No. 01-1897




Appeal from the United States District Court
for the Eastern District of Pennsylvania



Anthony J. Majestro
Marvin W. Masters
Masters & Taylor, L.C.
181 Summers Street
Charleston, WV 25301
(304) 342-3106
Attorney for Appellant Lloyd
Brian Wolfman
Alan B. Morrison
Public Citizen Litigation Group
1600 20th Street, N.W.
Washington, DC 20009
(202) 588-1000
Attorney for Appellants
Masters and Majestro

June 19, 2001


            Before addressing appellees' specific responses, we reply to two overarching points that shed light on what is and is not at issue in this appeal.

            1. In an effort to show that Ms. Lloyd's suit will cut into class members' recoveries, AcroMed repeatedly asserts that it is indemnified by the settlement fund for the cost of any bone-screw related litigation brought outside the class action. E.g., AcroMed Br. ("AM") 5-6, 22, 37. That assertion is inaccurate. AcroMed sought fees and costs from Ms. Lloyd below, which the court treated as a claim for indemnification and denied. JA 243-44. However, even if it were accurate, it would be irrelevant to whether Ms. Lloyd's West Virginia suit may proceed. No class action judgment is binding without proper notice and an opportunity to be heard, adequate representation, and, in most instances, a right to opt out. Phillips Petroleum v. Shutts, 472 U.S. 797, 812 (1985). Those basic requirements cannot be circumvented by the defendant's insertion of indemnification language into the very settlement under collateral attack.

            In any event, AcroMed misconstrues the settlement agreement. AcroMed is "indemnified by the AcroMed Settlement Fund, by appropriate petition to the Court" for costs involving "Settled Claims of Settlement Class Members ... that are filed in the future despite this Agreement." AA 76. However, if the Fanning judgment is not binding on Ms. Lloyd, then, by definition, she would not be asserting a "Settled Claim" and would not be a member of the "Settlement Class." Put differently, the premise of Ms. Lloyd's arguments on the merits is that she cannot be considered a "Settlement Class Member," and, if that is correct, the settlement's indemnification provision does not apply. That reality explains why AcroMed is fighting this appeal so tenaciously. If the indemnification provision were the "get-out-of-jail-free card" that AcroMed claims it is, Ms. Lloyd's suit would pose no threat to the company.

            AcroMed's reliance on the indemnification provision is tied to its repeated assertion that it is a "limited fund" that paid all it had into the settlement fund, and therefore cannot pay defense costs or a judgment in Ms. Lloyd's suit. That is not correct. Shortly after settlement approval, AcroMed was acquired by a solvent French conglomerate, JA 182-84, and thus AcroMed need not raid the settlement fund to defend Ms. Lloyd's case. Furthermore, even before the settlement became final, AcroMed settled the appeals of 54 objectors for "a substantial payment" from outside the supposed limited fund. S. Elizabeth Gibson, Case Studies of Mass Tort Limited Fund Class Action Settlements & Bankruptcy Reorganizations 153 (Federal Judicial Center 2000). As one commentator asked, "If the company had devoted all that it had to the settlement, where did the additional funds come from?" Id. at 158.(1)

            Moreover, it is undisputed that AcroMed's principal business is still the sale of pedicle screws and that the settlement did not apply to lawsuits brought by people implanted with those screws after December 31, 1996. JA 75. Its CEO testified that AcroMed would not alter the product's design, manufacturing process, or product labeling, or its sales practices after the settlement. See 4/24/97 Tr. 224-25 (Addendum to this brief, 1a). Thus, as the court below noted, JA 88-89, AcroMed must pay any defense and judgment costs associated with post-December 31, 1996 bone-screw implants with non-settlement money. In sum, AcroMed's frequent "limited fund" refrain is plainly mistaken.(2)

            2. Much of appellees' briefs considers whether the AcroMed settlement is binding on Ms. Lloyd. See, e.g., AM 23-35. But that is the secondary question presented here, which will be reached only if the principal question--whether the district court had the power to determine the binding effect of the Fanning judgment--is resolved against Ms. Lloyd. If this Court rules otherwise, as we believe it should, AcroMed's res judicata arguments will remain open in the West Virginia trial court, which has both the jurisdiction and the obligation to consider them. If those arguments are rejected there, AcroMed can appeal any final judgment to the West Virginia appellate courts and to the Supreme Court.


I. Because The District Court Lacked Personal Jurisdiction Over Ms. Lloyd, AcroMed's Res Judicata Defense May Only Be Asserted In Ms. Lloyd's West Virginia Action.

