Mount Sinai School of Medicine ("Mount Sinai") and The American Cancer Society ("ACS") appeal from a final order of the United States District Court for the Southern District of New York, Kevin Thomas Duffy, Judge, holding them in civil contempt for failing to comply with orders of the court requiring them to respond to subpoenas served by appellees The American Tobacco Company ("American"), R.J. Reynolds Tobacco Company ("Reynolds"), and Philip Morris, Inc. ("Philip Morris") (collectively the "tobacco companies"), requesting research data for use in lawsuits to which Mount Sinai and ACS are not parties. On appeal, Mount Sinai and ACS challenge interlocutory orders that (1) denied their motion to quash the subpoenas on grounds of res judicata, collateral estoppel, and privilege, and (2) denied in part their motion for a protective order permitting redaction of the subpoenaed materials to prevent identification of persons who were subjects of the research. For the reasons below, we uphold the orders of the district court.

Each of the tobacco companies is a defendant in one or more product liability suits pending in state or federal courts around the country (the "underlying suits"), in which the plaintiffs allege that their decedents died of lung cancer caused by a combination of cigarette smoking and exposure to asbestos. Mount Sinai and ACS are not parties to those suits, and neither they nor members of their medical staffs are expected to be called as witnesses. The tobacco companies expect, however, that the plaintiffs will rely on expert testimony that in turn will rely on seminal studies made by certain members of the medical staffs of Mount Sinai and ACS. The tobacco companies seek to subpoena the data underlying these studies.

Dr. Irving J. Selikoff, a professor of medicine at Mount Sinai, is the principal author of, inter alia, two articles whose medical conclusions support the plaintiffs' claims in the underlying suits. See Hammond, Selikoff & Seidman, Asbestos Exposure, Cigarette Smoking and Death Rates, 330 Annals N.Y. Acad. Sci. 473 (1979); Selikoff, Seidman & Hammond, Mortality Effects of Cigarette Smoking Among Amosite Asbestos Factory Workers, 65 J. Nat'l Cancer Inst. 507 (1980). Earlier studies had shown that cigarette smoking may cause cancer and that exposure to asbestos may also cause cancer. The Selikoff articles suggested that when cigarette smoking was combined with occupational exposure to asbestos, the risks of developing cancer increased geometrically rather than arithmetically. The authors concluded that these data suggested a synergistic relationship between the hazards of smoking and of exposure to asbestos.

For the 1979 article, the period of study ran from 1967 through 1976; for the 1980 article, the period of study ran from 1961 through 1977. Some 11,000 asbestos workers were subjects of the research reported in these two articles. Using a wide range of sources, Dr. Selikoff accumulated a variety of personal data on his subjects. Some information was received through the personal examination of approximately 500 subjects who were treated by Dr. Selikoff in connection with research for the two articles. Along with basic identifying details such as names, places of residence, and union affiliations, the assembled information included such items as dates of birth, dates and causes of death, periods of exposure to asbestos, smoking habits, evidence of cancer, and other relevant medical history. Dr. Selikoff assured his research subjects that the information they provided would remain confidential.

Much of the raw data was eventually recorded on computer tapes to facilitate statistical analysis and convenient storage. The Selikoff research continued after the publication of the 1979 and 1980 articles, and the computer tapes were updated frequently with new information.

In 1986, Reynolds served Mount Sinai and ACS with subpoenas issued by a state court in New York in connection with an action pending in state court in California, see Page v. Lincoln Electric Co., No. 257046 (Cal.Super.Ct.) (the "Page subpoenas"). The Page subpoenas sought all the data underlying the 1979 and 1980 Selikoff articles, as well as data underlying a 1968 article. Each subpoena provided, in pertinent part, as follows:

1. This subpoena covers documentation, data collections, or data bases (the "raw data") that formed the basis for the [1968, 1979, and 1980 articles];

2. Documents which describe, constitute, comment upon, criticize, review or concern the research design, methodology, sampling protocol, and/or conduct of any of the studies;

3. Copies of the questionnaires, answers to questionnaires, interview forms, responses to interviews, death certificates, autopsy reports, and other cause of death data....;

4. Data sheets, computer tapes and/or copies of computer discs containing all coded data....;

....

6. This request is intended to cover all available data used, in as "raw" a form as possible ....

(Emphasis added.)

