Opinion
MDL Docket No. 4:03CV1507 WRW.
November 12, 2004
Lyn P. Pruitt, MITCHELL, WILLIAMS, SELIG, GATES WOODYARD, P.L.L.C., Little Rock, AR, John W. Vardaman, Stephen L. Urbanczyk, F. Lane Heard III, WILLIAMS CONNOLLY LLP, Washington, D.C., Attorneys for Wyeth.
Introduction
The Women's Health Initiative clinical trial ("WHI") is the cornerstone of plaintiffs' case. It is fair to say that, absent the WHI report in July 2002 and the early termination of the trial, the parties would not be here. Scores of epidemiological studies had been done on hormone therapy drugs in the preceding years, and the labeling for Prempro, Premarin and other hormone therapy drugs accurately reflected the reasonably consistent findings of those studies. But when the WHI investigators reported their findings relating to breast cancer and cardiovascular disease, the race to the courthouse began. The story that the complaints tell — and the story that Lead Counsel told in its presentation on the scientific issues at the April hearing — is that the findings of the decades' worth of epidemiological studies and other data that existed prior to WHI should be tossed out the window, and substituted for the most part with the results of this single study.
Plaintiffs are happy to take the WHI report at face value, and they have already encouraged the Court, in effect, to require defendants to stipulate to the truth of the report. The nature of scientific inquiry, however, is to question, analyze, and attempt to verify. And, of course, that is also the nature of the adversarial process in litigation. Just as other scientists will in due course take the WHI data, evaluate it, and consider whether it supports the conclusions being drawn by the WHI investigators, so Wyeth and the other defendants are entitled to evaluate the data and "cross examine" it.
See Memorandum in Opposition to Defendant Wyeth's Motion For Leave to Conduct Limited Discovery [Docket No. 40] (opposing Wyeth subpoena because "Plaintiffs can only presume that Wyeth seeks the raw WHI data in order to re-analyze and manipulate it so that it can challenge and undermine the previously published findings. . . .").
The courts have so held. In Deitchman v. E.R. Squibb Sons, Inc., 740 F.2d 556 (7th Cir. 1984), as here, a major scientific study was the springboard for products liability litigation. When the defendant sought the underlying data from the researchers in order to evaluate whether the data truly supported the conclusions being drawn in the litigation, the researchers objected. Id. at 558. The Deitchman court gave short shrift to the objections and ordered disclosure, explaining:
The value of the conclusions turns on the quality of the data and the methods used by the researcher. . . . So if the conclusions or end product of a research effort is to be fairly tested, the underlying data must be available to others equally skilled and perceptive. In the eyes of the law, there is no such thing as an infallible witness.Id. at 562 (emphasis added); see also Burka v. U.S. Dep't of Health Human Servs., 87 F.3d 508, 521 (D.C. Cir. 1996).
The importance of "cross examining" the data has also been demonstrated in In re PPA Products Liability Litigation, MDL-1407, where, as with hormone therapy, the litigation got its impetus from a single article — in the case of PPA, a New England Journal of Medicine article that reported for the first time in 2000 a causal link between the use of PPA and hemorrhagic stroke. In response to a subpoena duces tecum, Yale University (where the research had been conducted) produced the underlying data and related documentation. Only because defendants were able to look behind the reported conclusions to the data where they able to learn that:
There was agreement that Yale would redact five items of information in order to protect the identities of the study participants: (1) names; (2) street addresses; (3) last four digits of telephone numbers; (4) the day of birth (not month and year); and (5) physician names involved in care of the study participants. See pp, 12, infra, regarding the longer list of so-called identifiers that NIH and FHCRC insist upon redacting.
