Reiter v. Zimmer Inc., 897 F. Supp. 154 (S.D.N.Y. 1995)

U.S. District Court for the Southern District of New York - 897 F. Supp. 154 (S.D.N.Y. 1995)
September 8, 1995

897 F. Supp. 154 (1995)

Nita REITER, Plaintiff,

No. 91 Civ. 8599 (LMM).

United States District Court, S.D. New York.

September 8, 1995.

*155 Lewis Rosenberg & Barry I. Levy, Shapiro, Shiff, Beilly, Rosenberg & Fox, New York City, for plaintiff.

James P. Barrett & Daniel Keenaghan, Simpson Thacher & Bartlett, New York City, Daniel K. Leininger, Lemon, Armey, Hearn & Leininger, Warsaw, IN, for defendant.



McKENNA, District Judge.

Plaintiff Nita Reiter ("Reiter") commenced this action against Defendant Zimmer Incorporated ("Zimmer") alleging that the premature hardening of Zimmer bone cement during hip replacement surgery caused her debilitating personal injuries. Zimmer moved for summary judgment, arguing that Reiter's tort claims were preempted by the Federal Food, Drug, and Cosmetic Act ("FDCA"), 21 U.S.C. §§ 301 et seq., and the Medical Device Amendments of 1976 ("MDA"), 21 U.S.C. §§ 360c et seq.

In an Order dated August 17, 1993, the Court held that Reiter's strict liability claim was preempted by federal law, but her negligent manufacture claim was not, as Reiter "allege[d] that Zimmer did not comply with *156 its own FDA approved manufacturing process." Reiter v. Zimmer, Inc., 830 F. Supp. 199, 204 (S.D.N.Y.1993). Following the completion of discovery on this issue, Zimmer renewed its motion for summary judgment, arguing both that Reiter has produced no evidence of non-compliance with the FDA approved process, and that Reiter's negligent manufacturing claim is preempted. For the reasons discussed below, Zimmer's motion is denied.


I. Failure of Zimmer Bone Cement

Dr. Victor Frankel explained the difficulties he encountered replacing Nita Reiter's left hip with a prosthesis:

My technique is to use a cement gun [to fill the femoral canal so as to receive the prosthesis], mix the cement in a cup and then pour it into the gun in two minutes and then inject it. During the process of injecting the cement, it started to become very firm. The cement was then removed from the gun and finger packed into the canal. The prosthesis was immediately inserted. Prior to being fully seated the cement hardened with excessive speed. Within five minutes of the time the cement was first mixed it was completely hard and the prosthesis was fixed in the femur half in and half out. In the attempt to insert the femur in the presence of rapidly hardening cement, part of the femoral shaft fractured due to the increased hoop stresses. It was then necessary to remove the prosthesis and the cement.
. . . . .
The average setting time for the Zimmer bone cement is usually 13 to 17 minutes in our operating room. Setting time of five minutes is quite unexpected and does not allow the surgeon to properly implant the prosthesis.

(Letter from Dr. Frankel to Pl.'s Att'y, dated April 15, 1991.)

Zimmer bone cement consists of two components: a powder and a liquid. When the two are mixed in the operating room, a chemical reaction begins, and the mixture starts to set or "cure." After a few minutes the mixture becomes a soft putty and can be used to fix the prosthesis. In about 10 minutes, the mixture hardens and the prosthesis cannot be moved. (Stanton[1] Report at 2.) The cement hardens due to the presence of benzoyl peroxide ("BPO") in the powder. (Id. at 4.) "The BPO concentration is critical to the reaction rate of the mixed Zimmer Bone Cement and, so, to the length of time available to the surgeon to use the mixed cement." (Id. at 17.)

Reiter contends that the batch of cement used in her operation may have hardened prematurely because it was improperly formulated and blended. It is uncontested that batch # 66688200 twice failed a cure test during the blending stages of the formulation:

The BPO apparently was not added in accordance with the batch instructions because a note on the Set-up Sheet for this stage reads, "Because of failure on the 1st 45,000g (grams) cure test ... only 3,435g (grams) was added ...," page 3 of 4. From the dates involved, this failure appears to be reported on CPL 88-989, "Failed," December 7, 1988, and on CPL 88-992, "Reblend, Failed," December 7, 1983. Two days later, "Reblend # 2, Passes," was recorded on CPL 88-998, December 9, 1988.

