Eli Lilly & Co.v.Marshall

Supreme Court of TexasApr 14, 1993
850 S.W.2d 155 (Tex. 1993)

No. D-2003.

February 3, 1993. Dissenting Opinion by Justice Doggett April 14, 1993.

Mark E. Smith, Robert G. Hogue, Dallas, Joe C. Freeman, Jr., Atlanta, GA, Wade C. Smith, Dallas and John L. Hill, Houston, for relators.

Paul L. Smith and William V. Dorsaneo, III, Dallas, for respondent.


OPINION


CORNYN, Justice.

Today we consider a second mandamus arising out of a products liability suit against the manufacturer of the antidepressant drug Prozac. Eli Lilly and Company, the manufacturer, seeks to compel the Honorable John Marshall, Respondent, to set aside his order requiring it to disclose certain information rendered confidential by federal regulation and sanctioning it for its failure to do so. Concluding that the trial court did not apply the proper legal standard, and that appeal is not an adequate remedy, we conditionally grant the writ.

See Eli Lilly Co. v. Marshall, 829 S.W.2d 157 (Tex. 1992) (issuing conditional writ of mandamus directing trial court to conduct hearing under Tex.R.Civ.P. 76a concerning trade secrets).

Michael Hays Biffle committed suicide six days after he began taking Prozac. His family and estate filed suit against Lilly and sought production of, among other things, various documents that Lilly had submitted to the federal Food and Drug Administration (FDA) to secure approval to market Prozac. The request for production included any adverse reaction or drug experience reports, which are submitted by physicians and other health care providers to report post-approval possible adverse reactions to a drug. The FDA uses this information to monitor clinical reactions to a drug to assess the terms and conditions of FDA approval or to consider whether to recall its approval entirely. 21 U.S.C. § 355(e). Although submission of a report to the manufacturer is voluntary by the health care provider, the manufacturer must submit any such reports it receives to the FDA. 21 U.S.C. § 355(k)(1). According to federal regulation, the FDA must keep confidential the identities of the patient and of the person or institution that reported the adverse reaction. 21 C.F.R. Sec. 314.430(e)(4)(ii) (1991). The present dispute concerns the identity of the reporters of possible adverse reactions to Prozac.

A report does not presume causal relation to the product. Drug Experience Report-Form 1639.


(4) Adverse reaction reports, product experience reports, consumer complaints, and other similar data and information after deletion of the following:

(ii) Names and any information that would identify any third party involved with the report, such as a physician or hospital or other institution.

The same confidentiality provisions can be found at 21 C.F.R. Sec. 20.111(c)(3)(ii)(b) (1991) (data and information submitted voluntarily to the FDA), and 21 C.F.R. Sec. 312.130(b) (1991) (data and information in an investigational new drug application).

The regulation provides in pertinent part:


Sec. 314.430(e) After FDA sends an approval letter to the applicant [the manufacturer], the following data and information in the application are immediately available for public disclosure, unless the applicant shows that extraordinary circumstances exist.

On October 23, 1991, following a hearing on the Biffles' motion to compel and Lilly's motion for a protective order based in part on the regulations at issue, the trial court ordered Lilly to produce the adverse reaction reports with only the patients' names redacted. When the Biffles' attorney arrived at Lilly's headquarters to inspect the documents, he found, among other alleged instances of noncompliance with the discovery order, that the names and addresses of the reporters had been redacted as well. Lilly's proffered explanation for its noncompliance is that its lawyers expected that maintenance of reporter and patient anonymity would be agreed upon by the parties. The Biffles, however, moved for sanctions, and following another hearing the trial court ordered Lilly to disclose the identity of the reporters and to pay the travel and copy expenses of the Biffles' attorney, and ruled that failure to comply would result in a default judgment in favor of the Biffles.