A. Personal Jurisdiction Was Lacking Because Class Members Were Not Permitted To Opt Out.

            Appellees concede that, under this Court's decision in In re Real Estate, a class action seeking a "money judgment" must provide an opt-out right before it can obtain personal jurisdiction over the absentees. AM 37; PLC 7. They argue only that this case eludes In re Real Estate because it does not involve a "money judgment," but rather a "limited fund" certified under Rule 23(b)(1)(B). See In re Real Estate Title & Settlement Services Antitrust Litig., 869 F.2d 760, 768 n.8 (3d Cir. 1989). That argument is wrong, as revealed by Ortiz v. Fibreboard Corp., 527 U.S. 815 (1999), and its predecessor, Shutts, 472 U.S. 797.

            Shutts held that, at a minimum, a court does not obtain personal jurisdiction over absentees in a class action "wholly or predominantly for money judgments" unless they are provided the right to opt out. Because Ortiz reiterated that Shutts' holding applied only to "money judgment" claims, Ortiz, 527 U.S. at 848 n.24, AcroMed asserts that there is no right to opt out under Ortiz where, as here, the defendant claims to be a "limited fund." AM 41 n.8. If that argument were correct, however, Ortiz would be at war with itself. Ortiz noted that Shutts had "raised the flag on this issue of due process more than a decade ago," 527 U.S. at 847 (emphasis added), and then discussed the opt-out issue at some length. AcroMed would have this Court believe that Ortiz was musing over something having nothing to do with the circumstances in that very case. However, the only sensible interpretation of Ortiz is that it "reinforce[d] the opt-out right recognized in Shutts,"AM 41 n.8, because it believed that the right applied in the purported limited-fund settlement at issue there. Since the "limited fund" rejected in Ortiz is legally indistinguishable from the purported limited fund here, the district court did not obtain jurisdiction over Ms. Lloyd because it never gave her a right to opt out.(4)

            After Shutts and Ortiz, then, the only question is whether this case is one wholly or predominately for money judgments. It is. The sole relief sought in individual bone-screw cases is money damages. The sole relief sought by Ms. Lloyd in her West Virginia action is money damages. JA 156-57. The sole relief provided to class members in Fanning is money damages. And, most fundamentally, the principal claims purportedly released by the Fanning judgment are for money damages. JA 122.

            In arguing that Fanning does not involve money damages, AcroMed relies on cases involving "traditional" limited funds, see Ortiz, 527 U.S. at 834, which do not involve in personam personal-injury claims, but identical claims against a pre-existing res. Indeed, every case cited by AcroMed fits the traditional mold. E.g., Smith v. Swormstedt, 57 U.S. (16 How.) 288 (1854) (suit to recover common tangible property); Mullane v. Central Hanover Bank & Trust, 339 U.S. 306 (1950) (rights to pre-existing trust fund); see Ortiz, 527 U.S. at 834 (collecting authorities).

            This Court's decision in Walsh v. Great Atlantic & Pacific Tea Co., 726 F.2d 956 (3d Cir. 1983), relied on heavily by AcroMed, is in the same vein. There, the plaintiffs sought a declaration under ERISA that their employer's unilateral amendments to a retirement plan provided the employer an illegal benefit. A class was certified under Rules 23(b)(1) and(2), and the settlement reallocated a preexisting fund by prohibiting certain monies from reverting to the employer. Id. at 958-62. That situation, where all class members sought and received identical treatment, is nothing like Fanning, where the class members possess personal claims, the validity and value of which depend on the particular facts of their individual circumstances. In short, Fanning involved claims for "money damages" and, as in Ortiz, did not involve a traditional "limited fund" that could proceed on a non-opt-out basis under the Constitution and Rule 23.

            AcroMed makes three other arguments regarding In re Real Estate. First, AcroMed asserts that Ms. Lloyd "consented" to the district court's jurisdiction because she did not seek to opt out at the 1997 fairness hearing. AM 43. AcroMed argues that In re Real Estate is distinguishable because the Arizona school boards there had notice and moved to opt out at the fairness hearing, evidencing their lack of consent to jurisdiction.

            That argument stands In re Real Estate on its head. AcroMed admits that if an out-of-state class member receives notice, she may appear solely to contest the lack of an opt-out right without consenting to personal jurisdiction. See In re Real Estate, 869 F.2d at 770-71. Yet AcroMed maintains that where, as here, the class member is deprived of two of the essential due process elements in Shutts, 472 U.S. at 812--notice and an opportunity to opt out--personal jurisdiction is present because the class member did not appear to contest it. That argument is absurd on its face, because a class member must have notice before she can consent and because the rationale for due process limits on a court's jurisdiction is that an absentee is not required to appear in a foreign court in certain circumstances. See International Shoe v. Washington, 326 U.S. 310 (1945).