• a number of the stroke subjects were cocaine abusers, a far more compelling explanation for their strokes than use of cough/cold remedies;
• when the data was broken down by type of stroke, and one considered only those types of strokes for which there was a biologically plausible link between PPA use and stroke, the results were not statistically significant;
• the researchers knowingly, and inexplicably, presented an odds ratio that was a textbook example of residual confounding (and that was thus invalid);
• halfway through the study, the researchers changed the definition for "focal time" (when the subject's stroke occurred) and therefore collected data inconsistently;
• the researchers initially determined to report the results for subarachnoid hemorrhage and intracerebral hemorrhage separately, as they are distinct diseases, but when the results proved statisitically insignificant, they lumped the results together in order to report an odds ratio that was statistically significant; and
• the researchers manipulated the data to record a stroke as occurring after PPA use when contemporaneous medical records documented the first signs of stroke before PPA use.
In sum, only because the defendants were able to evaluate the data were they able to determine that there were significant issues with the study's integrity.
Judge Anthony Mohr of the California Superior Court, speaking at a March 18, 2004 hearing, said: "I mean, you could almost say that there was some unethical activity with that Yale Study."
Needless to say, scientists (and litigants) will differ about how the data should be "cut and sliced." But the adversarial system is grounded in the belief that judges and juries reach the truth through exposure to competing perspectives based "cutting and slicing" data differently.
For the moment, the public has only the perspective provided by the WHI researchers themselves (and what little evaluation can be made based on what data has been reported by them in their articles). Even the limited evaluation of the WHI study that has been possible without access to the underlying data, however, demonstates how significant can be "cutting and slicing" the same data in a new way. As was reported last month:
Menopause hormones may not be as risky as once thought for the vast majority of women who take them, according to a new analysis of two major hormone studies.
The analysis . . . strongly supports the notion that the age a woman starts taking hormones has a big impact on her risk for hormone-related heart problems. Specifically, younger women who take hormones for menopause symptoms appear to be at far lower cardiac risk and may even gain some benefit, according to the new review of the two largest hormone studies, the Women's Health Initiative and the Nurses Health Study. Most women who take hormones today fall into this category.
. . . Although many doctors have speculated that the [WHI] data may not apply to the typical 50-year old patient with menopausal sumptoms, the report presented at the N[orth] A[merican] M[enopause] S[ociety] meeting is the first detailed scientific analysis to back up that theory.
"Rethinking Hormones Again: Heart Risk May Be Lower in Women Who Start Early," Wall Street Journal, Oct. 12, 2004 (emphasis added).
For these reasons, the WHI data is important to Wyeth's defense of the litigation. Wyeth has sought the data from both the National Institutes of Health ("NIH"), which sponsored the study, and the Fred Hutchison Cancer Research Center ("FHCRC"), which conducted it. Each has responded separately to their identical subpoenas duces tecum.
BACKGROUND
Over plaintiffs' opposition, Wyeth sought leave to subpoena NIH for the data underlying the WHI report and related material. The Court granted leave, and in December 2003 Wyeth served identical subpoenas on six NIH and/or HHS administrators. The subpoenas called for nine categories of documents, but this motion to compel relates only to the first and most important category, which concerns the data underlying the WHI report:
1. All electronic data and programming code (including but not limited to all Statistical Analysis Software (SAS) data sets, instruction files, and transport files) from the Women's Health Institutive clinical trial of combined postmenopausal hormones (the combined estrogen plus progestin, or PERT, trial — (preliminary results of which were published in the Journal of the American Medical Association in July 2002, volume 288 at pages 321-33), including but not limited to all electronic data and programming code that can be used to analyze, manipulate and/or merge the data. This request includes all databases from which the SAS data sets were derived, and includes the aggregate data as well as the data stratified by site.
Counsel for Wyeth and NIH have conferred several times by telephone and exchanged letters. NIH initially took the position that it has already produced the underlying data to Wyeth. Then NIH offered to negotiate over how much data it would produce, providing Wyeth a spreadsheet of of data variables and asking Wyeth to consider whether its interest was limited to some subset of the data. Then, when Wyeth reiterated that it needed all the data, NIH said that it did not have possession, custody or control of any of the data — only the clinical investigators at various study sites around the country did.
