campbell, who provided over the MDL and had the greatest familiarity with the common discovery conducted in that proceeding, issued exactly such an order in his own cases. See Kathleen A. Shea v. C.R. Bard, Inc. et al., 2:16-cv-03088-DGC, Doc. No. 16 at 3 (D. Ariz. August 7, 2020). Plaintiff may well be able to offer the design of the Denali filter as a reasonable, feasible alternative to the original Recovery filter; and if that is the case, the PRESERVE study may be a useful repository of information on which plaintiff’s expert(s) may rely. But having access to the study itself provides plaintiff with sufficient information. Expanding discovery to Bard’s participation in the PRESERVE study is minimally relevant, if at all, to this case because the PRESERVE study exclusively tracked patients implanted with the Denali filter, which is five generations removed from the filter placed in plaintiff. Plaintiff has not convinced the Court that the production of Bard’s ancillary communications (whether internal or with third parties) will do anything to advance the merits of the case. Rather, the discovery sought is not proportional to the needs of this case. The Motion to Compel further responses to RFP No. 7 is therefore DENIED. (B) Request for Production No. 8 RFP No. 8 states: Produce all DOCUMENTS AND ESI EVIDENCING or referring to all data analysis or trends of adverse events that were reported to YOU regarding the Recovery Filter, G2 Filter, G2 Express Filter, Eclipse Filter, Meridian Filter and/or the Denali Filter, including any studies, research or documents prepared to reflect any analysis or trend.

This RFP essentially seeks discovery of any documentary evidence in Bard’s possession that relates in any way, shape, or form to failures and other adverse events for all of Bard’s IVC filters. As the Court understands it, an “adverse event” occurs when the filter fails to perform as intended, for example, if a portion of the filter breaks off while implanted in the patient or if the filter “migrates” along the patient’s vena cava, regardless of whether it causes any particular injury to a patient. The parties agree Bard has produced the following categories of information pertinent to adverse events: (1) Trackwise data; (2) fracture reports; and (3) Quality Management Board Reviews (“QMBRs”). As the Court understands based on the parties’ explanations, Trackwise is a software program Bard uses to track the various failures and adverse events attributed to its IVC filters when Bard receives patient complaints, which it maintains in Trackwise complaint files. Bard contends the entire system itself is not suitable for production and inspection without imposing an extraordinary burden on Bard, but it has produced a spreadsheet of data exported from Trackwise. Bard argues the export is sufficient, but plaintiff characterizes the export as containing some, but not all, of the data in Bard’s complaint files. More specifically, plaintiff suggests the exported spreadsheet excludes some of the supporting documentation—e.g., communications from patients and physicians or relevant medical records—included in Bard’s complaint files.2 However, plaintiff is generally satisfied with the Trackwise data because Bard’s reports, including the fracture reports and (to a lesser extent) QMBRs, that have already been produced contain sufficient adverse event data. There remain some narrow disputes concerning Bard’s reports.

2 Plaintiff’s counsel suggests the supporting documents are discoverable because Bard may have imprecisely translated the information contained in the supporting documents into Trackwise. At this point, plaintiff’s concern is speculative at best. Bard’s counsel explains that producing all the supporting documents from every complaint file would be excessive, particularly because not all complaint files are germane to plaintiff’s case, but if plaintiff were able to articulate a need for a particular document or documents, counsel could meet-and-confer about having them produced. The Court directs the parties to do exactly that. If plaintiff can articulate a reason why specific documents stored in Trackwise but not produced in this matter are relevant to this case, as opposed to merely relevant to Bard IVC products liability litigation in general, discovery of those documents might be Fracture reports and QMBRs are two different kinds of reports maintained by Bard in the ordinary course of business. They differ in matters of form and the level of detail provided, but both include information about adverse events related to Bard’s various IVC filters. Plaintiff is particularly concerned with fracture reports here, which her counsel characterizes as more detailed in terms of adverse event data, and therefore more useful on the merits. Plaintiff first notes the production of Bard’s fracture reports ended years ago, and plaintiff suspects that more recent fracture reports, or analogous documents that serve the same function as fracture reports, have not been produced. Second, plaintiff seeks ancillary internal documents that generally concern Bard’s internal analysis and assessment of the adverse events, rather than the raw data about the adverse events themselves. As to the first issue, Bard’s counsel represents that Bard stopped preparing fracture reports in approximately March 2017, which explains why Bard’s document productions do not include reports post-dating that time. Plaintiff’s counsel rejoins by pointing out that federal regulatory compliance obligates Bard to track failure rates in its devices, and he questions whether Bard simply stopped tracking fracture and failure rates, and/or whether the discontinued fracture reports have been replaced by another compilation and reporting process not yet produced in discovery. In a show of admirable candor, Bard’s counsel acknowledged during oral argument that he was uncertain whether the reporting requirement formerly satisfied by the fracture reports is currently being satisfied with another reporting procedure. He suggested Bard may simply be tracking fracture rates as part of its internal complaint files, the relevant portions of which have been exported from Trackwise and provided to plaintiff. However, after conferring with his client, Bard’s counsel was able to confirm that Bard also produces