            Second, AcroMed claims that the school boards' attack in In re Real Estate was limited to an assertion that the class representatives had "inadequately represented them by ignoring state-by-state variances in the strength of the state action defense," while Ms. Lloyd contemplates "a wholesale collateral attack." AM 44-45. Actually, the school boards' attack was broader than Ms. Lloyd's: They claimed that the MDL 633 judgment was non-binding for all Arizonans because the named plaintiffs had "sold them out" by failing to obtain monetary relief. In re Real Estate, 869 F.2d at 765 n.2; see JA 283-88. In any event, In re Real Estate does not remotely suggest that its personal jurisdiction holding hinges on the kind of defect alleged by the absentee. Rather, the crux of In re Real Estate is that a ruling on an absentee's challenge to a prior judgment must be made by the court in which the absentee brought suit, if no opt-out right is provided. In response, AcroMed relies on Carlough v. Amchem Products, 10 F.3d 189, 204 (3d Cir. 1993), see AM 45, but that case actually supports Ms. Lloyd. Carlough held that an out-of-state class action filed for the purpose of disrupting the Georgine asbestos settlement could be enjoined, but only after the class members were given notice and an opportunity to opt out, id. at 200, 203-04, precisely the essential due process rights denied Ms. Lloyd.

            Finally, AcroMed asserts that, because Fanning was part of a multi-district litigation, the district court had jurisdiction over Ms. Lloyd, although she did not learn of MDL 1014 until two years after the fairness hearing. AcroMed acknowledges that its position conflicts with In re Real Estate, AM 46 n.9, but nonetheless claims that if Ms. Lloyd had been a part of a West Virginia class action, which had been transferred to MDL 1014, she would have been subject to the jurisdiction of the district court. AM 46. Although that did not occur here, AcroMed would be incorrect even if it had.

            Any application of the MDL statute, like any statute, must comport with due process, as demonstrated by In re Real Estate. The Arizona school boards were, in fact, part of a federal class action in Arizona that was transferred to MDL 633 in Philadelphia, 869 F.2d at 763, the very hypothetical that AcroMed claims illustrates the folly of Ms. Lloyd's position. Despite these facts, this Court never suggested that the MDL statute overrode the school boards' due process right to sue in their home forum.

            AcroMed raises two other hypotheticals that are even further afield. AcroMed says that if it had succeeded in removing Ms. Lloyd's West Virginia suit to federal court, or if it had sued Ms. Lloyd in West Virginia federal court seeking a declaration that her suit violated the Fanning judgment, those cases would have been transferred to MDL 1014. Because under those circumstances, AcroMed asserts, the district court would have had personal jurisdiction over Ms. Lloyd, it is present here as well. AM 47-48.

            Those hypotheticals are not this case, and Ms. Lloyd's suit has not made its way to MDL 1014. In any event, the premise of AcroMed's argument is wrong. A tort defendant cannot end-run personal jurisdiction rules simply by suing a potential or actual plaintiff in a separate federal-court action. Cf. Skelly Oil v. Phillips Petroleum, 339 U.S. 667, 671-74 (1950) (declaratory judgment action cannot create federal jurisdiction where none exists in affirmative suit on which declaratory judgment action is premised). Thus, even if AcroMed had sued Ms. Lloyd in West Virginia, or even if her suit had been removable, those procedural machinations would not have magically created contacts with, or consent to be sued in, the MDL court sitting in Philadelphia. Those are the teachings of In re Real Estate and Shutts, which would be meaningless if they could be evaded whenever a defendant brought a declaratory judgment action seeking to obtain the forum of its choice. Under AcroMed's theory, the Arizona school boards would have been forced to litigate their due process claims in Philadelphia if the defendants, rather than seeking an injunction in MDL 633, had sued for declaratory relief in Arizona federal court, had the case transferred to MDL 633, and then sought an injunction. To state the proposition is to refute it.(5)

* * *

            To sum up: In re Real Estate prohibits the out-of-state assertion of jurisdiction that AcroMed urges here; Shutts requires receipt of notice and an opt-out right before an absent class member can be bound with respect to any claim that is wholly or predominantly for money damages, such as those foreclosed by Fanning and pled by Ms. Lloyd in West Virginia; and Ortiz closes the loop by applying Shutts to purported limited fund cases. Even if Ortiz leaves some breathing room for mass-tort class actions, it leaves no doubt that the constitutional principles enunciated in Shutts and In re Real Estate apply here. Therefore, the decision below should be reversed.(6)

B. The District Court Did Not Have Personal Jurisdiction Over Ms. Lloyd Because The Notice Was Constitutionally Defective.

            Appellees make several arguments to rationalize their feeble attempt at notifying people like Ms. Lloyd who had not already sued AcroMed, but do not contest two basic facts: For the tens of thousands of purported class members who had not already sued AcroMed, no effort was made at first-class notice, and the only attempt to inform them of the settlement consisted of a few short ads in national publications.