Because Wyeth donated the Prempro that was used in the WHI study, it had an agreement with NIH that the company would be provided with the data resulting from the study. NIH has interpreted that agreement, however, as obligating it to give Wyeth only the limited data supporting the tables and charts published by the investigators. Putting aside whether that interpretation of the agreement is the right one, it is akin to a prosecutor disclosing the material that supports conviction, but not Brady material. Because of NIH's narrow interpretation of its obligation to Wyeth, and because the litigation heightens the importance of evaluating the data, Wyeth subpoenaed all the data.
Although dubious, Wyeth accepted that representation. It seemed implausible that NIH did not have control of the data for a study that it sponsored and paid for, and which is everywhere touted as an "NIH study." And it seemed inconsistent for NIH to contend that it did not have control of the data after initially responding to the subpoena by saying, "We gave you the data already." Still, because we had no legal basis to challenge NIH's claim — and, frankly, because it was "the government" telling us — we turned our attention to serving subpoenas on the Fred Hutchinson Cancer Research Center, the principal investigator and the entity which contracted with NIH to perform the study. FHCRC almost immediately produced its contract with NIH for the WHI study. The contract requires FHCRC to provide NIH — as a "deliverable" under the contract — such "data and documentation" as the NIH contracting officer may request. In light of that provision, Wyeth asked NIH counsel to explain:
See Letter to Paul J. Robertson from F. Lane Heard III, May 25, 2004 (attached as Exhibit 2):
As your initial response regarding Request No. 1 was that your client had provided information to Wyeth already, we find it hard to understand the subsequent response that your clients do not have any of the data. We also find it difficult to believe that this information regarding WHI (an NIH -sponsored study) is not in the possession, custody and control of the Director or Administrative Officer of WHI, nor the Secretary of HHS or the Director of NIH, [but] we will proceed on that basis for now unless you advise me that I have misunderstood what the respondents have and don't have.
Wyeth has served subpoenas on persons associated with FHCRC in Seattle, Washington. FHCRC conducted the study under a contract from NIH.
We have understood you to take the position until now that you cannot supply the data called for by paragraph 1 of the Attachment [to the subpoena], because that data is not in your possession or control. Please explain how that position can be reconciled with Article F-1(1)(W) [of the contract]..
Letter to Paul J. Robertson from F. Lane Heard III, August 11, 2004 (attached as Exhibit 3) (emphasis added).
NIH never responded — a telling commentary on its position that it does not have control of the data.
It is also telling that counsel for FHCRC indicated that the first thing she did when served with Wyeth's subpoenas was to get in touch with NIH and ask what to do.
Wyeth therefore moves to compel NIH to produce the data. NIH has never objected to the subpoenas, except to say initially that the requests were duplicative and possibly burdensome — objections rendered moot by NIH's subsequent claim that it does not have control of the data. The sole issue, then, is whether NIH has legal control of the data. If it does, the data should be produced forthwith, subject to a confidentiality order that bars Wyeth and other recipients of the data from attempting to identify the study participants.
Wyeth is also moving to compel FHCRC, which was served with identical subpoenas on July 7, 2004. FHCRC has possession of the data underlying the WHI study and is prepared to produce it, but only subject to redaction of a long list of so-called "participant identifiers" that would both substantially delay the production and impair Wyeth's ability to analyze the data — redactions, moreover, that are unnecessary because Wyeth has proposed a confidentiality order that would prohibit any person to whom the data is disclosed from attempting to identify the study participants. Because FHCRC takes the position that it cannot compromise about the redactions absent the agreement of NIH — indeed, says that NIH requires it to make the redactions — it is necessary to bring NIH "to the table" and establish its control over the data.
Wyeth served subpoenas on the following four individuals: FHCRC Custodian of Records; FHCRC Custodian of Records for the WHI Clinical Coordinating Center; Mary Pettinger; and Garnet L. Anderson, Ph.D. The subpoenas are attached as Exhibit 4.