We first address the propriety of mandamus relief. Mandamus will issue only to correct a clear abuse of discretion when there is no adequate remedy by ordinary appeal. Walker v. Packer, 827 S.W.2d 833, 839 (Tex. 1992). Although this court may not substitute its judgment for that of the trial court with regard to fact issues, Johnson v. Fourth Court of Appeals, 700 S.W.2d 916, 918 (Tex. 1985), we accord the trial court's analysis and application of the appropriate law far less deference. Walker, 827 S.W.2d at 840. When the undisputed facts and the applicable law permit of but one lawful decision, this court is called on to ensure that decision is reached. Id. We will exercise our mandamus jurisdiction in a case, such as we conclude this to be, in which an ordinary appeal could not cure the error presented. Id. at 843.

Lilly contends that the federal regulations establishing confidentiality preempt the trial court's order compelling disclosure of the reporters' identities. It also contends that revealing the identities of reporters would destroy the voluntary reporting system that is essential to post-approval monitoring of Prozac and thus jeopardize the vital public interest in the free flow of adverse reaction reports essential to assessing the long-term safety of this and other FDA-approved drugs. The Biffles, on the other hand, contend that the FDA regulations apply only to the FDA, that the regulations themselves permit disclosure under proper court order, and that the regulations were not meant to and do not preempt Texas law of tort or discovery.

21 C.F.R. § 20.83(a) (1991) provides in pertinent part:


Records of the Food and Drug Administration which the Commissioner has determined are not available for public disclosure, either in the form of a regulation . . . or by a written determination . . ., shall nevertheless be made available for public disclosure in compliance with a final court order requiring such disclosure.

The FDA has filed a statement of interest with this court. While disclaiming any interest in the merits of this lawsuit, the FDA emphasizes the importance of post-approval reporting because certain kinds of problems, particularly those arising from long-term use, would not come to the attention of the FDA otherwise, or at least not in the volume reported under the current system. If not assured confidentiality, the FDA contends, reporters would probably not report possible adverse reactions because of fear of entanglement in litigation and potential violation of the physician-patient privilege. Finally, although conceding that the regulations themselves apply only to the FDA, the FDA asserts that the public policy interest the regulations embody are of sufficient importance to preempt any disclosure that would undermine the FDA's post-approval monitoring system.

The FDA argued the same interests to the trial court at a hearing on Lilly's motion for reconsideration. The trial court denied that motion.

The regulations at issue were promulgated in an effort to comply with the Freedom of Information Act. 5 U.S.C. § 552. In responding to comments on the proposed regulations, the FDA Commissioner stated:


The agency has concluded that the release of any names contained in a medical file is clearly unwarranted, except in extraordinary circumstances. . . . With respect to complaints received voluntarily from third parties, usually health professionals, i.e., doctors, nurses, pharmacists, and so forth, relating to such matters as adverse reactions they have observed, and which thus relate to complaints made on behalf of other persons, the Commissioner concludes on the basis of the longstanding experience of the Food and Drug Administration that it is essential to pledge that all identifying information will be deleted prior to public disclosure, and [21 C.F.R. Sec. 314.430(e)(4)(ii) ] so provides. If such a pledge is not made, the possibility of persuading health professionals voluntarily to submit important adverse reaction information on marketed products to the Food and Drug Administration is substantially diminished, and indeed perhaps wholly destroyed. Such information is important to the Food and Drug Administration and to the public, since it may well lead to action by the Food and Drug Administration designed to protect the public health. Accordingly, the Commissioner concludes that deletion of all such identifying information from such reports prior to release to the public is fully within the intent of the personal privacy and confidential commercial information exemptions.

Affidavit of Joyce M. Johnson, Acting Director of the FDA's Division of Epidemiology and Surveillance, Center for Drug Evaluation and Research, p 7.