            Before confronting appellees' specific notice arguments, one recurring theme requires a reply. Appellees claim that prior publicity regarding bone screws alerted people to the potential for litigation, and that, therefore, they were justified in providing scant notice to class members other than those who had already sued AcroMed. AM 25; PLC 21-22. As discussed below, that point is wrong on its own terms, not only because hundreds of class members were not notified on a timely basis, see In re Orthopedic Bone Screw Prods. Liab. Litig. (Sambolin), 246 F.3d 315, 320 (3d Cir. 2001), but also because it is unknown how many injured bone-screw patients still have not been notified or how many learned of the registration deadline after May 1997 and took no action because the notice said it was too late. Cf. JA 129 (order instructing Claims Administrator not to accept registrations proffered after April 6, 1999).

            However, even if appellees are correct that the mailed notice reached the great majority of people with claims against AcroMed, that would hardly support the result that they seek. To the contrary, if that were accurate, the obvious solution would be to bind only those class members who in fact were notified by first-class mail, just as occurred in Shutts, 472 U.S. at 801. Cf. Fed. R. Civ. P. 23(c)(2). Such a rule would comport with due process, would encourage appellees to find and notify class members through doctors and others, and would constitute a very small threat to AcroMed's assets if, as appellees claim, "[i]t is reasonable to assume that those individuals with compensable claims against AcroMed are aware of this litigation and have [already] filed complaints against AcroMed." See Settling Parties' Proposed Findings of Fact, at 153 (filed July 3, 1997), Addendum 3a. On the other hand, the rule urged by appellees would bind bone-screw patients through the fiction of constructive notice, even where they had no inkling of the settlement or, as in Ms. Lloyd's case, were not injured at the time notice was provided. That result is intolerable.

            Regarding appellees' specific arguments, we turn first to "future" claimants, who, like Ms. Lloyd, were implanted with bone screws by December 31, 1996, but were not injured when the notice went out a month later in January 1997. That issue presents the most egregious ramification of the inadequate notice and the narrowest basis on which to rule in this case. Despite the PLC's protestations, PLC 10 n.1, Fanning is indisputably a "futures" class action because it "specifically includ[es] claims for alleged injuries and damages not yet known or manifest." JA 74 (class definition). The PLC relies on the district court's statement that, because most bone-screw injuries arise within four to six months of implantation, the class members would have suffered injury by the time they had to file their claims. PLC 10 n.1. That argument is both irrelevant and factually incorrect. It is irrelevant because the key question is not whether class members would be required to file claims before they were injured (although that would be problematic), but whether they, or their friends, family, and doctors, would be notified of the settlement before their injuries occurred. Even assuming a latency period of only four months, the notice program predated the injury for class members who had surgery in the last month or two of the class period, and thus provided no notice at all.

            The district court cited an AcroMed expert, who declared that, although most bone-screw injuries occur within four to six months, some take eight months to arise, and a small number may occur up to a year after implantation. Addendum 6a-7a. AcroMed acknowledges these facts, AM 39-40 n.7, and the district court implicitly recognized them, because it justified inclusion of future claimants on the ground that they would have "a minimum" of a year after their surgeries to file claims. JA 82. But that is demonstrably wrong. The settlement agreement included a May 1997 registration deadline, which, as noted in Sambolin, the district court enforced almost without exception. 246 F.3d at 320. Thus, the supposed one-year filing window from injury to deadline, was, in fact, as short as approximately four months, and further underscores the inadequacy of the notice for class members, like Ms. Lloyd, whose implantation occurred shortly before the notice program. See also Addendum 8a (expert opinion premised on same inaccuracy).

            Quite apart from the "futures" problem, the notice program was constitutionally defective principally because a few short ads in national publications are insufficient to provide constructive notice. See Lloyd Opening Br. 43-47. Appellees cite a few cases where notices were published in newspapers and class members were sent full notices by first-class mail. AM 24; PLC 20. In those cases, the settling parties had lists of all, or almost all, of the class members, each of whom was sent a notice by first-class mail. See, e.g., Zimmer Paper Prods. v. Berger & Montague, 758 F.2d 86 (3d Cir. 1985). That is what occurs, for instance, in ERISA cases, where the employer or fiduciary maintains a list of names and addresses of plan participants. See Walsh, 726 F.2d at 962-63. In those cases, publication notice, a poor substitute for individual notice, Mullane, 339 U.S. at 315, is used only to find the few people that the list may have missed.(7)

            Fanning presents an entirely different situation. The settlement purports to bind between 100,000 and 117,000 implantees, AA 85, but first-class notice was provided only to the few thousand people who had already sued AcroMed or who were already known to the PLC. Obviously, more than scant publication notice was constitutionally required to reach the tens of thousands of other AcroMed bone-screw implantees. In this regard, Sambolin's dicta could not be more on point: Although Sambolin does not say precisely which additional tools would have brought the notice program up to constitutional standards, it does say that the notice here did not meet the "best notice practicable" standard, 246 F.3d at 327 n.11, the hallmark of both the Constitution and Rule 23. See Mullane, 339 U.S. at 316; Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 175 (1974). Thus, although appellees are correct that Sambolin's due process observations were dicta, the content of that dicta is not in doubt.(8)

            Since appellees rely principally on Walsh, it bears mentioning that Walsh is inapposite on another ground. As noted earlier, Walsh involved a true mandatory class because the sole issue was whether the company's unilateral changes to the ERISA retirement plan should be declared unlawful; any relief to the class would simply take money claimed by the defendant and return it to the fund for ultimate distribution based on the class members' pre-existing interests in that fund. 726 F.2d at 959-60. Thus, although notice was far more extensive than it was here (the affected class members were notified by first-class mail), its sole purpose was to allow class members to object to the settlement (not claim benefits), and those who did not receive notice would still be entitled to share in the retirement fund. Here, by contrast, no notice means no compensation in perpetuity.