Mount Sinai and ACS moved to quash the Page subpoenas in New York state court. That court, after discussing both the interests of scholars with respect to their research and the burdens of producing the data called for by the Page subpoenas, granted the motions to quash. In re R.J. Reynolds Tobacco Co., 136 Misc. 2d 282, 518 N.Y.S.2d 729 (Sup.Ct.1987) ("Reynolds "). The court stated, in pertinent part, as follows:

Reynolds' subpoenae ... are not selective. Rather, they are sweeping and indiscriminate.

A subpoena may be challenged on the grounds that it is overbroad, burdensome or oppressive.... A subpoena may be vacated for reasons of privilege, whether statutory or constitutional, or having its genesis under common law....

CPLR 3101 provides that "[t]here shall be a full disclosure of all evidence ...", which means all relevant information calculated to lead to relevant evidence.... The data requested must be material and necessary although it need not be indispensable.... The rules of evidence may be considered when determining necessity of the requested material. But inadmissibility by itself may not be a reason for denying access to the information, since such information may lead to evidence. In addition, its use in cross-examination can be a consideration.

... [T]he fact that the information is available from other sources or even that Reynolds has or could have obtained the material needed at trial is not reason alone for denying the production of the data. Material subpoenaed may be used for corroborative purposes.

However, when compliance with subpoenae would be so oppressive as to hinder the normal functioning of a Department of the medical school and/or of the American Cancer Society for a prolonged period of time, the Court on motion or on its own initiative may provide appropriate relief.

This is especially so because Mt. Sinai and the American Cancer Society are complete strangers to the underlying litigation. Neither Dr. Selikoff nor his co-authors will be witnesses at the Page trial. They are not consultants in that matter, nor was decedent Page a subject of their medical investigation.

....

Policy in New York, unlike most other jurisdictions, has accorded privilege to experts.... The United States Supreme Court held in the notable case of Branzburg v. Hayes, 408 U.S. 665, 688, [92 S. Ct. 2646, 2660, 33 L. Ed. 2d 626 (1972),] "that 'the public ... has a right to every man's evidence,' except for those persons protected by a constitutional, common law or statutory privilege." ...

If the production of the material would become oppressive and unreasonably burdensome, the court, in balancing the hardships, should consider whether there are other sources for obtaining the material needed to disprove the conclusions reached by the three studies. In addition, the probative value of the material, if produced, should be considered.

....

Dr. Selikoff determined by a sample test that it would take thousands of hours to redact all the material which is of a confidential nature. The time required would be the same whether the information is in raw data form or on computer tape.

Reynolds has offered to reimburse Mt. Sinai and the American Cancer Society for the reasonable expenses incurred in their complying with the subpoenae. However, it appears that Dr. Selikoff and his colleagues would still have to spend over a thousand hours of time removing the data identifying the individuals who participated in the study. The loss of the scientists' time in complying with the subpoenae would be an unreasonable burden and would unduly interrupt their ongoing medical investigations.

It is also clear that the medical material requested is not of an archival nature but rather is in constant use as part of ongoing medical research.

....

Mount Sinai and the American Cancer Society protest, on behalf of their medical scientists, the interference with their ongoing scientific research in this area. While these medical investigations are still in progress, they should not be subjected to examination and criticism by people whose interests are arguably antithetical to the medical scientists. It would have the effect of denying to these doctors the opportunity of first publication of their studies. It could also have a chilling effect and discourage future scientific endeavors.

Mount Sinai, the American Cancer Society, the Dean at the Yale University School of Medicine, and other eminent medical scholars claim that forced compliance with the subpoenae would result in an interference with their academic freedom. It was stated in Dow Chemical Co. v. Allen, 7th Cir., 672 F.2d 1262, 1275 [1982] that "whatever constitutional protection is afforded by the First Amendment extends as readily to the scholar in the laboratory as to the teacher in the classroom." Of course, like all other constitutional rights, academic freedom is not absolute and must be balanced against competing interests.... For present purposes, respondent's interest in academic freedom may properly figure into the legal calculation of whether forced disclosure would be reasonable.

For the reasons stated above, the court finds that compliance with the subpoenae would place an unreasonable burden upon the medical and scientific institutions involved and would unduly disrupt the ongoing research at both Mount Sinai and the American Cancer Society.