(Stanton Report at 5.) Stanton suggests that this "attempt to `coax' th[e] defective batch into conformity" by adding more BPO may have been a substantial factor in the cement's premature hardening. (Stanton Aff. ¶ 24.)


II. Regulation of Class III Medical Devices

Zimmer's Bone Cement is classified as a Class III medical device since "it presents *157 a potential unreasonable risk of illness or injury." 21 U.S.C. § 360c(a) (1) (C) (ii) (II) (1972 & 1995 Supp.) A manufacturer seeking approval of a Class III device submits a New Drug Application ("NDA") to the FDA, which includes the company's proposed procedures for manufacturing the device. If the procedures are approved, the manufacturer is thereafter required to follow them, unless permission to alter those procedures is later obtained:

The premarket approval process is designed to provide a "reasonable assurance of ... safety and effectiveness" for medical devices which are too dangerous or unknown to permit less regulation. 21 U.S.C. § 360c(a) (1) (C). Post-approval regulation is designed to keep the FDA apprised of ongoing safety findings or any other information about the device as it becomes available. Id. §§ 360e(e) & 360i(a).
Pursuant to the premarketing approval process, the FDA requires applicants to submit proposed labeling, extensive safety testing data and descriptions of manufacturing methods and materials. Id. § 360e(c) (1). Upon reviewing the materials in a comprehensive manner, the FDA may approve the device for sale or return the application to the applicant for further information or testing. Id. § 360e(d) (1) ... Once the device is approved, the FDA retains the power to withdraw approval of the product permanently or suspend its approval temporarily if it determines that the device has become unsafe or its labeling inadequate. Id. § 360e(e) (1)-(3). To assist the FDA in making these determinations, manufacturers must maintain records and make reports to the FDA on information pertinent to the device. Id. § 360i(a).

King v. Collagen Corp., 983 F.2d 1130, 1131-32 (1st Cir.), cert. denied, ___ U.S. ___, 114 S. Ct. 84, 126 L. Ed. 2d 52 (1993).

From 1976 through the date of Nita Reiter's surgery, Zimmer could make no change in the design, manufacture or label affecting the safety or effectiveness of Zimmer bone cement without the prior review and approval of the FDA. (See the FDA's listing of examples of changes requiring FDA pre-approval at 21 CFR § 814.39.) In addition to the requirement that the FDA pre-approve such changes, Zimmer continued with the same record-keeping and reporting duties as had existed when bone cement was marketed under the NDA. See 21 CFR §§ 814.82 and 814.84.

(Def.'s Reply Mem. at 4.)


III. Alleged Failure to Comply with the NDA

Stanton contends that the premature hardening of the bone cement used in the operation resulted from Zimmer's failure to comply with the FDA-approved NDA, that is, with the very procedures proposed by Zimmer. In his final report, Stanton concludes:

[I]t is my professional opinion to a reasonable degree of chemical engineering certainty, [that] the batch of Zimmer bone cement from which the kit used in Mrs. Reiter's surgery was not produced in accordance with FDA approved standards which resulted in the failure (premature hardening) of this critical device, which in turn, resulted in injury to her during the surgery of September, 1989.

(Stanton Aff. ¶ 26.)

While Stanton enumerates several different "failures to comply," the Court need not at this time reach highly technical questions relating to Zimmer's alleged non-compliance with procedures described explicitly in the NDA.[2] The Court, instead, focuses on Zimmer's *158 alleged failure to comply with industry and FDA standards that it chose to "incorporate by reference" in the NDA.

Stanton complains that Zimmer failed to employ procedures to ensure that the batch of bone cement was sufficiently homogeneous (uniform). The issue of homogeneity was addressed in the NDA both explicitly, by the detailing of certain blending procedures, and implicitly, by the incorporation of industry standards. One such implicit reference appears in Zimmer Engineering Specification ("ZES") 2R-01, which states that a quality program in conformity with industry standard ASQC C1 1985 will be maintained. (ZES 2R-01 at 5.2) This standard, promulgated by the American Society for Quality Control, requires, inter alia, that "[c]ontrol shall be maintained over production processes to prevent product ... non-conformity and excessive variability." (ASQC C1 1985 at §§ 3.5, 3.5.1) The standard also requires that "methods and facilities shall be provided to assure conformance with requirements for special process specifications, such as ... homogenization ..." (Id. at 3.5.3) Stanton contends that Zimmer's bone cement was formulated in a fashion that resulted in "excessive variability" and did not produce a homogeneous product, which were "substantial contributing factors causing the failure of the Zimmer Bone cement leading directly to Ms. Reiter's injury." (Stanton Aff. ¶ 13.)