The genesis of federal preemption is Article VI, clause 2 of the United States Constitution: "the Laws of the United States . . . shall be the supreme law of the Land; and the Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding." Federal law may preempt state law in several ways. When acting within constitutional limits, Congress may preempt state law by so stating in express terms. Hillsborough County, Fla. v. Automated Medical Labs., Inc., 471 U.S. 707, 713, 105 S.Ct. 2371, 2375, 85 L.Ed.2d 714 (1985). Alternatively, Congress' intent to preempt state law may be inferred from its complete and comprehensive regulation of an area. Finally, even if Congress has not completely displaced state regulation, state law is preempted to the extent it actually conflicts with federal law. English v. General Elec. Co., 496 U.S. 72, 78-79, 110 S.Ct. 2270, 2274-75, 110 L.Ed.2d 65 (1990); NCNB Tex. Nat'l Bank v. Cowden, 895 F.2d 1488, 1494-95 (5th Cir. 1990).

Regulations have the same preemptive effect as statutes. Hillsborough County, 471 U.S. at 713, 105 S.Ct. at 2375; Fidelity Fed. Savings and Loan Ass'n v. De la Cuesta, 458 U.S. 141, 153, 102 S.Ct. 3014, 3022, 73 L.Ed.2d 664 (1982). Moreover, federal law preempts conflicting judicial action as well as conflicting statutes and regulations. Wisconsin Dept. of Indus., Labor and Human Relations v. Gould Inc., 475 U.S. 282, 286, 106 S.Ct. 1057, 1060, 89 L.Ed.2d 223 (1986); Texas Employers' Ins. Ass'n v. Jackson, 820 F.2d 1406, 1412 (5th Cir. 1987); Macmillan v. Redman Homes, Inc., 818 S.W.2d 87, 95 (Tex.App. — San Antonio 1991, writ denied).

When determining if a state law actually conflicts with federal law, the question presented is whether compliance with both state and federal law is impossible, or whether the state law stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress. Capital Cities Cable, Inc. v. Crisp, 467 U.S. 691, 699, 104 S.Ct. 2694, 2700, 81 L.Ed.2d 580 (1984); Hines v. Davidowitz, 312 U.S. 52, 67-68, 61 S.Ct. 399, 404, 85 L.Ed. 581 (1941); Jackson v. S.P. Leasing Corp., 774 S.W.2d 673, 678 (Tex.App. — Texarkana 1989, writ denied). Lilly and the FDA urge that just such a conflict is presented here: by compelling disclosure of what the regulations promise will remain confidential, the trial court's order stands as an obstacle to the effective operation of the FDA's reporting system, and they argue, will effectively destroy it.

Although no court has expressly determined that the FDA regulations "preempt" disclosure, the courts of four other states have had occasion to consider the discoverability of reporters' identities. In Newsom v. Breon Laboratories Inc., the Supreme Court of Tennessee held that in permitting disclosure of the identities of reporters, the lower courts erred in not considering the burdens on the parties or the reporters' expectations of confidence. 709 S.W.2d 559, 560 (Tenn. 1986). The court determined that disclosure of the names and addresses of twelve out of approximately 400 reporters would be "sufficient to satisfy plaintiffs' interests." Id. Citing Newsom, the Louisiana Supreme Court ordered a trial court to amend its protective order so as to delete reporters' identities, but "[reserv[ed] to plaintiffs the right to apply for disclosure in a particular case upon showing of relevance." Wesley v. Rye, 490 So.2d 272 (La. 1986). Striking a similar balance between the parties' burdens and needs, a New York appellate court also ordered redaction of reporters' identities. Stahl v. Rhee, 136 A.D.2d 539, 523 N.Y.S.2d 159, 160 (1988). Noting the public policy reasons in support of redaction, the court found that "at least at this juncture, . . . the identities of the reporting sources are not material and necessary to the prosecution of the plaintiffs' case." Id. Lilly has also successfully obtained protective orders masking reporters' identities in trial courts in Kentucky and Texas. Fentress v. Shea Communications, No. 90-CI-06033 (Jefferson Cir.Ct., Ky. March 29, 1991); Morris v. Eli Lilly Co., Inc., No. 240, 313-401 (Probate Court No. 2, Harris County, Tex. Sept. 4, 1991).