Accordingly, the motion to quash the subpoenae is granted.

Id. at 284-88, 518 N.Y.S.2d at 731-34. The tobacco companies did not appeal.B. The Present Subpoenas and the Decision Below

Shortly after the decision in Reynolds, the tobacco companies served on Mount Sinai and ACS the subpoenas at issue here, in connection with actions pending in federal court in Louisiana and Pennsylvania. See Lejeune v. Armstrong World Indus., No. 86-0421 (W.D.La.); Shires v. The Celotex Corp., No. 85-7141 (E.D.Pa.). In these new subpoenas, the tobacco companies seek fewer items than were sought in the Page subpoenas, concentrating primarily on the computer tapes storing the relevant raw data. To the extent relevant here, these subpoenas provide as follows:

1. The computer tape(s) for the 1979 study containing all information gathered, or that relates to matters occurring, prior to January 1, 1977.

2. The computer tape(s) for the 1980 study containing all information gathered, or that relates to matters occurring, prior to January 1, 1978.

3. The supporting documentation for the 1979 and 1980 studies.

The subpoenas define "supporting documentation" as comprising

computer code books for the 1979 and 1980 studies; related documentation that would assist in identifying what information is on the computer tapes for the 1979 and 1980 studies and where it is located on the tapes; a blank copy of the coding forms used in transferring information onto the computer tapes for the 1979 and 1980 studies; a blank copy of each type of questionnaire used to gather information for the 1979 and 1980 studies; a copy of the protocol documents for the 1979 and 1980 studies and any addenda thereto; and documents showing the calculations and the analytical methods and assumptions used in developing the information with respect to the 1979 and 1980 studies ....

The subpoenas also make the following provision for confidentiality:

Where necessary to protect the anonymity of the study participants, identifying information (such as names and street addresses) may be redacted....

Mount Sinai and ACS moved in the district court to quash the subpoenas on the grounds that (1) in light of Reynolds, the subpoenas were barred by res judicata and collateral estoppel; (2) under New York law, the data sought by the tobacco companies was subject to the absolute privilege afforded experts; and (3) even if they had no absolute privilege, the researchers enjoyed a qualified privilege and the tobacco companies had not shown that their interests outweighed the interests of researchers. In the alternative, Mount Sinai and ACS moved for a protective order allowing (a) redaction of all identifying information on the computer tapes, (b) a summary by decade of all specific dates, e.g., dates of birth, death, and employment, and (c) restrictions on access by third parties to the subpoenaed materials. They argued that such a protective order was necessary to protect the participants' anonymity and physician-patient privilege.

The district court denied the motion to quash but fashioned a protective order designed to preserve confidentiality. The court rejected the res judicata and collateral estoppel arguments on the ground that "New York State discovery rules and rules [in federal court] are different." It rejected the privilege argument on the ground that no state court precedent definitively established such a privilege; the district court viewed the Reynolds court's decision as quashing the Page subpoenas on the ground of burdensomeness and viewed the state court's discussion of privilege as dictum. Finally, though Mount Sinai and ACS argued that even the process of redacting the materials would be burdensome, the court stated that it would not be necessary for Dr. Selikoff himself to perform all the redaction and that a scholar such as he should anticipate that others would wish to scrutinize the bases for his published work. It concluded that "the burdensomeness is not something which is unduly burdensome ...."

Persuaded nonetheless that the study participants were entitled to some degree of privacy protection, the court entered a protective order granting the Mount Sinai-ACS motion in part. In relevant part, the protective order provides as follows:

1. The motion of Mount Sinai and the ACS for a protective order ... is hereby GRANTED to the extent that Mount Sinai and the ACS may redact the names of the subjects of the two studies ..., and their street addresses, town or village of residence, names of employers, social security numbers and union registration numbers, provided, however, that Mount Sinai and the ACS shall provide the county of residence of the subjects. The motion is DENIED in all other respects.

....

3. All persons, parties or other entities who are permitted access to the subpoenaed documents ... shall be ... prohibited from using any of the information contained therein to identify or attempt to identify the names of any of the subjects of the two studies, and each such person, party or other entity shall execute an Acknowledgment, in the form annexed hereto ....