Stanton further observes that ZES 2R-01 references the FDA's Good Manufacturing Practices for Medical Devices, 21 C.F.R. § 820 et seq. Section 820.5 of "Good Manufacturing Practices" requires "a quality assurance program that is appropriate to the specific device manufactured." Section 820.161 requires an investigation when a critical device fails to meet performance specifications.

Reiter's position is that if Zimmer had chosen to omit such references, and FDA approval were received, then Zimmer would only be obligated to follow the manufacturing processes explicitly described in the NDA. Since Zimmer chose to reap the benefits of incorporating industry standards (presumably making FDA approval more likely), Reiter argues, Zimmer must be held to those standards as well, even if they impose requirements not specifically enumerated in the NDA.

To clarify the technical arguments relating to the early formulation problems, the Court conducted a hearing on June 8, 1995:

Def.'s Att'y: [T]hey are taking one 45-kilogram barrel, and they are putting in the ingredients. They then mix it and test it. And they found that they had to make more additions. That was a failure, No. 1. So they put in more ingredients into this one barrel, mixed it, tested it again. That was failure No. 2. And then they put in more ingredients, mixed it, and it passed the test.

(Tr. at 31.)

Pl.'s Att'y: Mr. Stanton says that, in order to cure that failure, they coaxed it into conformity by altering the concentration.
At this particular point, Judge, we were dealing with something which the FDA does not control by any Zimmer specifications. They said to the FDA that they were going to blend A, B, C and D. But they didn't say to the FDA, "When it fails, here is how we are going to correct it. We have a procedure to correct it." They adopted this procedure.
And the point I'm making, your Honor, is that, although there are a good many details of the process that have been submitted to the FDA, there were some gaps. Those gaps are filled by the engineering specifications where they adopt industry standards to fill those gaps because the objective of the homogeneity is essential for this stuff to work the way it is supposed to....
The FDA has been told that they will put so many grams of this into so many grams of that. And they will achieve totals and then comply with the tests. But they don't tell the FDA how they will coax it into conformity. There's no specific part of the formula that has been submitted and approved that governs that. That's where the general specifications take over.
Mr. Leininger is very critical of Mr. Stanton's illustration of these industry standards. If they were not useful or *159 analyzed or necessary, why would Zimmer adopt these? They didn't have to. Nobody twisted their hands, as far as I know.

(Id. at 35. (emphasis supplied))

The Court understands the essence of this argument to be that while it may be possible to pass one particular test, the cure test, by altering the proportion of ingredients, the alteration might result in a defective final product. Furthermore, the alteration was not specifically contemplated by the FDA during its review of the NDA and would not constitute a proper discretionary practice. Good engineering practice as well as common sense might suggest discarding a batch that has twice failed a cure test, or inquiring into the basis for the failure (such as improper environmental controls, or substandard ingredients), or conducting additional testing.[3]

The Court finds Stanton's argument as to the incorporation of industry and FDA standards to be persuasive, particularly in light of the affidavit submitted by Carl A. Larson, who served as the FDA's Director of the Division of General and Restorative Devices from 1979 to 1992. This Division had primary regulatory responsibility for bone cement, among other medical devices. (Larson Aff. at ¶ 3.) Larson, who during his tenure was acquainted with both Zimmer personnel and Dr. Frankel (a member of the FDA's Orthopaedic Advisory Panel, which made recommendations to Larson's division), reviewed the affidavit and reports of George Stanton. Larson states: "I agree with Mr. Stanton that bone cement manufacturers, such as Zimmer, are bound to adhere to Good Manufacturing Practices promulgated by the FDA and I agree with the construction that Mr. Stanton places upon these regulations." (Id. ¶ 8.)[4]

The Court might well have come to the same conclusion even if Zimmer had not chosen to incorporate industry and FDA standards by reference. An obligation to follow good engineering practice (as distinct from the FDA's Good Manufacturing Practices) may arise by operation of state law, see, e.g., N.Y.Educ.Law §§ 7200-09, §§ 6509-15 (McKinney 1985) (concerning, respectively, the licensing and the misconduct of professional engineers), or, for the unlicensed engineer, as a professional obligation.[5] Rather than seeing such requirements *160 as additional burdens imposed by state law, which, as discussed infra, would be preempted by federal law, the Court instead views such professional obligations as the FDA undoubtedly does as inherent in all medical manufacturing. Since there will undoubtedly be formulation issues that "fall between the cracks" of the NDA, the FDA must assume, particularly in the context of a Class III medical device, that the manufacturer will follow "good engineering practice."