and maintains Bard Quarterly Monthly Reports (“BQMRs”), which include information about IVC filters and associated adverse events. Bard’s counsel represents the BQMRs are a kind of antecedent to the more comprehensive QMBR, and he suggests the relevant information from the BQMR is contained in the QMBR. He also suggests it would be cost- prohibitive to produce BQMRs in discovery because they cover more than Bard’s IVC filters, and they must be reviewed to ensure non-discoverable patient personally identifying information is not inadvertently produced. Plaintiff’s counsel suggests the BQMR contains more granular detail about adverse IVC filter events than the QMBR, and he suggests it is a better means of discovering adverse event data than the unwieldy Trackwise exports. Plaintiff argues the BQMRs are not distinguishable from fracture reports and QMBRs, and they should therefore be produced. The Court will follow the line of reasoning employed by Judge Campbell, who, in managing cases on remand, required Bard to supplement disclosures of “adverse event data,” even though it would technically constitute “common” discovery. See Kathleen A. Shea v. C.R. Bard, Inc. et al., 2:16-cv-03088-DGC, Doc. No. 16 at 3 (D. Ariz. August 7, 2020); see also Lalond v. C.R. Bard, Inc., 1:19cv01516-DAD-BAM, 2020 U.S. Dist. LEXIS 22846, at *6 (E.D. Cal. Feb. 7, 2020) (imposing identical post-remand adverse event data supplementation and noting two other District Court cases adopting the same rationale). Adverse event data is germane to the central issues of liability in this case, and, when collected by Bard in a report or other repository (whether as a matter of regulatory compliance or otherwise) is relevant and not unduly burdensome to produce in discovery. Bard’s BQMR’s fall within the scope of adverse event data reasonably required to be supplemented here after the MDL, and Bard is ordered to produce the BQMRs to plaintiff in this action. The ancillary, internal documents related to Bard’s analyses of its device failures are another matter. Plaintiff suggests Bard had a practice of tracking the failure rates of its devices and periodically evaluating those rates against predetermined guidelines to determine whether to recall any of the retrievable IVC filters. Plaintiff now seeks to discover all the documentary evidence of Bard’s internal deliberations about whether any of its filters should be recalled. Plaintiff describes these documents as probative of Bard’s ongoing duty to warn of the potential dangers associated with a fractured recovery filter and whether such duty was breached in this case. The Court perceives an outcome determinative flaw in this line of reasoning. The parties hotly dispute whether California law imposed a duty to warn of dangers that became known to Bard only after the IVC filter was placed into plaintiff. If such an ongoing duty exists, plaintiff contends Bard should have warned plaintiff (or more likely her physicians) of dangers inherent in the Recovery once it became aware that a large number of recovery filters were fracturing and injuring patients. Bard contends no such ongoing duty exists as a matter of California law, and it discharged any duty to warn when it told plaintiffs doctors about all dangers known to Bard at the time the device was originally sold and implanted. The Court does not need to decide the issue because, assuming California law imposes a post-sale duty to warn, that duty indisputably sounds in negligence, and California negligence law is axiomatically based on an objective “reasonableness” standard. See Regents of Univ. of Cal. v. Superior Ct., 29 Cal. App. 5th 890, 904 (2018). Indeed, the post-sale duty to warn at issue here is based on section 10 of the Restatement (Third) of Torts: Products Liability, which looks at whether a “reasonable” seller would provide a post-sale warning. See also Restatement (Third) of Torts: Products Liability, § 10 cmt. b (“The standard governing the liability of the seller is objective: whether a reasonable person in the seller’s position would provide a warning. This is the standard traditionally applied in determining negligence.”). Thus, assuming an ongoing duty to warn applied, plaintiff can establish whether Bard should have issued a warning here based solely on the raw adverse event data. What Bard did with that data as a subjective matter is generally irrelevant to the ultimate calculus of what an objectively reasonable seller in Bard’s position would or should have done.3 3 Bard’s response to the various adverse events is much more relevant to the issue of breach. However, as the Court understands it, the question of duty is the focus of this case, at least as to the post-sale duty to warn, because it is essentially undisputed that Bard, believing it had no such duty, issued no warnings other than those provided at the time of sale. Thus, even to the extent the discovery at issue here is relevant to proving a potential breach of the post-sale duty to warn, it is cumulative and thus disproportionate to the needs The value to this case of Bard’s internal deliberations is therefore de minimis at best, production of the documents is not proportional, and Bard’s internal communications about the adverse event data in its possession fall beyond the scope of discovery. Accord Shea, 2:16-cv-03088-DGC, Doc. No. 16 at 3 (noting the “voluminous and comprehensive fact and expert discovery” conduct during the MDL and ruling “the Court will not reopen general fact or expert discovery”). Plaintiff’s Motion to Compel further responses to RFP No. 8 is accordingly GRANTED IN PART AND DENIED IN PART. Bard must produce the BQMRs that it has not yet produced, but it is not obligated to collect and produce ancillary internal documents related to IVC filter adverse event data. (C) Request for Production No. 11 Plaintiff’s RFP No. 11 states: Produce all DOCUMENTS AND ESI EVIDENCING or RELATING to any communications whether by e-mail, telephone, facsimile, or verbal between YOU and Plaintiff's health care providers Brian Moore, M.D., Jong Hwan Yun, M.D., Sharp Grossmont Hospital in La Mesa, California, or Kaiser Permanente Zion Medical Center in San Diego, California regarding the safety and/or efficacy of the Recovery Filter, G2 Filter, Eclipse Filter, Meridian Filter, and/or Denali Filter.