4. In the event that American, Reynolds or Philip Morris, or any other party, is served with a subpoena ... or other discovery request ... seeking production of the subpoenaed documents, such party shall request the person seeking the documents to execute an Acknowledgment as described in paragraph 3. If the person ... refuses to execute an Acknowledgment and moves or applies to compel production, the party from whom such discovery is sought shall notify counsel for Mount Sinai and the ACS of such motion or application....

The Acknowledgment would require any person seeking access to the data to "agree to be bound by the terms and conditions contained [in the protective order], and ... consent to the jurisdiction of the Court for the enforcement thereof." The court rejected the Mount Sinai-ACS contention that considerations of physician-patient privilege required further redaction.

Eventually, following an aborted appeal to this Court, see In re American Tobacco Co., 866 F.2d 552, 556 (2d Cir.1989) ("orders requiring production and denying a further protective order are not final in the absence of a contempt adjudication"), the district court held Mount Sinai and ACS in contempt when they refused to comply with the orders requiring them to respond to the subpoenas. The court imposed on them a sanction of $500 for each day that they remained in contempt, but stayed enforcement of that order pending appeal.

This appeal followed.

On appeal, Mount Sinai and ACS contend principally that the district court erred in failing to give Reynolds the proper preclusive effect and in failing to recognize the state-law privileges afforded to experts and research scholars. In the alternative, they contend that the district court's protective order fails to protect adequately against disclosure of confidential information. We have considered all of their contentions and find no basis for reversal.

A. The Claimed Preclusive Effect of Reynolds

Mount Sinai and ACS contend that the present subpoenas arise out of the same operative facts and give rise to the same issues that were before the state court in Reynolds and that therefore principles of res judicata or collateral estoppel bar their enforcement. We disagree.

In order to determine the preclusive effect of the decision in Reynolds, we look to the law of New York. See Marrese v. American Academy of Orthopaedic Surgeons, 470 U.S. 373, 380, 105 S. Ct. 1327, 1331-32, 84 L. Ed. 2d 274 (1985) (federal court is to "refer to the preclusion law of the State in which judgment was rendered"); Ruiz v. Commissioner of the Dep't of Transportation of the City of New York, 858 F.2d 898, 902 (2d Cir.1988); 28 U.S.C. Sec. 1738 (1982). Under the transactional approach to res judicata adopted by New York, "once a claim is brought to a final conclusion, all other claims arising out of the same transaction ... are barred, even if based upon different theories or if seeking a different remedy ...." O'Brien v. City of Syracuse, 54 N.Y.2d 353, 357, 445 N.Y.S.2d 687, 688, 429 N.E.2d 1158, 1159 (1981). Principles of collateral estoppel will bar relitigation of an issue that is identical to an issue which has necessarily been decided in the prior action. Ryan v. New York Telephone Co., 62 N.Y.2d 494, 500, 478 N.Y.S.2d 823, 826, 467 N.E.2d 487, 490 (1984). Issues are considered identical if a different decision in the second suit would necessarily " 'destroy or impair rights or interests established by the first.' " Id. at 501, 478 N.Y.S.2d at 826, 467 N.E.2d at 490 (quotingSchuylkill Fuel Corp. v. B & C Nieberg Realty Corp., 250 N.Y. 304, 307, 165 N.E.456 (1929) (Cardozo, J.)).

These principles have been applied in federal court to bar an attack on a subpoena, where a New York state court had previously denied a motion to quash an identical subpoena. Sreter v. Hynes, 419 F. Supp. 546, 548 (E.D.N.Y.1976) (Pratt, J.); see also Westwood Chemical Co. v. Kulick, 656 F.2d 1224, 1227 (6th Cir.1981) (res judicata barred party from deposing corporate officers where prior decision quashed subpoenas for depositions of other officers of the same corporation). Where, however, the first subpoena has been quashed as overly broad and a second subpoena is served which is clearly narrower or more specific, New York law does not give preclusive effect to the decision quashing the earlier subpoena. See, e.g., Cunningham and Kaming, P.C. v. Nadjari, 53 A.D.2d 520, 384 N.Y.S.2d 383 (1st Dep't 1976) (mem.). Rather, "any possible future subpoena[ ] ... is to be examined on its own merits." Id. at 522, 384 N.Y.S.2d at 384.