The Court's conclusion regarding "incorporation by reference" renders many of Zimmer's arguments irrelevant. For example, Zimmer has argued that the two failed cure tests were not required by the NDA as approved by the FDA in 1982. (Letter from Def.'s Att'y to Court of June 22, 1995 at 3.) Even if these tests were not explicitly detailed in the NDA, the two failures may have triggered further obligations, such as an inquiry into their cause, either under the standards incorporated by reference, or under the general professional standard of good engineering practice.


IV. Scope of FDCA/MDA Preemption

Zimmer's primary argument on its renewed motion for summary judgment is that Reiter's negligent manufacturing claim is preempted by federal law. The Court disagrees, finding the caselaw barring state claims to be distinguishable from the case at bar.

The MDA states that:

(a) Except as provided in subsection (b) of this section, no State or political subdivision of a State may establish or continue in effect with respect to a device intended for human use any requirement
(1) which is different from, or in addition to any requirement applicable under this chapter to the device, and
(2) which relates to the safety or effectiveness of the device or to any other matter included in a requirement applicable to the device under this chapter.

21 U.S.C. § 360(k) (emphasis supplied).

If the FDA has issued requirements for a device, subsection (a) prohibits states from imposing any requirements which differ from or add to the FDA requirements, or which relate to the safety or effectiveness of the device. A "State ... requirement" in subsection (a) may emanate from any requirement established by a state including statutes, regulations, court decisions or ordinances. 21 C.F.R. § 808.1(b); see also San Diego Building Trades Council v. Garmon, 359 U.S. 236, 247 [79 S. Ct. 773, 780, 3 L. Ed. 2d 775] (1959) ("[State] regulation can be as effectively exerted through an award of damages as through some form of preventive relief. The obligation to pay compensation can be, indeed is designed to be, a potent method of governing conduct and controlling policy.").
The language of subsection (a) and the definition of state requirement promulgated under it demonstrate a field of preemption which is broad, but limited. Any state requirement which, in effect, establishes a new substantive requirement for the device in a regulated area such as labeling, is preempted.

King, 983 F.2d at 1134 (emphasis supplied) (parallel citations omitted).

While § 360(k) bars tort claims that impose requirements "different from, or in addition to" federal requirements, the plain language of the statute supports Reiter's claim that a tort claim is not preempted where it seeks only to enforce existing federal requirements.[7]

When Congress has considered the issue of preemption and has included in the enacted legislation a provision explicitly addressing that issue, and when that provision provides a "reliable indicium of congressional intent with respect to state authority," Malone v. White Motor Corp., 435 U.S., [497] at 505, 98 S.Ct., [1185] at 1190 *161 [55 L. Ed. 2d 443], "there is no need to infer congressional intent to preempt state laws from the substantive provisions" of the legislation. California Federal Savings & Loan Assn. v. Guerra, 479 U.S. 272, 282 [107 S. Ct. 683, 689, 93 L. Ed. 2d 613] (1987). Such reasoning is a variant of the familiar principle of expressio unius est exclusio alterius: Congress' enactment of a provision defining the preemptive reach of a statute implies that matters beyond that reach are not preempted.

Cipollone v. Liggett Group, 505 U.S. 504, 517, 112 S. Ct. 2608, 2618, 120 L. Ed. 2d 407 (1992); see also CSX Transp., Inc. v. Easterwood, ___ U.S. ___, ___, 113 S. Ct. 1732, 1737, 123 L. Ed. 2d 387 (1993) ("If the statute contains an express preemption clause, the task of statutory construction must in the first instance focus on the plain wording of the clause, which necessarily contains the best evidence of Congress' preemptive intent."); United States v. Goldberger & Dubin, P.C., 935 F.2d 501, 506 (2d Cir. 1991) ("[W]ords of a statute should be given their normal meaning and effect in the absence of a showing that some other meaning was intended.").