In the pending federal multi-district litigation concerning Prozac, the Southern District of Indiana recently reviewed these same contentions. After balancing the competing interests of and burdens on the parties, pursuant to Federal Rule of Civil Procedure 26(c), the court ordered the redaction of the reporters' identities, reserving to the plaintiffs the opportunity to request the identity of a particular reporter based on a showing of "relevance, necessity to resolve disputed facts and that the information [would not be] otherwise available." In re Eli Lilly Co., Prozac Prods. Liab. Litigation, 142 F.R.D. 454, 459 (S.D.Ind. 1992).

See also Langer v. Dista Prods. Co., No. 90C-4598, 1991 WL 349606 (N.D.Ill. Nov. 12, 1991); Harris v. The Upjohn Co., 115 F.R.D. 191 (S.D.Ill. 1987).


For analysis of confidentiality in slightly different contexts, see Deitchman v. E.R. Squibb Sons, Inc., 740 F.2d 556, 565 (7th Cir. 1984) (district court directed to permit some discovery of reports in university registry of injuries caused by certain drug, but cautioned that identity of those submitting information to the registry need not necessarily be revealed); Farnsworth v. The Procter Gamble Co., 101 F.R.D. 355, 357 (N.D.Ga. 1984), aff'd, 758 F.2d 1545 (11th Cir. 1985) (confidentiality interest of patients and doctors providing information to the Center for Disease Control outweighed manufacturer's need for identities).

The plaintiffs in In re Lilly supported their arguments with the trial court's order in this case and that of a federal court compelling disclosure in Durham v. Hoffman-LaRoche, involving the drug Accutane. No. CV 89-L-0075-S (N.D.Ala. 1989). In Durham, the trial court ordered the defendants to produce the names and addresses of reporting physicians. The defendants' petition for writ of mandamus to the Eleventh Circuit was denied by a divided panel; their petition for writ of certiorari to the United States Supreme Court was also denied. 498 U.S. 890, 111 S.Ct. 232, 112 L.Ed.2d 192 (1990). As noted by the In re Lilly court, the Durham order contains no statement of facts or reasons, and we note that the failure of the Supreme Court to grant review should not be viewed as a pronouncement on the merits. E.g., Hopfmann v. Connolly, 471 U.S. 459, 461, 105 S.Ct. 2106, 2107, 85 L.Ed.2d 469 (1985).

Judge Clark would have granted the petition to order the district court to keep confidential the reporters' identities except as necessary to resolve disputed fact issues when the evidence would not be available from another source. In re Hoffman-LaRoche, No. 89-7896 (11th Cir. Oct. 9, 1990, Clark, J., dissenting).

To effectuate the truth-finding function of the legal system, discovery is not limited to what may be admissible at trial, but includes any information relevant to the pending subject matter that is reasonably calculated to lead to the discovery of admissible evidence. TEX.R.CIV.P. 166b(2)(a). Moreover, under the doctrine of shared discovery, the fruits of discovery are available not only to the parties in a particular case but may be disseminated in turn to other litigants and potential litigants. Garcia v. Peeples, 734 S.W.2d 343, 347 (Tex. 1987). The broad scope of discovery may be circumscribed, however, by the legitimate interest of the opposing party in avoiding discovery based on a compelling, particularized interest in nondisclosure. See Jampole v. Touchy, 673 S.W.2d 569, 573 (Tex. 1984); TEX.R.CIV.P. 166b(3), 166b(5). In this case, although we are not persuaded that the FDA regulations preempt the trial court's order, we nevertheless conclude that disclosure of otherwise discoverable information is circumscribed by the compelling public interest considerations manifested by those regulations.

We assume, for the sake of argument, that the Biffles would be able to persuade some doctors to in turn persuade their patients to waive the physician-patient privilege.