This RFP has provoked a dispute between the parties about how to search for and produce responsive ESI. The parties agree that Dr. Moore implanted the Recovery filter in plaintiff, and that Dr. Yun is her primary care physician. The two medical centers identified in the RFP are where plaintiff has received medical care related to the Recovery filter, including but not limited to the original implantation, and where she is currently receiving treatment. Plaintiff wants Bard to turn over every communication in its possession that is (a) pertinent to all of the retrievable IVC filters identified in the RFP; and (b) between Bard and either of the doctors or medical facilities listed in the RFP, whether or not such communications pertain to plaintiff. There are multiple sources of ESI at issue here. Plaintiff’s first dispute concerns a large volume of ESI that was collected as part of the MDL proceedings. The documents are kept in a database, and they can be searched with keywords to locate responsive ESI on a case-by-case basis. Bard has agreed to search these collected-but-not-produced documents using a series of keywords, which includes the plaintiff, her treating physicians, and the sales representatives responsible for covering the general geographic area during the period when the device was implanted in plaintiff—all of which Bard’s counsel represents are run using certain “anchor terms.” Plaintiff disagrees with the search terms Bard has used. The second prong of the dispute is that plaintiff also wants Bard to collect and review ESI from other ESI custodians. Specifically, two Bard employees who held the role “Vice President of Quality.” This new ESI would fall beyond the scope of the ESI that was collected during the MDL proceedings. Bard objects to engaging in a second round of ESI collection and estimates the costs of doing so could exceed $100,000, which it considers disproportionate to the needs of the case. Plaintiff suggests Bard’s estimate of the economic burden here is purely speculative and not founded on any actual evidence. Moreover, plaintiff has had no opportunity to challenge Bard’s assessment here, which may well include sources of financial burden (such as reviewing for privilege claims) not properly included as a matter of law. As to the first issue, the Court does not relish the task of refereeing disputes about keyword search terms because the parties, not the Court, have a much better idea of the most effective way to gather responsive documents. The thrust of plaintiff’s complaint here is that Bard, particularly through its use of “anchor terms” in many search strings, is using search terms that are too narrow. Plaintiff fears Bard’s keyword searches will exclude information relevant to this case. Bard argues plaintiff’s keywords are too broad and would return a glut of documents not relevant to this case. However, because the ESI at issue has been collected and stored in a readily searchable format, Bard generally admits the burden of running additional search strings is minimal. Bard is accordingly ordered to search the collected-but-not-produced ESI using plaintiff's keywords, but with the following caveat: plaintiff must restrict her proposed keywords to collect information pertinent to the Recovery filter, not the subsequent generations of filters. As to whether Bard should engage in a second round of ESI collection to broaden scope of discovery, plaintiff has not shown the Court how the information is relevant to plaintiff's injuries. Given the low relevance of this information, which is cumulative of common discovery already conducted, the Court finds that conducting fact discovery collect more ESI that would sweep in additional custodians is accordingly not proportional to the needs of this case. Plaintiff's Motion to Compel further responses to RFP No. 11 is accordingly GRANTED IN PART AND DENIED IN PART #4 IV. Conclusion Plaintiff's motion to compel further responses to RFP No. 7 is DENIED. □□□□□□□□□□□ }motion to compel further responses to RFP Nos. 8 and 11 is GRANTED IN PART AND DENIED IN PART as explained in this Order. Dated: June 26, 2023 Wi (Be “a Hori. Karen 8S. Crawford United States Magistrate Judge || The Court urges the parties to be reasonable when negotiating search terms as /ordered. If plaintiff produces a voluminous list of search strings that, by design or by accident, unreasonably burden Bard, Bard may yet object. But the Court expects Bard to grant plaintiff some latitude, as the burden of searching the pre-collected ESI is minor, and Court has spared Bard the burden of collecting ESI beyond that which has already been collected. If the parties reach a bonafide dispute about search terms, they must meet and confer before contacting the Court.