As discussed in Part I above, the subpoenas at issue in the present case are plainly narrower than the subpoenas quashed in Reynolds. For example, whereas the Page subpoenas requested the raw data in its original form (e.g., interview notes, completed questionnaires, x-rays), the present subpoenas seek only the computer tapes plus such information as is necessary to interpret those tapes. Further, the present subpoenas, unlike the Page subpoenas, do not seek information that pertains to events occurring subsequent to the periods covered by the published articles.

Since the two sets of subpoenas are significantly different, the district court properly rejected the contention that enforcement of the present subpoenas was precluded by the decision in Reynolds quashing the broader subpoenas.

Mount Sinai and ACS also contend that they should not have been ordered to produce the subpoenaed materials in light of the state-law privileges accorded to experts and research scholars. Though in a diversity case the existence of a privilege is to be determined by reference to state law, Dixon v. 80 Pine Street Corp., 516 F.2d 1278, 1280 (2d Cir.1975); see Fed.R.Evid. 501, an existing privilege should be interpreted no more broadly than necessary. Since privileges shielding information from the reach of the court "contravene the fundamental principle that 'the public ... has a right to every man's evidence,' " such privileges "must be strictly construed." Trammel v. United States, 445 U.S. 40, 50, 100 S. Ct. 906, 912, 63 L. Ed. 2d 186 (1980) (quoting United States v. Bryan, 339 U.S. 323, 331, 70 S. Ct. 724, 730, 94 L. Ed. 884 (1950)); see Gray v. Board of Higher Education, 692 F.2d 901, 904 (2d Cir.1982). In light of these principles, we are unpersuaded that the orders of the district court should be overturned.

The claim to protection under the state-law privilege accorded to experts need not detain us long. In New York, experts who have no personal connection to a case enjoy an absolute privilege not to be compelled to give their opinions. Gilly v. City of New York, 69 N.Y.2d 509, 511, 516 N.Y.S.2d 166, 167, 508 N.E.2d 901, 901-02 (1987) (per curiam). This privilege has been extended to cover requests, pursuant to a subpoena, for preparation by an expert of a written medical report. See Plummer v. R.H. Macy & Co., 69 A.D.2d 765, 414 N.Y.S.2d 921 (1st Dep't 1979) (mem.).

In the present case, no expert is being asked to testify or to prepare a report. Mount Sinai and ACS have referred us to no New York decision extending the privilege to all existing documentary evidence in the possession of an expert, and our own research has revealed none. Accordingly, we conclude that this privilege is not applicable here.

In support of their claim to protection by a privilege accorded to research scholars, Mount Sinai and ACS are forced to rely on Reynolds, for no other New York case can be said to have recognized a scholar's privilege impeding production of research data. The Seventh Circuit has recognized such a qualified privilege, principally to protect scholars from the premature disclosure of their research, see, e.g., Deitchman v. E.R. Squibb & Sons, Inc., 740 F.2d 556, 560-61 (7th Cir.1984); see also Dow Chemical Co. v. Allen, 672 F.2d 1262, 1274-76 (7th Cir.1982), and the Reynolds court, citing Dow Chemical, explored the possibility that there might be such a privilege. The Reynolds court discussed the interests of scholars in their research data, as it did concepts of academic freedom, and it concluded that such concepts were permissible considerations in determining whether to quash the broad subpoenas before it. The court focused primarily on the need to protect the scholar against threats of interference with his ongoing research and against the release of his research findings prior to having had his own opportunity to publish them. 136 Misc. 2d at 287-88, 518 N.Y.S.2d at 734.

It is not clear to us, however, that the decision in Reynolds establishes the existence of a scholar's privilege as a matter of New York law. The court's discussion was discursive, and its conclusion that the subpoenas should be quashed did not state that the ruling was based on the existence of a privilege. Its concern for burden was plain, as it noted that the Page subpoenas were "not selective" but rather were "sweeping and indiscriminate," and it repeatedly referred to the principles that control where subpoenas are "overbroad, burdensome or oppressive," or "so oppressive as to hinder the normal functioning" of the recipient, or where "the production of the material would become oppressive and unreasonably burdensome," or the burdens would be "unreasonable." Id. at 284-86, 518 N.Y.S.2d at 731-33. The court concluded its discussion by stating that a scholar's "interest in academic freedom may properly figure into the legal calculation of whether forced disclosure would be reasonable," and that for all the reasons discussed, compliance with the Page subpoenas would "place an unreasonable burden upon the medical and scientific institutions involved and would unduly disrupt the ongoing research at both Mount Sinai and the American Cancer Society." Id. at 287-88, 518 N.Y.S.2d at 734. In the absence of a more explicit ruling, it is possible that the court regarded the scholar's interest in his research data as merely a factor to be taken into account in weighing the burdens of production, and that it did not intend to rule that the scholar has a privilege protecting him from having to disclose that data.