            Appellees argue that the publication notice was deliberately meager because they knew that bone screws had received media attention in the mid-1990's and, again, when the settlement was reached. The Claims Administrator made the same argument in favor of affirmance in Sambolin, see Welsh Brief in No. 99-2054, at 20-21, and this Court apparently was not persuaded.(9)

            For good reason: There is no evidence that the paucity of formal notice reflected appellees' judgment that prior newspaper articles or lawyer advertisements were an effective means of notifying the class. See AM 26. Rather, they endorsed a minimal notice program because they believed (wrongly) that virtually everyone with a claim had already sued AcroMed. See Addendum 3a.

            Moreover, the articles and ads on which appellees rely did not bring the notice program up to constitutional standards. Most of them pre-date the settlement by more than two years and deal only with the general issue of bone-screw injuries. See AA 1-42. The articles that concern the settlement are from early to mid-December 1996, AA 59-74, shortly after the parties reached an informal understanding, but before the settlement's terms were public. Indeed, the settlement agreement was not signed until January 8, 1997, JA 16, about a month after publication of the articles cited by AcroMed, and the court did not approve the notice until January 16. JA 24. Therefore, the articles simply announce the parties' understanding and the aggregate settlement amount; they do not, because they could not, state the settlement's specific terms, how a class member could obtain more information, that typical opt-out rights were abrogated, or that there was a May 1997 registration deadline. Thus, these articles did not make an otherwise defective notice constitutional.(10)

            Appellees' other major point is that additional notice would have been too costly. Cost is a proper consideration in determining the kind and amount of notice to a class. In one respect, however, the PLC's focus on cost is irrelevant because the best and least expensive way to find potential claimants would have been through doctors and hospitals, i.e. AcroMed's customers. Had AcroMed notified Cabell Huntington Hospital and/or Ms. Lloyd's doctors, asking them to notify their bone-screw patients, Ms. Lloyd likely would have learned of the settlement.

            The PLC admits that notice of this kind would have cost only about two dollars per mailing, PLC 19 n.8, negating any serious cost concern. AcroMed's only response, quoting Judge Bechtle's July 10, 2000 order, is that notifying doctors would have been impracticable because the "structure of the industry for the distribution of bone screw products" would have required a "full blown investigation" of the bone-screw distribution process. AM 32. The company does not explain what that means or deny that it has a list of hospitals and doctors to whom it sold its expensive products.

            Indeed, in Sambolin, AcroMed acknowledged that it could have contacted the hospitals where surgeries took place and the doctors who performed them. AcroMed Brief in No. 99-2054, at 30 ("notice could have been sent to hospitals and surgeons for distribution to their patients"). It argued only that health care providers would have been uncooperative because they were also defendants in bone-screw suits. Id. However, Sambolin emphatically rejected that argument as "unconvincing." 246 F.3d at 327 n.11. Actually, according to AcroMed's CEO, only a "small percentage of bone screw plaintiffs have advanced claims ... against treating physicians and hospitals." AA 93. And of those that were sued, AcroMed's customers were more likely than most doctors and hospitals to be cooperative because Fanning purported to immunize them from the kinds of bone-screw lawsuits that they had faced in the past. AA 93-98. In any event, even if the hospitals and doctors refused to provide notice, they could have (and should have) been required to do so, just as brokers are routinely required to provide notice to their clients in securities class actions. See Manual on Complex Litigation Third § 30.211, p. 226 (1995); 2 Newberg on Class Actions § 8.08, p. 8-29 (3d ed. 1992); accord Sambolin, 246 F.3d at 327 n.11; see also PLC 18 n.7 (noting that brokers can be required to notify class members who hold stock in "street name" if brokers' costs are covered).

            Furthermore, AcroMed's non-cooperation claim is belied by experience. Other medical-device settlements were very successful in locating class members, who otherwise never would have learned of the settlement, by using doctors and hospitals to notify their patients. Lloyd Opening Br. 39-42. Those results do not prove that the same success would have occurred here; however, the PLC is comprised of lawyers at the forefront of mass-tort litigation nationwide, and it has never disputed those results nor provided contrary data indicating that notice through health care professionals is ineffective.