In Cipollone, the Court held that the word "requirement," in a statutory context, was not limited to positive enactments by states and localities, but could reach to common law actions. 505 U.S. at 521-22, 112 S. Ct. at 2620. The Supreme Court observed, however, that such language did not automatically bar all state claims: "we must fairly but in light of the strong presumption against preemption narrowly construe the precise language of [the statute] and we must look to each of petitioner's common law claims to determine whether it is in fact preempted." Id. at 523, 112 S. Ct. at 2621.

Since the determination of which common law claims are preempted requires a substantive statutory analysis, a state action does not impose an "additional requirement" simply because it might give rise to an obligation to pay money damages that might not arise under federal law. The phrase "additional requirement" appears instead to relate to concerns that common law standards will interfere with federal standards.

In the instant action state law may not require that Zimmer manufacture its bone cement in a different manner than has already been approved by the FDA. For example, a determination that Zimmer was negligent for failure to perform additional testing of the cement during formulation, beyond any testing required by the FDA, would clearly be a "different" or "additional" requirement imposed by state law that is preempted. It is not improper, however, to subject Zimmer to a state claim grounded solely in non-compliance with existing FDA requirements such as the failure to follow "good engineering practice" if it was the FDA's understanding that such practices would be employed.

The Court rejects Zimmer's argument that any jury determination in the instant action would amount to the imposition of jury standards for the manufacturing of its bone cement:

Plaintiff's state law claims would impose requirements related to the safety and effectiveness of Zimmer bone cement. Using the words of King, supra, "if successful, the claims[s] would require [Zimmer] to redesign [bone cement]...."

(Zimmer Reply Mem. of August 9, 1993, at 9.) Clearly, a tort action that merely enforces existing FDA manufacturing standards does not require a "redesign" of Zimmer's bone cement and does empower a jury to rewrite federal standards.

The caselaw relating to MDA preemption is largely concerned with the adequacy of labeling and with jury-set standards, neither of which is at issue in the instant action. See, e.g., Cipollone, 505 U.S. at 523-25, 112 S. Ct. at 2621-22 ("[I]nsofar as claims under either failure to warn theory require a showing that respondents' post-1969 advertising or promotions should have included additional, or more clearly stated, warnings, those claims are preempted."); CSX Transp., ___ U.S. at ___, 113 S. Ct. at 1743 (where train had complied with federally established speed limit, claim of excessive train speed which would have allowed jury-set standard was necessarily preempted, but not remainder of negligence claim); Slater v. *162 Optical Radiation Corp., 961 F.2d 1330, 1333 (7th Cir.) ("The plaintiff wishes in the name of state tort law to impose additional requirements namely that [the medical device] have had design characteristics that it lacked...."), cert. denied, ___ U.S. ___, 113 S. Ct. 327, 121 L. Ed. 2d 246 (1992); King, 983 F.2d at 1136 ("[A] finding of negligence would force [the manufacturer] to alter these aspects of [its product] in response to the finding of liability, or be subject to liability."); cf. Hunsaker v. Surgidev Corp., 818 F. Supp. 744, 751 (M.D.Pa.1992) (new requirements are imposed where the fear of liability will "induce manufacturers to conform their conduct to jury-set standards"), aff'd, 5 F.3d 1489 (3d Cir.1993).

Zimmer would have the Court uncritically read the caselaw to suggest a wholesale ban on state tort litigation. While other circuits have held on rather different facts that the FDCA does not provide or imply a private cause of action for damages, see, e.g., Pacific Trading Co. v. Wilson & Co., 547 F.2d 367, 370 (7th Cir.1976); Gile v. Optical Radiation Corp., 22 F.3d 540, 544 (3d Cir.1994) (stating, in the context of an "investigational device exemption" that "violations of the FDCA do not create private rights of action"), cert. denied, ___ U.S. ___, 115 S. Ct. 429, 130 L. Ed. 2d 342 (1994), no blanket prohibition on tort claims has been endorsed by this circuit. See, e.g., LaMontagne v. E.I. Du Pont De Nemours & Co., 41 F.3d 846, 855 (2d Cir.1994) (affirming district court's determination that state tort claim was not preempted by the MDA where factors such as "different or additional" requirements were not satisfied).