Further, if the Reynolds court did mean to rule that there is such a privilege, it plainly did not mean to suggest that such a privilege was absolute, for the discussion reveals at least two limiting concerns. First, as discussed above, the court considered at some length whether production would be burdensome, a concern that would have no relevance in the context of an absolute privilege. Second, the court indicated that its concern for research scientists focused in part on their interest in being the first to publish the results of their studies. Thus, the court stated that "[w]hile these medical investigations are still in progress, they should not be subjected to examination and criticism by people whose interests are arguably antithetical to the medical scientists," for a premature disclosure "would have the effect of denying to these doctors the opportunity of first publication of their studies." Id. at 287, 518 N.Y.S.2d at 733-34 (emphasis added).

In the present case, the district court viewed the Reynolds court's discussion of privilege as dictum and ruled that Mount Sinai and ACS had failed to establish that New York recognized a scholar's privilege. Even were we to disagree with this ruling and conclude that a qualified privilege does exist here, we cannot conclude that reversal is required. Where a qualified privilege is found to exist, a "court 'must apply a balancing test to determine whether the need of the party seeking disclosure outweighs the adverse effect such disclosure would have on the policies underlying the [claimed] privilege.' " Deitchman v. E.R. Squibb & Sons, Inc., 740 F.2d at 559 (quotingEEOC v. University of Notre Dame du Lac, 715 F.2d 331, 338 (7th Cir.1983)) (brackets in Deitchman ); see also United States v. Burke, 700 F.2d 70, 77 (2d Cir.) (discussing balancing test to be applied where discovery requests implicate the reporters' qualified privilege), cert. denied, 464 U.S. 816, 104 S. Ct. 72, 78 L. Ed. 2d 85 (1983). On appeal, we will not disturb the district court's application of such a balancing test absent an abuse of discretion. See Baker v. F & F Investment, 470 F.2d 778, 783, 785 (2d Cir.1972), cert. denied, 411 U.S. 966, 93 S. Ct. 2147, 36 L. Ed. 2d 686 (1973).

Here, notwithstanding its view that New York law did not establish a scholar's privilege, the district court asked the tobacco companies to "[a]ssume that there is a qualified privilege," and in that context discussed the burdens of compliance. The court ultimately concluded that the burdens on the scholars did not outweigh the tobacco companies' interests in obtaining the research data. Mount Sinai and ACS challenge this conclusion, emphasizing principally that scientific research efforts will be chilled by the notion that data underlying reported findings may be discoverable in actions to which the scientists are not parties, that the tobacco companies could conduct their own studies, and that compliance with the subpoenas will require Dr. Selikoff to spend a substantial amount of time in redacting the data on the tapes, time that will be taken away from his ongoing research. It was within the discretion of the district court to reject these arguments.

The principal legitimate chilling effect on scientific research, adverted to by the Reynolds court, is the possibility that research results discovered prior to their publication would be vulnerable to preemptive or predatory publication by others. But this possibility does not appear to be a factor here since the tobacco companies have narrowed their subpoenas to request only data that was relied upon by Dr. Selikoff in preparing articles that were published some years ago.

Further, if the mere fact that production will require time and effort were dispositive, only rarely would any discovery attempt be successful since most gainfully employed discovery targets could contend that compliance would take time away from their regular work. And though it seems clear that Mount Sinai, ACS, and Dr. Selikoff will be burdened somewhat in complying with the subpoenas, much of the detailed redaction work required in the interests of confidentiality may undoubtedly be performed, given proper instructions, by someone other than Dr. Selikoff or his fellow researchers. The tobacco companies have offered to reimburse Mount Sinai and ACS for the reasonable expenses of compliance.