            Even apart from the failure to notify through doctors and hospitals, appellees' "cost" argument fails. The notice program cost about $325,000. PLC 19 n.8. In a classic red herring, the PLC contrasts that amount with an assertion that spending 20% of the fund would have cut too far into the claimants' recoveries. Id. However, the notice here did not cost 20%, 10%, 2%, or even 1% of the fund, but a piddling .327% (less than one-third of one percent) of the fund. If only 1% of the fund had been spent, notices could have been published in regional and local papers, see AA 135 (notice in San Juan general circulation paper cost $1223), and selected electronic media, not to mention the "free media," such as public service announcements, that were ignored in this case. However, spending 6%, or $6 million--which would have allowed extensive electronic and newspaper notice, cf. Ahearn v. Fibreboard Corp., 1995 U.S. Dist. Lexis 11532, *294-*304 (E.D. Tex. July 27, 1995)--could have easily been justified on cost grounds, as that would have been about one year's interest on the fund.(11)

            Finally, in an effort to explain away the overwhelming evidence of the defective notice, AcroMed suggests that the hundreds of "late" registrations were the result of neglect, not poor notice. AM 28-29. Not so. Registration materials are kept by the Claims Administrator and are not in the record. However, in December 1999, the Claims Administrator denied nearly 600 claims as untimely, the "greatest number" of which "asserted that they had not heard of the settlement or their ability to participate," because they ran "squarely into the Court's holding that notice was adequate as initially conducted." Addendum 10a. The district court's decision affirming that denial also analyzed the issue as one of notice, JA 232-35, which would have been irrelevant, if neglect, not lack of actual notice, was the problem.

            In fact, the Claims Administrator received hundreds of late registrations from late 1997 through early 1999. See Welsh Brief in No. 99-2054, at 16-17. In September 2000, at oral argument in Sambolin, he noted that he continued to receive registrations "over the transom" despite the district court's April 1999 order directing him not to accept further registrations. It defies logic to suggest, as does AcroMed, that large numbers of class members learned of the settlement and its registration requirement during the notice period in early 1997, but waited months or even years to register. The only rational conclusion is that thousands of class members were not notified of the settlement on a timely basis.(12)

C. Ms. Lloyd Did Not Consent To The District Court's Jurisdiction Through Her November 1999 Appearance.

            Appellees argue that Ms. Lloyd's appearance before the district court in November 1999 constituted consent to jurisdiction in the Fanning settlement approved in October 1997. Appellees rely on statements from briefs filed below that outlined our inadequate representation arguments. We will not repeat our opening brief, except to reiterate that the only ruling that Ms. Lloyd sought below was one that would allow her to pursue her West Virginia action. Ms. Lloyd never sought a ruling that her due process rights were violated by the Fanning judgment, which is an issue that she was (and remains) content to take up in West Virginia.

            However, it is important to note the internal contradiction in AcroMed's "consent" defense. AcroMed acknowledges, as it must, that a class action judgment is not binding without notice, or if it is the result of inadequate representation, and thus concedes that Ms. Lloyd has a right to challenge the Fanning judgment in some court. AM 20. That said, it is difficult to understand what AcroMed believes Ms. Lloyd should have done differently. If Ms. Lloyd had not outlined her inadequate representation arguments, AcroMed would still have argued that she could not pursue her West Virginia suit (after all, AcroMed sought to enjoin that suit in October 1999, before Ms. Lloyd's counsel first appeared in MDL 1014, see JA 142). And had Ms. Lloyd said nothing below about why the settlement did not bind her, AcroMed surely would be arguing that she waived any due process challenge to the Fanning judgment in federal court.

            Seen from this perspective, Ms. Lloyd did the most sensible thing. Her counsel appeared before the district court when summoned to do so, seeking only the right to proceed in West Virginia. However, they outlined the arguments Ms. Lloyd would make on the merits of her inadequacy claims, so if AcroMed were successful in blocking her West Virginia action, she could at least make those arguments in federal court. See JA 223.

            In any event, AcroMed's focus on Ms. Lloyd's 1999 filings exposes the fundamental legal flaw in the company's "consent" argument. As noted in our opening brief (at 48-49), the time for judging Ms. Lloyd's consent (or lack thereof) was the Spring 1997 fairness proceeding, at which point she knew nothing about the settlement and so could not have consented to jurisdiction in MDL 1014. Appellees simply ignore In re Real Estate, the precedent directly on point. In a critical passage, wholly disregarded by appellees, this Court explicitly noted that the school boards had also argued a lack of adequate representation in response to the defendants' attempt to enjoin their Arizona state court action. 869 F.2d at 765 & n.2; see JA 283-88. This Court obviously did not believe that raising merits arguments at that time barred consideration of the personal jurisdiction issue, which was, in fact, resolved in the school boards' favor. That ruling applies fully here and demonstrates that Ms. Lloyd did not consent to the district court's jurisdiction.