Where a state statute conflicts with or frustrates federal law, the former must give way. U.S. Const. art. VI, cl. 2; Maryland v. Louisiana, 451 U.S. 725, 746, 101 S. Ct. 2114, 2128, 68 L. Ed. 2d 576 (1981). Where a common law action does no more than help to enforce existing federal law, with no substantive modifications, no such conflict has occurred. See, e.g., Silkwood v. Kerr-McGee Corp., 464 U.S. 238, 104 S. Ct. 615, 78 L. Ed. 2d 443 (1984) (incidental regulatory consequences of application of state tort law may be acceptable where direct regulatory authority is preempted). Since state and federal manufacturing requirements are identical in such a case, it cannot be said that the two standards are merely consistent, in which case preemption might still be appropriate. See Stamps v. Collagen Corp., 984 F.2d 1416, 1424 (5th Cir.) (citing Morales v. Trans World Airlines, Inc., 504 U.S. 374, 385, 112 S. Ct. 2031, 2038, 119 L. Ed. 2d 157 (1992)), cert. denied, ___ U.S. ___, 114 S. Ct. 86, 126 L. Ed. 2d 54 (1993). Reiter's negligent manufacturing claim, limited in scope to questions of compliance with existing FDA regulations and questions of proximate causation and damages, is an appropriate mechanism in support of federal law,[8] and is therefore not preempted.


V. Negligent Manufacture

Having determined that Reiter's claim of negligent manufacture is not preempted by federal law, the Court must determine whether Zimmer is entitled to summary judgment on the alternate ground of a failure of Reiter's proof of negligent manufacture. On a motion for summary judgment, the Court's proper role is identifying *163 factual issues, not resolving them. In support of this branch of its motion, Zimmer has argued, inter alia, that George Stanton lacks the necessary qualifications to come to the technical conclusions that he has reached and that Reiter has failed to adduce evidence of failure to comply with FDA regulations. The Court, however, finds that the Larson and Stanton affidavits raise a number of substantial factual questions relating to Zimmer's manufacturing practices that must be addressed at trial.[9]

In sum, Reiter has raised substantial factual questions as to Zimmer's compliance with FDA requirements regarding the formulation of batch # 66688200 requiring Zimmer's motion for summary judgment be denied.



[1] George B. Stanton, Jr., Reiter's expert witness, is a licensed professional engineer with a B.Ch.E. and M.Ch.E. in Chemical Engineering and an M.A. in Health and Safety. Stanton has submitted both a technical analysis and report dated May 16, 1994 ("Stanton Report") and an updated report dated May 4, 1995 ("Stanton Aff."). Stanton's reports were based on a review of Zimmer's New Drug Application to the FDA, Zimmer's correspondence with the FDA, Zimmer's engineering specifications, and entries made in laboratory notebooks during the formulation of the batch of cement used in Reiter's operation, # 66688200. Stanton also examined a sample of cement from this batch, but could not directly confirm that the batch was defective, since it was already "out-of-date." (Stanton Dep. at 8.)

[2] By limiting its review of Stanton's arguments to the relatively non-technical issue of "incorporation by reference," the Court sought to avoid playing amateur scientist. As Chief Justice Rehnquist noted, the law does not impose on judges "either the obligation or the authority to become amateur scientists." Daubert v. Merrell Dow Pharmaceuticals, ___ U.S. ___, ___, 113 S. Ct. 2786, 2800, 125 L. Ed. 2d 469 (1993) (Rehnquist, C.J., concurring in part and dissenting in part). See also In re Joint E. & S. Dist. Asbestos Litig., 52 F.3d 1124, 1137 (2d Cir.1995) (reversing the district court's determination that plaintiff had submitted insufficient scientific evidence of causation, where the court had "impermissibly made a number of independent scientific conclusions").

[3] Zimmer has argued that altering the proportion of ingredients was harmless, analogizing it to a man working on a lathe: "What are the odds of a production worker putting a piece of steel on the lathe, grinding it, stopping it and hitting the specifications on that particular one. Everybody knows that it is impossible or it would be entirely coincidental if he did." (Tr. at 46.) The analogy between the routine working of a piece of steel and the tinkering with the chemical formulation of a federally regulated critical medical device is a poor one. Altering the proportion of ingredients in Zimmer bone cement might have substantially altered its carefully designed properties.

[4] Larson also agrees with Stanton that Zimmer appears to have implemented an engineering change that "substantially degraded the quality standard for BPO," without receiving the requisite FDA approval for the change. (Stanton Aff. ¶ 15.) Larson states:

I have also observed in these papers, that Zimmer's attorney claims that this change was approved by the FDA by means of Zimmer's reference to this change included in its First Annual Progress Report filed with the FDA in 1979.