The FDA regulations clearly embody a vital public interest in confidential voluntary reporting that is eviscerated as equally by a manufacturer's compelled disclosure as by the FDA's disclosure. While Lilly claims no privilege per se in maintaining reporter confidentiality, we do not doubt its protectible economic interest — in addition to the public interest, as asserted here by the FDA — in maintaining the free flow of information derived from adverse reaction reports. Consequently, we agree that the congressional objective of fostering post-approval reporting of possible adverse reactions for all FDA-approved drugs is severely compromised by the trial court's order of wholesale disclosure of reporters' identities. While the need for confidentiality as determined by the FDA, and as promised on form 1639 and expressed in the regulations, may yield to a proper court order under 21 C.F.R. Sec. 20.83(a), here the trial court ordered full disclosure without according any weight to the public interest in the current voluntary reporting system. By ordering disclosure without a showing of particularized relevance and need by the Biffles, the trial court failed to apply the correct legal standard for determining if this confidential information should be released.

The Biffles are clearly entitled to all the substantive information on the reports and to share that discovery with their expert witnesses and litigants in other cases. But while we agree that the FDA regulations do not preempt Texas law of tort or discovery, to the extent that Lilly has been ordered to act in a manner inconsistent with the public interest concerns manifested by federal law, and without due consideration having been given to those concerns, that order is erroneous as a matter of law.

In compliance with the Freedom of Information Act, 5 U.S.C. § 552, after a drug is approved, adverse reaction reports are "immediately available for public disclosure" following redaction of patient and reporter identities. 21 C.F.R. § 314.430(e). Lilly has already produced some of the reports. We expect it to produce the rest of the reports promptly in compliance with the modified order which we anticipate Respondent will render.

Accordingly, we hold that the trial court abused its discretion by directing disclosure of the reporters' identities without a showing of particularized relevance and need, in contravention of important congressional objectives. We therefore conditionally grant the writ and direct Judge Marshall to modify his order in accordance with this opinion. The writ will issue only should he fail to do so.

DOGGETT and GAMMAGE, JJ., note their dissent.


Once again the majority has intervened in an ongoing trial court proceeding to rewrite Texas law. This time the special treatment accorded in creating a previously unknown discovery privilege threatens the public health and safety by posing formidable obstacles to the search for truth in pharmaceutical and medical device litigation. I dissent.

I.

The events leading to this particular case began with the suicide of Michael Hays Biffle, six days after beginning his prescription of Prozac, an anti-depressant manufactured by Eli Lilly. Alleging that the drug was responsible for the death, his family sought production of premarketing documents, including all reports Lilly forwarded to the FDA containing descriptions by health care providers of their patients' adverse reactions to Prozac.

Lilly objected to discovery of this "confidential information," which included alleged trade secrets and the identities of patients who had experienced adverse reactions, as well as their reporting physicians. Alternatively, the company sought a protective order preventing any disclosure of this information to other litigants or the public. On October 17, 1991, Judge John McClellan Marshall held a hearing on both the Biffles' request for production and Lilly's motion for a protective order; arguments centered on trade secrets and the need to prohibit discovery of the identities of reporters of adverse reactions. On October 23 Marshall granted the plaintiffs' request in part and issued a limited protective order entitling Lilly to withhold from discovery technical materials concerning the manufacture of Prozac and to redact from adverse reaction reports the names of patients, but not reporters. This order did not limit public disclosure of the documents.

On December 2, 1991, Lilly sought mandamus relief to prevent public disclosure, and the majority immediately granted an emergency stay. See Eli Lilly Co. v. Marshall, 829 S.W.2d 156 (Tex. 1991) (Doggett, J., dissenting to order granting leave to file petition for writ of mandamus). This court later ordered that Lilly's trade secret claim be considered in a hearing on sealing under Tex.R.Civ.P. 76a. See Eli Lilly Co. v. Marshall, 829 S.W.2d 157, 158 (Tex. 1992, orig. proceeding).

Although expressing concern in this court only with the question of public disclosure, Lilly continued its efforts in the trial court to prohibit any discovery of adverse reaction reporters' identities. Under the October 23 order, Lilly had represented that all documents ordered produced would be made available at its headquarters in Indianapolis. When the Biffles' counsel arrived at the scheduled time, however, Lilly refused to provide many of the documents and redacted reporters' names from those which were produced. Following a hearing, the trial court in January 1992 required Lilly to pay the Biffles' future costs in obtaining those materials as sanctions for defiance of the prior order.