II. Alternatively, Fanning Does Not Bind Ms. Lloyd.

            Ms. Lloyd's principal contention is that the district court did not have jurisdiction over her and that, therefore, the injunction should be lifted, leaving Ms. Lloyd free to prosecute her West Virginia suit, and AcroMed free to defend that suit on res judicata grounds. If the Court disagrees, however, it must consider whether Fanning is binding on Ms. Lloyd. We rely mainly on our opening brief (at 51-56), and address only one point concerning the purported $100 million "limited fund."

            Something is terribly wrong here: The district court adopted the PLC's expert's $104 million valuation based on "what a willing buyer would pay a willing seller for this company," JA 69 (quoting report of Dr. Harvey Rosen), but the company was then sold for $325 million plus the assumption of the entire settlement liability. AcroMed's brief simply ignores the issue and does not disclose whether the sale was in negotiation prior to settlement approval, and, if so, whether the company made any effort to disclose that critical information to the PLC or the court.

            For its part, the PLC says nothing about whether it ever inquired into the possible sale of AcroMed, or whether it did any investigation other than obtain Dr. Rosen's (obviously flawed) report. It does, however, try to rationalize the huge difference between the court's finding and the actual sale price by stressing that the sale came "five months after the settlement" "with the cloud of litigation removed." PLC Br. 10 n.2 (emphasis in original). But, as the PLC well knows, Dr. Rosen's valuation was supposedly based on the sale value of the company "without the financial constraints of the litigation costs and the uncertainty of litigation outcomes," JA 69, i.e. "with the cloud of litigation removed." Indeed, that is the only type of valuation that made sense: Why would the settling parties want to trumpet the value of the company absent the settlement? Why would the class members care about anything other than AcroMed's post-settlement value? After all, the whole point of the settlement was to put the litigation behind the company, and thus the valuation purported to be based on AcroMed's cash-flow with the settlement in place. See JA 68-70.(13)

            Based on the current record, the only reasonable conclusion is that the PLC's representation regarding the amount of the "limited fund" was inadequate. If the Court reaches that issue, it should hold that Fanning does not bind Ms. Lloyd or, alternatively, remand the issue for further development.


            The decision below should be reversed.

Respectfully submitted,

Brian Wolfman
Alan B. Morrison
Public Citizen Litigation Group
1600 20th Street, N.W.
Washington, D.C. 20009
(202) 588-1000

Attorneys for Appellants Majestro and Masters

Anthony J. Majestro
Marvin W. Masters
Masters & Taylor, L.C.
181 Summers Street
Charleston, WV 25301
(304) 342-3106

Attorneys for Appellant Lloyd

June 19, 2000


            I, Brian Wolfman, counsel of record in this appeal, am a member of the bar of this Court.

Brian Wolfman

June 19, 2001


            Pursuant to Federal Rule of Appellate Procedure 32(a)(7)(C), I state that this brief complies with the type-volume limitations of Rule 32(a)(7)(B), and contains 6993 words according to WordPerfect 7, the word-processing program used to prepare this brief.

Brian Wolfman


            I hereby certify that, on June 19, 2001, I caused to be served by regular U.S. mail two copies of the foregoing brief on each of the following counsel who have entered an appearance in this appeal:

Arnold Levin
Levin, Fishbein, Sedran & Berman
510 Walnut Street, Suite 500
Philadelphia, PA 19106

Richard I. Werder, Jr.
Jones, Day, Reavis & Pogue
North Point, 901 Lakeside Avenue
Cleveland, OH 44114

Brian Wolfman

1. Moreover, there was obviously no "limited fund" with respect to doctors who implanted AcroMed bone screws and hospitals where surgeries took place, and yet the settlement purported to release them, on a non-opt-out basis, from a wide range of claims, JA 63, including Ms. Lloyd's claim against Cabell Huntington Hospital. These health care providers were not defendants in Fanning and did not contribute financially to the settlement, and their release in a "limited fund" settlement was plainly unlawful. See In re Telectronics Pacing Sys., 221 F.3d 870 (6th Cir. 2000).

2. AcroMed claims that Ms. Lloyd is the only person who lacked notice and subsequently filed a collateral attack. AM 29. In fact, at least one other class member, David Moon, filed such a case, which was also enjoined by the district court. See PTO 1914; see also JA 130-39. Mr. Moon did not appeal. We agree with AcroMed that Ms. Lloyd's suit is not part of an onslaught of litigation by purported members of the Fanning class. See AM 29. Although large numbers of class members were not properly notified of the settlement, see infra at 22-23, at this point, it is unlikely that many people implanted with AcroMed screws on or before December 31, 1996 will bring suit.(3)

3. AcroMed claims that Ms. Lloyd is the only person who lacked notice and subsequently filed a collateral attack. However, class member David Moon did file such a collateral attack, which was also enjoined by the district court. See PTO 1914; see also JA 130-39. Mr. Moon did not appeal.