ARGUMENT
A. This Court Has the Authority to Enforce the Subpoenas
Although the subpoenas to NIH were issued out of the District Court for the District of Columbia, they are enforceable by this Court, as an MDL court. See In re Subpoena served on Wilmer, Cutler Pickering, 255 F.Supp.2d 1, 2-3 (D.D.C. 2003); see also In re Subpoena Issued to Boies, Schiller Flexner LLP, No. M8-85, 2003 WL 1831426 (S.D.N.Y. Apr. 3, 2003) ( 28 U.S.C. § 1407 gives MDL transferee judge the power to hear motion regarding subpoena issued by another court); In re Diet Drugs Phentermine/Fenfluramine/Dexfenfluramine Prods. Liab. Litig., 2004 WL 838074 (D. Kan. March 24, 2004) (MDL judge in Eastern District of Pennsylvania "has all the powers of a federal district judge in the District of Kansas for the purpose of conducting depositions in this multidistrict litigation").
It is noteworthy that the District Court for the District of Columbia, which is the alternative venue for seeking enforcement of these subpoenas to NIH, has recognized and defended the rule that the MDL court has the power to do so. See United States ex rel. Pogue v. Diabetes Treatment Ctrs. of Am., Inc., 238 F.Supp.2d 270, 273, 275 (D.D.C. 2002) ("the weight of authority and effectuation of the purposes of multi-district litigation support a finding of jurisdiction" in the MDL court to enforce subpoenas issued elsewhere) (citing 28 U.S.C. § 1407(b); Manual for Complex Litigation (Third) § 21.424 (2002)).
B. NIH Has Legal Control of the WHI Data
1. The test is control, not possession
It does not matter that NIH does not have physicial possession of the WHI data. Where discovery under the Federal Rules of Civil Procedure is concerned, the test is not physical possession of the documents or data being requested, but control. The question that the law puts to the recipient of a subpoena is not simply, "Do you have the material in your possession right now?", but, if you don't, "Do you have the right to take possession of the material?" See Marc Rich Co., A.G. v. United States, 707 F.2d 663, 667 (2d Cir. 1983) ("The test for production of documents is control, not location."), cert. denied, 463 U.S. 1215 (1983); Kissinger v. Reporters Comm. for Freedom of the Press, 445 U.S. 136, 165 n. 6 (1980) (Brennan, J. concurring in part and dissenting in part) ("In construing [Federal Rules of Civil Procedure 34 and 45] the courts have rejected a narrow physical-possession test, focusing instead on whether the subpoenaed party has a legal right to custody or control of the documents in question."); Fed.R.Civ.P. 45, advisory committee's note. Thus, a party that has a contractual right to documents has "control" of the documents for purposes of Rules 34 and 45. See Rosie D. v. Romney, 256 F. Supp. 2d 115, 119 (D. Mass. 2003) ( citing Anderson v. Cryovac, Inc., 862 F.2d 910, 928-29 (1st Cir. 1988), on remand sub nom. Anderson v. Beatrice Foods Co., 127 F.R.D. 1, 10 (D. Mass. 1989).
Treatises on federal practice also point to control, rather than location. See 9A Charles Alan Wright Arthur R. Miller, Federal Practice and Procedure § 2456, at 31 (1995 2003 Supp.) ("Even records kept beyond the territorial jurisdiction of the district court issuing the subpoena may be covered if they are controlled by someone subject to the court's jurisdiction.").
In Rosie D. v. Romney, the plaintiffs sought production of documents containing personal data concerning clients of various agencies, including the Executive Office of Health and Human Services. 256 F. Supp. 2d at 117. The state Division of Medical Assistance did not have possession of the data, but a contract between the Division and a partnership of entities charged with delivering behavioral health services provided that the Division had the right to examine and copy the data. Id. at 119. Consequently, the court ordered production of the documents, holding that "[g]iven these contractual provisions, there is little doubt that Defendants have the right to control and obtain the documents that are in the possession of the various non-defendant agencies." Id.
2. NIH has control under the NIH-FHCRC contract
NIH's position is that it does not have possession or control of the WHI data. Wyeth accepts that NIH does not have physical possession of the data, but contends that NIH does in fact have control of the data — for three reasons.