On January 22, Lilly returned to Austin, this time complaining of the trial court's abuse of discretion in ordering the discovery of reporters' identities. This second mandamus request came despite Lilly's earlier representation to this court that its only complaint as to the October 23 order was "on the dissemination question." Motion for Temporary Relief, Petition for Writ of Mandamus, and Brief in Support at 9. The following day John Luke Hill, the former Chief Justice of this court and now counsel for Eli Lilly Co., filed here a letter of concern he had just personally received from an employee of the Food and Drug Administration (FDA). Within a few hours, the majority granted emergency relief. When after another hearing in which the FDA participated, the trial court for the third time rejected the reporter confidentiality arguments, the majority granted leave to file this second petition for writ of mandamus, described by Lilly's counsel at oral argument as a "pure [federal] preemption case."

Lilly's claim that FDA rules prevent a state court from authorizing access to reporter identities conflicts with the explicit language of an applicable regulation never cited by Lilly in any of its briefing:

Records of the Food and Drug Administration which the Commissioner has determined are not available for public disclosure . . . shall nevertheless be made available for public disclosure in compliance with a final court order requiring such disclosure.

21 C.F.R. Sec. 20.83(a) (1991). The FDA clearly anticipated production of adverse reaction reports pursuant to an order like that issued by Judge Marshall. Understandably, no published appellate court opinion anywhere has ever agreed with the type of preemption argument advanced by Lilly.

Despite its unambiguous language, counsel for Lilly responded at oral argument that this regulation "has nothing to do with the issues before this court."

Reluctantly accepting the reality that preemption does not apply here, the majority nonetheless pontificates about this subject at length without noting that

[c]onsideration of [preemption] . . . starts with the assumption that the historic police powers of the States [are] not to be superseded by . . . Federal Act unless that [is] the clear and manifest purpose of Congress.
Cipollone v. Liggett Group, Inc., 505 U.S. ____, ____, 112 S.Ct. 2608, 2617, 120 L.Ed.2d 407, 422 (1992) (citations omitted). This presumption against preemption is even more difficult to overcome in the specific context of administrative regulations:

[B]ecause agencies normally address problems in a detailed manner and can speak through a variety of means, including regulations, preambles, interpretive statements, and responses to comments, we can expect that they will make their intentions clear if they intend for their regulations to be exclusive.
Hillsborough County v. Automated Medical Laboratories, Inc., 471 U.S. 707, 718, 105 S.Ct. 2371, 2377, 85 L.Ed.2d 714 (1985); see also California Coastal Comm'n v. Granite Rock Co., 480 U.S. 572, 583, 107 S.Ct. 1419, 1426, 94 L.Ed.2d 577 (1987) ("it is appropriate to expect an administrative regulation to declare any intention to pre-empt state law with some specificity"). Hence, given the FDA's express disavowal in 21 C.F.R. Sec. 20.83 of the preemptive effect for confidentiality regulations in court proceedings, the majority was left with the task of finding some other basis for its preconceived result.

Despite my vigorous disagreement with this creation of a new privilege, I certainly do join in the majority's strong reaffirmation of "the doctrine of shared discovery" as recognized in Garcia v. Peeples, 734 S.W.2d 343, 347 (Tex. 1987) and the right of the Biffles "to share [any] discovery with their expert witnesses and litigants in other cases." 850 S.W.2d at 160.

II.

To meet that challenge, the majority simply creates a new exception to discoverability nowhere previously recognized in our state rules or statutes. Were there a true public need for such absolute reporter confidentiality, the FDA could initiate rulemaking to address it or certainly the drug manufacturers could voice their concerns to Congress or the Legislature. Rushing to legislate its own new basis for secrecy, the majority, however, pursues a course long ago abandoned by most courts by creating an entirely new common law privilege from discovery. As one eminent scholar has noted, "[t]he development of judge-made privileges halted a century ago." Charles McCormick, The Scope of Privilege in the Law of Evidence 16 Tex.L.Rev. 447, 469 (1938). Unlike the majority, which appears determined to reverse this longstanding trend, most jurisdictions exercise judicial restraint by interpreting statutes rather than enacting new privileges.