4. There is no doubt that Ortiz and Fanning presented the same "limited fund" issue. One need look no further than the district court's opinion approving the Fanning settlement, which chiefly relied on the Fifth Circuit's decision in In re Asbestos Litig., 90 F.3d 963 (1996), see JA 99-100, the very decision reversed in Ortiz. Cf. In re Diet Drugs Prods. Liab. Litig., 1999 U.S. Dist. Lexis 14881 (E.D. Pa. Sept. 27, 1999) (Bechtle, J.) (relying on Ortiz to reject non-opt-out "limited fund" settlement indistinguishable in all relevant respects from Fanning).

5. AcroMed's arguments regarding MDL 1014 illustrate a fundamental misunderstanding of Ms. Lloyd's personal jurisdiction argument. The question is not whether the district court did have, or could have had, jurisdiction over Ms. Lloyd when her counsel appeared in MDL 1014 in 1999 (although it did not and could not), but whether that court had jurisdiction over Ms. Lloyd when it certified the class and approved the settlement in 1997. See infra Part I.C.

6. Even if there were a question as to whether there is a due process right to opt out in a purported limited-fund class action such as Fanning, it should be resolved in the West Virginia action. That is true because "[i]t is well settled that the court adjudicating a dispute cannot predetermine the res judicata effect of its own judgment; that can be tested only in a subsequent suit." 7B Federal Practice and Procedure § 1789, p. 245 (2d ed. 1986); accord, e.g., Taunton Gardens v. Hills, 557 F.2d 877, 878 (1st Cir. 1977); Gonzales v. Cassidy, 474 F.2d 67, 74 (5th Cir. 1973); 3 Newberg on Class Actions § 16.24, p. 16-131 (3d ed. 1992).

7. In other cases cited by appellees, where lists were incomplete, there was far more extensive publication notice than occurred here. See, e.g., In re Domestic Air Trans. Antitrust Litig., 141 F.R.D. 534 (N.D. Ga. 1992).

8. In many cases where the class was notified by first-class mail, each plaintiff's claim was relatively small. See, e.g., Shutts, 472 U.S. at 801 ($100 per claimant). If anything, notice here should have been far more extensive because the plaintiffs' individual interests are far larger. See JA 71-72 (average AcroMed settlement $131,000; average verdict $561,500). For that reason, actual notice is constitutionally required in this case. See Mathews v. Eldridge, 424 U.S. 319, 335 (1976). Appellees' arguments regarding actual notice are answered by our opening brief (at 31-34).

9. AcroMed goes so far as to argue that had Ms. Lloyd purchased a newspaper with the Parade insert containing the settlement notice, she would have been notified of the settlement. AM 27-28. To put it mildly, that position is difficult to square with Sambolin or any notion of due process. AcroMed also states incorrectly that Ms. Lloyd's declaration is vague regarding whether she received actual notice. AM 14. She declared unequivocally that she did not have notice of the settlement until told of it by counsel in May 1999. JA 194.

10. Appellees have provided no evidence suggesting that even one class member registered after learning of the case through press reports. On the other hand, hundreds of class members registered "late" because they did not have notice. See infra at 22-23.

11. Appellees' argument assumes a $100 million "limited fund." Actually, the amounts available were larger because the settlement included AcroMed's insurance proceeds and because the company was actually worth far more than $100 million. See infra at 26-27.

12. Because the actual content of the publication notice was constitutionally insufficient to apprise Ms. Lloyd of her rights and obligations, the district court did not obtain jurisdiction over her for that reason as well. Since appellees ignore that issue, we rest on our opening brief (at 45-47).

13. The same effort to "reinterpret" Dr. Rosen's report has occurred before. Dr. Rosen testified at a November 1998 fairness hearing in support of a purported limited fund settlement in In re Telectronics Pacing Systems, MDL 1057 (S.D. Ohio), which has since been rejected by the Sixth Circuit. In re Telectronics Pacing Sys., 221 F.3d 870 (2000). When questioned about his testimony from the AcroMed fairness hearing, Dr. Rosen stated that his valuation of AcroMed was based on the company's value in the absence of the class action settlement. See JA 188. As indicated in the text, however, Dr. Rosen's valuation of AcroMed in Fanning, and the district court's finding premised on that valuation, allegedly was based on AcroMed's value with the class action settlement, i.e. with the litigation behind the company. If appellees believed that $104 million was AcroMed's value absent the settlement (i.e., as reflected in Dr. Rosen's Telectronics testimony, not his Fanning testimony), then appellees presumably knew that the great majority of the company's value would go to AcroMed's shareholders, not to the class members, as had been represented at the fairness hearing. See JA 74. That gross undervaluation would explain why AcroMed sold for $325 million shortly after the settlement was consummated. The lead plaintiffs' lawyer in Telectronics successfully prevented further inquiry on this topic. See JA 189. That lawyer is also a member of the PLC.