First, the contract for the WHI study — Contract No. 1-WH-2-2110 between the Research Contracts Branch, National Institutes of Health, and FHCRC ("WHI contract") — expressly provides that FHCRC must deliver the data to NIH. Article F-1 of the WHI contract describes the deliverables necessary to accomplish "[s]atisfactory performance of this contract." The final item to be delivered is the "Data and Documentation" for the clinical trial. When Wyeth received the contract (from FHCRC, not NIH), we inquired how it was that NIH could take the position that the study data was not in its control:
The FHCRC served as the clinical coordinating center for the WHI study.
See Article F-1(1)(W) of the WHI contract (relevant portion attached as Exhibit 5).
We have understood you to take the position until now that you cannot supply the data called for by paragraph 1 of the Attachment [to the subpoena], because that data is not in your possession or control. Please explain how that position can be reconciled with Article F-1(1)(W) [of the contract].
Letter to Paul J. Robertson from F. Lane Heard III, August 11, 2004 (attached as Exhibit 3).
NIH did not respond.
Second, counsel for FHCRC has told us that NIH controls the data. When Wyeth subpoenaed the data from FHCRC, the first thing its counsel did, she said, was to seek instructions from NIH on how to respond. As Wyeth and FHCRC have conferred concerning the terms of a protective order to govern the data, FHCRC has stated that it needs "sign off" from NIH. And on the point of dispute between the parties (a question of what information needs to be redacted to protect the identity of study participants), FHCRC says that it is bound by what NIH has told it to do.
Third, NIH's current position that it does not have control of the WHI data is inconsistent with its own statements and actions. When NIH and Wyeth counsel first conferred about the subpoenas, NIH counsel said with regard to Request No. 1 that NIH had already given Wyeth the data. (Wyeth acknowledged that NIH had given it some of the data; the purpose of the subpoenas was to obtain a complete set of the data.) In an effort to narrow the subpoena, NIH counsel then gave Wyeth counsel a spreadsheet listing all the variables for which WHI had collected data and asked Wyeth to review with an eye to limiting its request to the data for a subset of variables. It was only when Wyeth rejected that proposal and asked again for all the data that NIH claimed it did not have possession or control of any of the data. In asserting that it had produced all (or part) of the data already, in seeking to negotiate a production of part of the data (based on Wyeth's selection of certain data variables), and in giving directions to FHCRC, NIH has acted in a manner completely inconsistent with the claim — implausible enough on its face, given that NIH sponsored the study — that NIH does not have control of the data.
The spreadsheet is attached as Exhibit 6.
If NIH has control of the data, then it must produce the data forthwith, because the only objections that it ever made to the subpoena are makeweight. NIH objected that Wyeth's request for the underlying data is duplicative, but NIH concedes that it has not provided the data for hundreds of variables. NIH also objected that the request was burdensome, but all that is required of NIH is agreement to permit the data to be disclosed subject to an appropriate confidentiality order. Wyeth has agreed to pay the reasonable costs associated with FHCRC's actual production of the data.
We are at a loss to understand why NIH is playing this shell game, but it should end. NIH should not be permitted to say, "Get the data from FHCRC" while FHCRC says (as we explain below), "We can only disclose the data that NIH tell us we can disclose."
C. The Court Can Protect the Privacy of Study Participants
1. "Participant identifiers"
FHCRC has the WHI data on its computer system and is prepared to produce the data subject to two conditions: (1) that Wyeth agrees to pay the cost of extracting the data from the system and producing in a way that makes it usable, and (2) that the Court enters a confidentiality order that authorizes FHCRC to redact all so-called "participant identifiers." Wyeth has agreed to the first condition, but seeks to limit the second. Wyeth agrees to the entry of a confidentiality order; indeed, Wyeth and FHCRC have agreed to all the terms of such an order, but one. This motion to compel concerns only that one term.
FHCRC and Wyeth are continuing to refine Request No. 1 in an effort to reduce the time and cost involved in extracting and producing the data.