See State ex. rel. Chandra v. Sprinkle, 678 S.W.2d 804, 807 (Mo. 1984) (in bank); Sherman v. District Court, 637 P.2d 378, 384 (Colo. 1981) (en banc); Davison v. St. Paul Fire Marine Ins. Co., 75 Wis.2d 190, 248 N.W.2d 433, 441 (1977); Valley Bank of Nev. v. Superior Court of San Joaquin City, 15 Cal.3d 652, 125 Cal.Rptr. 553, 554-55, 542 P.2d 977, 978-79 (1975) (in bank); Nazareth Literary Benevolent Inst. v. Stephenson, 503 S.W.2d 177, 178-79 (Ky. 1973); Southern Bell Tel. Tel. Co. v. Beard, 597 So.2d 873, 876 n. 4 (Fla.Ct.App. 1992); Matter of Parkway Manor Healthcare Center, 448 N.W.2d 116, 120-21 (Minn.Ct.App. 1989); Scroggins v. Uniden Corp. of America, 506 N.E.2d 83, 85 (Ind.Ct.App. 1987).

Nor does this record show a sufficiently compelling basis to justify special treatment for Lilly. The sole evidence cited in support of this new privilege is the conclusory statement of a single individual, not even subject to cross-examination, that if reporters' identities were made public, "health care professionals and others would be much more reluctant to report adverse events for fear of involving themselves . . . in litigation." Majority opinion at 158 n. 7 (citing affidavit of Joyce Johnson at 4.) This statement in turn relies on a single study, which, in fact, concluded that only 11 percent of reporters of adverse reactions expressed concerns regarding the legal implications of their reports. See id. at 5; Julie Milstein, Gerald Faich, et al., Factors Affecting Physician Reporting of Adverse Drug Reactions, 20 Drug. Info. J. 157, 162 (1986).

Subsequent studies by one of the same authors, moreover, found that only 10 percent of all adverse reaction reports are sent directly to the FDA rather than to the drug manufacturer. Gerald Faich, et al., National Adverse Drug Reaction Surveillance, 257 J.Am.Med.Ass'n 2068 (1987). Since the FDA's confidentiality regulations do not prevent the drug maker from revealing the identity of reporters, the fact that 90 percent of the reports are sent to manufacturers indicates that health care providers are not particularly concerned that their identities be kept confidential. Additionally, almost half of the physicians surveyed were not even aware that the FDA has a system for reporting such reactions. Audrey Rogers, Gerald Faich, et al., Physician Knowledge, Attitudes, and Behavior Related to Reporting Adverse Drug Events, 148 Arch.Internal Med. 1596, 1599 (1988). Allowing discovery cannot deter reports from doctors unaware of the existence of the reporting system. Many health care providers are, however, keenly aware of the need to protect their patients' health — perhaps the real fear here is that some physicians will follow their ethical obligation to "make relevant information available to patients, colleagues, and the public. . . ." American Medical Association, Principles of Medical Ethics, reprinted in Law and Ethics in the Medical Office 126 (Marcia Lewis, ed., 1983).

As the majority acknowledges in rejecting the preemption argument, the FDA regulations do not themselves provide any "compelling public interest" justification, 850 S.W.2d at 160, for this newly established common law privilege. The FDA has never considered reporter confidentiality absolute. As the FDA Commissioner summarized public comment from health care providers, drug manufacturers, and others when the confidentiality rules were first proposed:

Comments pointed out that the Food and Drug Administration cannot guarantee confidentiality for any record, since a court may conclude that the information is subject to public disclosure.