Finally, though it may be that the tobacco companies could conduct their own studies in an effort to controvert the findings of Dr. Selikoff, it seems that an inordinate length of time would be required in order to duplicate the Selikoff study, and it is clear that scrutiny of the Selikoff data would provide a logically permissible manner in which to attack the findings. Since Dr. Selikoff is acknowledged to be the preeminent authority in the area and it is anticipatable that the expert witnesses to be called by the plaintiffs in the underlying suits will rely on his findings, the district court was not required to relegate the tobacco companies to undertaking independent studies rather than pursuing the most direct method of attack on the Selikoff findings.

We are unpersuaded that a contrary result is required by the fact that the medical researchers have no direct interest in the underlying lawsuits. The publication of their findings and conclusions invites use by persons whom the findings favor and invites reliance by the finders of fact. The public has an interest in resolving disputes on the basis of accurate information. Though under New York law this interest does not warrant requiring scientists to testify or prepare reports in actions with which they have no personal connection, we doubt that New York law would recognize a privilege to withhold the data on which already published findings are based and thereby preclude direct scrutiny of the findings' validity. Thus, even assuming that there is a qualified privilege for research scientists under New York law, we conclude that the scientists may be required to produce data underlying their published findings.

This obligation should not, of course, be without reasonable limitations. We are concerned by the information, given us by counsel for Mount Sinai and ACS at oral argument of this appeal, that there are some 40 extant subpoenas in various lawsuits seeking the data underlying the Selikoff findings. Though we uphold the decision of the district court in the present case requiring compliance with the present subpoenas under the protective order fashioned by the court, we do not foreclose the possibility that repeated requests for production of such data could become unduly burdensome. It may be that any excessive burden could be averted by the creation of a central repository of such data to which current and future suitors could have access, with the proper provisions for confidentiality, or by resort to some other mechanism such as centralized discovery through multidistrict litigation, see 28 U.S.C. Sec. 1407 (1982); see generally In re Asbestos and Asbestos Insulation Material Products Liability Litigation, 431 F. Supp. 906, 910 (J.P.M.L.1977) (per curiam) (declining to transfer to one district 103 pending actions involving asbestos exposure "under the[ ] circumstances," where all parties opposed transfer and some actions were far more advanced than others).

We are persuaded, however, that the proper balancing of interests does not warrant a bar against disclosure but only protection against needlessly repetitive disclosures. Hence, in the present case, we conclude that even if New York law recognizes a qualified privilege for research scholars, the decisions of the district court did not abrogate that privilege.

The protective order entered by the district court allows Mount Sinai and ACS to redact the names of the participants in the studies, their street addresses, towns or villages, social security numbers, employers, and union registration numbers. Mount Sinai and ACS contend that they should also be allowed to redact counties of residence and union local data and to summarize birth and death dates by decade. They assert that without these modifications, the tobacco companies and others could identify the subjects of the studies. We decline to modify the court's protective order.

District court decisions fashioning discovery protective orders are to be reversed only if there has been an abuse of discretion. See Galella v. Onassis, 487 F.2d 986, 997 (2d Cir.1973). While it might be possible for the tobacco companies to determine the identities of some of the research participants from the information remaining on the computer tapes after the redactions ordered by the district court, the court has enjoined them from doing so. Thus, the protective order provides that "[a]ll persons ... permitted access to the subpoenaed documents ... shall be ... prohibited from ... identify[ing] or attempt[ing] to identify ... the subjects of the two studies." Further, any person seeking access to the subpoenaed materials must first execute the Acknowledgment attached to the protective order, which subjects that person to the terms and conditions of the court's order. Presumably the court would impose suitable sanctions against any person who, in violation of these provisions, did attempt to identify a subject of the study. We cannot say that the court was required to grant further redactions as requested by Mount Sinai and ACS.

We also reject the contention that further redaction was required in order to preserve the physician-patient privilege. That privilege can be adequately protected by a court order that prohibits disclosure of the patient's identity. See Hyman v. Jewish Chronic Disease Hospital, 15 N.Y.2d 317, 258 N.Y.S.2d 397, 206 N.E.2d 338 (1965). Assuming the existence of facts prerequisite to the assertion of such a privilege, we conclude that the protective order's provisions for confidentiality are sufficient to preserve the privilege.

In sum, the protective order fashioned by the district court was not an abuse of discretion.

The orders of the district court are affirmed.