Wyeth and FHCRC agree that FHCRC may redact that information that would truly identify a participant in the WHI study — i.e., her name, street address, telephone number, social security number, place of birth, city of birth, and contact data. Wyeth and FHCRC part company at that point, because FHCRC also insists upon redacting as "participant identifiers" the following items of information: (1) clinical center identifiers; (2) sensitive data (drug, alcohol, HIV/AIDS, STDs, violent behavior, sexual behaviors); (3) genetic markers of participants; (4) dates of disease events; and (5) assignment to the dietary modification calcium trials. Wyeth objects to the redaction of this information for three reasons.
First, the information is important to analyzing the data and determining whether it supports the conclusions drawn by the researchers and plaintiffs:
In the case of "clinical center identifiers," this information is necessary to detect trends, anomalies or biases at particular centers that may affect the results. For example, if a particular center has a disproportionate number of breast cancer diagnoses, a question arises about the methods of diagnosis, or the diagnosticians, that may be unrelated to the therapy. Or there may be a higher rate of errors from that center, because of recording or other quality assurance problems.
In the case of "sensitive data," this information represents potentially confounding factors. For example, alcohol use and/or abuse has been associated with a number of outcomes in the WHI study — i.e., including breast cancer, stroke and cardiovascular disease. Similarly, illegal drug use of various kinds can contribute to cardiovascular problems, including stroke.
In the case of "genetic markers," this information, too, represents potential confounding factors. For example, if many (or most) study participants who had venous thromboembolisms also had a certain genetic mutation (Factor V Leiden), which predisposes the carrier to clotting problems, the mutation (and not hormone therapy) could explain the outcome.
In the case of "dates," FHCRC plans to recode all calendar dates to dates relative to a reference point (e.g., time of randomization). That is, instead of reporting that a participant suffered a stroke on June 10, 2000, FHCRC would recode this date and report it as 750 days since randomization. It is important, however, to receive actual dates for two reasons: (1) to determine if a global event at the study site introduced bias; and (2) to compare the date of events to the dates of interim letters sent to the study participants (because the letters warned them of increased rates of various outcomes, such that any spike in those outcomes after the letters may reflect a bias introduced by the letters).
Finally, in the case of assignment to the "dietary modification clinical trial" and "Calcium and Vitamin D clinical trial," some hormone therapy study participants were also enrolled in a WHI studies of dietary modification and calcium use. Because dietary modification can affect the risk of heart disease, stroke and breast cancer (indeed, the study hypothesizes that diet affects breast cancer rates), assignment to that trial is a confounder; similarly, assignment to the Vitamin D trial is a confounder for osteoporotic results. In other words, if more of the placebo group in the hormone therapy trial were also assigned to the dietary modification trial — and complied with that assignment — that fact (and not hormone therapy) could explain any apparent lower risk of breast cancer or stroke.
Second, these so-called "participant identifiers" are no such thing. Knowing, for example, that the study was carried out, in part, at a center in Los Angeles does not serve to identify any participant by name. Nor does knowing that participant No. 10,562 had an abortion, nor that participant No. 4,877 abused cocaine. One would need the combined forces of the FBI and CIA (operating with the success attributed to them in the movies) to take this kind of information and identify a participant. The same is true for the other "participant identifiers" in question — genetic markers, date of events, and assignment to the dietary modificiation trial — which are a bureaucrat's fantasy of what information someone with unlimited desire and unlimited resources could possibly use to come up with the name of someone in the WHI study.
Third, redacting the entire list of "participant identifiers" will slow down FHCRC's production of the data and make the process more expensive. As FHCRC estimates that the process will take more than 5000 hours and cost more than $800,000, any savings of time and money is meaningful. The litigation will not stop to wait for FHCRC to finish, and the cost is enormous by any reckoning.
2. A court order suffices
There is a better way. The NIH guidelines for redaction have almost certainly been written to be over-inclusive, because neither NIH nor FHCRC have any effective way to enforce an agreement not to identify study participants when data is disclosed to other researchers. The form agreement for such disclosure toothlessly provides:
Failure to comply with this Distribution Agreement could result in denial of further access to ____ Study data. Violation of the confidentiality requirements of this agreement is considered a breach of confidentiality and may leave requesting investigators liable to legal action on the part of ____ Study participants, their families, or the U.S. Government.