39 Fed.Reg. 44,619 (1974). In adopting what is now 21 C.F.R. Sec. 20.83, the FDA reflected a public policy determination that any need for reporter confidentiality can be superseded by the public interest in seeking truth during the litigation process.

III.

This court's once strong commitment to open discovery is quickly being replaced with a new double standard of justice that promotes secrecy. In the past, this court had emphasized that discovery represents

See Axelson v. McIlhany, 798 S.W.2d 550, 553 (Tex. 1990, orig. proceeding) (abuse of discretion to deny discovery of potentially relevant documents without reviewing them in camera); Garcia v. Peeples, 734 S.W.2d 343, 347-48 (Tex. 1987, orig. proceeding) (abuse of discretion to grant a blanket protective order against sharing discovery with other litigants); Peeples v. Hon. Fourth Supreme Judicial Dist., 701 S.W.2d 635, 637 (Tex. 1985, orig. proceeding) (burden is on party asserting a privilege from discovery to produce evidence concerning its applicability); Jampole v. Touchy, 673 S.W.2d 569, 573 (Tex. 1984, orig. proceeding) (abuse of discretion to deny discovery of alternative designs of product). But see National Tank Co. v. Brotherton, 851 S.W.2d 193 (Tex. 1993, orig. proceeding) (restricting access to post-accident investigations).

the linchpin of the search for truth, as it makes a trial less of a game of blind man's bluff and more a fair contest with the issues and facts disclosed to the fullest practicable extent. In recent years, we have sought to secure this objective through both revision of the Texas Rules of Civil Procedure and our opinions discouraging gamesmanship and secrecy.
State v. Lowry, 802 S.W.2d 669, 671 (Tex. 1991, orig. proceeding) (citations omitted). The unwarranted invention of a new discovery privilege here serves only to loosen the linchpin, and cast greater doubt into the search for truth.

In the past, the party with something to hide bore the burden of justifying a restriction on discovery. See Barnes v. Whittington, 751 S.W.2d 493, 494 (Tex. 1988, orig. proceeding) ("a privilege must be established to justify an exception to the general rule favoring discovery"); Peeples v. Honorable Fourth Supreme Judicial Dist., 701 S.W.2d 635, 637 (Tex. 1985, orig. proceeding) (party seeking to exclude matters from discovery must affirmatively plead and prove a particular privilege); Tex.R.Civ.P. 166b(3)(e) (exempting from discovery matters protected by existing privileges). But here, the majority cites no applicable statutory provision, or evidentiary or procedural rule that entitles Lilly to resist production of the identities of reporters of adverse drug reactions. In the past, the party seeking a protective order was required to show a "particular, articulated and demonstrable injury," see Masinga v. Whittington, 792 S.W.2d 940, 940-41 (Tex. 1990, orig. proceeding); Garcia v. Peeples, 734 S.W.2d 343, 345 (Tex. 1987, orig. proceeding). But for at least one drug maker, the majority alters all of that well-established law. Despite our prior refusal to shield discovery based on "conclusory allegations" of harm, see Masinga, 792 S.W.2d at 941; Garcia, 734 S.W.2d at 345, this manufacturer's unsubstantiated global claims are accepted without question. The burden of proof is then reversed by requiring the Biffles to show "particularized relevance and need" for the reporters' identities. 850 S.W.2d at 160.

While reciting the requirement that mandamus will issue "only to correct a clear abuse of discretion," 850 S.W.2d at 157, in the course of staying four trial court orders and twice issuing the "extraordinary" remedy of mandamus in a single lawsuit, the majority has effectively eliminated any exercise of discretion regarding the proper scope of discovery. Despite having conducted three separate hearings addressing the majority's concerns, Judge Marshall has been accorded no reasonable latitude to assess the validity of Lilly's claims.

The public interest in health and safety — the purported basis for the majority's action — has in fact been jeopardized by its writing. Manufacturers of drugs and medical devices are now presumptively free to conceal the identities of those who complain of the potential life-threatening qualities of their products.

GAMMAGE, J., joins in this opinion.