I do not however, agree that obscure references such as this, in an annual report filed with the FDA, constitutes FDA review and approval, required by 21 C.F.R. § 814.39, without specific written acknowledgment that such a change is approved by the FDA.

It is elementary that any change in the process described as approved in a PMA that affects the efficacy or safety of a medical device must be specifically and explicitly approved before it may be implemented.

(Larson Aff. ¶¶ 10-12 (emphasis supplied).)

[5] Fundamental Canon and Rules of Practice

I. Engineers shall hold paramount the safety, health and welfare of the public in the performance of their professional duties.

a. Engineers shall at all times recognize that their primary obligation is to protect the safety, health, property and welfare of the public. If their professional judgment is overruled under circumstances where the safety, health, property or welfare of the public are endangered, they shall notify their employer or client and such other authority as may be appropriate.

Deborah G. Johnson, Ethical Issues in Engineering 98-99 (1991) (quoting National Society of Professional Engineers Code of Ethics); see also Mike W. Martin & Roland Schinzinger, Ethics in Engineering 341-61 (1989) (sample Codes of Ethics).

[7] FDA preemption applies only where state requirements are "different from, or in addition to" federal requirements and "relate[] to the safety or effectiveness of the device." 21 U.S.C. §§ 360(k) (a) (1)-(2). The broad reach of the second clause, Morales v. Trans World Airlines, 504 U.S. 374, 112 S. Ct. 2031, 119 L. Ed. 2d 157 (1992), need not be considered as the Court holds that Reiter's negligent manufacturing claim fails to come within the first clause.

[8] In Bravman v. Baxter Healthcare Corp., 842 F. Supp. 747, 760-61 (S.D.N.Y.1994), Judge Sweet, although holding that a state "duty to warn" claim imposed "additional" requirements, and was therefore preempted by the MDA, nevertheless presented compelling reasons against federal preemption. First, the state law tort system acts as an incentive for manufacturers to continue to improve their products. Second, blanket preemption may encourage manufacturers to withhold disclosure of potential problems with their devices to the FDA. Finally, preemption would lead many deserving plaintiffs without any remedy. "Too ready a tendency to declare the state protective shield replaced by the still somewhat spotty federal protections will leave many injured persons without recourse." Id. at 761 (citing Burke v. Dow Chem. Co., 797 F. Supp. 1128, 1132 (E.D.N.Y.1992) (Weinstein, J.)). Finally, Judge Sweet quotes the former Chief Counsel of the FDA: "[the] FDA has no expertise or authority for managing systems of redress for private injuries. The value judgment necessary for such management how the scales for plaintiffs and defendants should be set up are best left to legislatures and courts." Id. (citing Richard M. Cooper, Drug Labeling and Products Liability: The Role of the Food and Drug Administration, 41 Food Drug Cosm.L.J. 233, 233 (1986).)

[9] The record contains several other indications that Zimmer may have disregarded "good engineering practice." For example, Zimmer appears to have consistently employed the minimum permitted blending times:

Each load was blended for 10 minutes, sifted and reblended for a further 10 minutes. The Set-up Sheet calls for 10-25 minutes for the first blend for each load and, for the reblend, 10-60 minutes.

The minimum times were reduced on June 14, 1988; for the first blend, from 20 to 10 minutes; for the reblend, from 30 minutes without the intensifier bar in operation to 10 minutes with it; and for additional blending if required by QC testing, from 15 minutes without the intensifier bar in operation to 10 minutes with it.

(Stanton Report at 5.)

Change number 8 on Set-Up Sheet XX-XXXX-XXX-XX.

FROM: ... Blend for twenty (20) + / - five (5) minutes ...

TO: ... blend for 10 to 25 minutes ...

(Zimmer Operations Control Procedure, Revisions Summary Sheet, Ex. 11 to Case Aff.) When an engineer specifies a range of numerical values to be used in a manufacturing process, it is commonly understood that the mean (or central) value is the desirable value, but that deviations may be permissible or unavoidable.

It appears that Zimmer simply chose the minimum value within the range, which may have been inappropriate for a critical medical device. See Martin & Schinzinger, supra note 5, at 100 ("one of the greatest moral problems in engineering ... is that of minimal compliance").