See Distribution Agreement (attached as Exhibit 7).
But where the disclosure of study data is made in response to a subpoena duces tecum, the Court can enter a confidentiality order with teeth. The Court can order anyone who receives the data pursuant to the subpoena not to attempt to identify the study participants upon pain of contempt of court. Such a provision — proposed by Wyeth and rejected by FHCRC (under direction from NIH) — protects the privacy of the participants and the ability of Wyeth (and others) to "cross examine" the data.
See proposed Confidentiality Order (attached as Exhibit 8).
This approach is consistent with the Privacy Act, which provides for the disclosure of protected information in response to discovery requests. It is also consistent with the Public Health Act, which prohibits the disclosure of the "identifying characteristics" of study participants. To begin with, Wyeth has agreed to redaction of true "identifying characteristics" — name, address, social security number, etc. In addition, the one reported opinion discussing the Pubic Health Act provision supports the approach Wyeth advocates here. In Murphy v. Philip Morris Inc., 1999 WL 33521196 (C.D. Cal. Dec. 28, 1999), a personal injury case alleging that the plaintiff contracted lung cancer from exposure to second hand tobacco smoke, the defendant subpoenaed the raw data underlying a study conducted by the University of Southern California on "Lung Cancer in Nonsmoking Women." The university objected on the ground that disclosure of the data would abrogate prohibitions against disclosure of the identities of study participants.
5 U.S.C. § 552a.
See Laxalt v. McClatchy, 809 F.2d 885, 889 (D.C. Cir. 1986); see also Mary Imogene Bassett Hosp., 136 F.R.D. 42, 49 (N.D.N.Y. 1991) (granting motion to compel disclosure of Medicare database of patient records from Secretary of Health and Human Services subject to appropriate protective order).
The court's holding had three important elements. First, the court held that Philip Morris' "interest in disclosure of the redacted data raw data outweighs the interest of the public in non-disclosure of that redacted data." Second, the court held the public's interest in non-disclosure of the data is actually limited only to non-disclosure of the actual identities of the study participants. . . ." Third, the court found that "the data itself without further inquiry does not automatically identify any participant;" that one could not ascertain the identity of any participant "without using the [raw data] to research other databases;" and that "Philip Morris has repeatedly assured this court that it has no intention of using the raw data to identify any of the study participants."
1999 WL 33521196, at *4. The court explained that "due process specifically and justice generally require that Philip Morris' ability and right to defend itself against the Murphy allegations . . . not be compromised; in contrast, the identities of the individuals whose personal stories have contributed to the collection of information on this subject is of little or no public concern." Id. at *3.
The court accordingly ordered the university to produce the raw data subject to a protective order that did two things: (1) permitted the redaction of actual identifying information (name, address, social security number, personal ID numbers, insurance policy numbers, etc.), and (2) prohibited any person having access to the data from "attempt[ing] in any way, directly or indirectly, to ascertain the identity of any such [USC] study participant." The court expressly took account of the Public Health Act and held that its protective order satisfied the statute.
Id. at *4.
These same elements are all present here. Wyeth has the same compelling need to defend itself by examining the data underlying the WHI study, upon which plaintiffs so heavily rely. Wyeth agrees to the redaction of information that reveals the actual identities of the study participants. And Wyeth also agrees to entry of an order prohibiting any attempt to identify the participants. Where this is true, there is no legal impediment to disclosure of the data.
CONCLUSION
For these reasons, the Court should order FHCRC and NIH to produce the data underlying the WHI study subject to the attached confidentiality order.
ORDER
Upon consideration of Wyeth's Motion to Compel Production of Documents From the National Institutes of Health and the Fred Hutchison Cancer Research Center, filed November 5, 2004, and the response thereto, it is ORDERED that the motion is GRANTED.
1. NIH and FHCRC will produce the documents responsive to Wyeth's Document Request No. 1 as expeditiously as possible.
2. Disclosure will be subject to the Confidentiality Order Concerning WHI Study Data, which is attached.
IT IS SO ORDERED.