US Legal Principles and Confidentiality of the Peer Review Process

When drawn into other parties' litigation or investigations, journals often receive requests or subpoenas for confidential peer review documents. We describe reasons for such requests and reasons that journals resist producing confidential documents, ways that journals respond, and steps that journals can take to minimize third-party breaches of journals' confidential processes. We discuss legal principles on which resistance may be based in the United States, including the journalist's privilege, the scholar's privilege, the burdensomeness of response, and the Federal Rules of Civil Procedure, which protect against attempts to obtain expert opinion testimony without compensation. Illustrative cases are described in which courts affirmed the confidentiality of peer review conducted by journals.

During the peer review process, information and documents are generated that authors, reviewers, editors, and journal policy may consider to be confidential. Examples of such information include the identity of the reviewers and potential reviewers, substantive and editorial comments by outsider peer reviewers and internal editorial reviewers, recommendations on acceptance for publication, written discussion about whether a manuscript should undergo peer review, and the manuscript itself in its various drafts. Confidentiality varies among these items, and it may be treated differently when third parties request such information.

In the United States, journals may receive either a subpoena for confidential information or a "request" from a federal agency, such as the Office of Research Integrity, the Food and Drug Administration, and the Office for Human Research Protections. Many federal agencies do not have subpoena power and thus attempt to secure information by citing their regulatory authority; they generally cannot compel a journal to provide the requested information. Subpoenas may be issued in conjunction with civil litigation between private litigants or in conjunction with a criminal investigation. The types of cases in which a civil litigant might serve a subpoena to a journal or editor include patent infringement or invalidation, toxic torts, product liability, and medical malpractice. Often the subpoena that seeks the production of certain documents is followed by or joined with a notice of deposition that seeks the testimony of an editor or reviewer.

Litigants seek facts that might assist their cases; for example, a reviewer provided information that indicated that the "invention" disclosed in a publication was well-known to those in their discipline and thus was not an invention. Other litigants seek to identify counterexperts who they can use in their case; that is, they might hope that the editors have selected or identified reviewers who have specialized expertise. Litigants sometimes hope to undermine the opposition's experts by culling critical remarks from the reviews about the opposition's experts' theories and or publications.

To protect confidentiality, journals and authors use an assortment of federal and state constitutional provisions, state shield laws, and common law protections. The legal doctrines a journal might cite when resisting a subpoena for confidential information will depend on the facts and circumstances of the subpoenas. It will also depend on the jurisdiction in which the subpoena is served. A journal would be subject to the jurisdiction of the courts in which it operates or has significant contacts. Often the authors and reviewers work in other jurisdictions and may even be in other countries. In such cases, litigants will try to establish that the reviewers, editors, and journal had sufficient contact with the jurisdiction in which the case is pending to be subject to the court's jurisdiction. If not, the litigants will attempt to get the information from a party who is subject to the court's jurisdiction.

Successful motions to quash have been premised on the journalist's privilege and the scholar's privilege, as well as the First Amendment. Motions may also cite the burdensomeness of responding¹ and Rule 45 of the US Federal Rules of Civil Procedure, which protects individuals against an attempt by a party to litigation to get expert opinion testimony without paying for the expert's time.

Inherent in every example of a privilege that shields documents or individuals from requests for discovery is a tension between society's need for relevant evidence and the need to protect a nonlitigant's confidential information from unwarranted disclosure. Thus, the following balancing test is conducted.

First, the court examines the hardship that the party resisting discovery will suffer if compelled to produce the requested documents. Second, the court will examine whether the disclosure is relevant and necessary to the party requesting discovery and will compare that to whether the discovery is deemed harmful to the resisting party. Third, the court will give more weight to interests that have a distinctly social value than to purely private interests.

The following cases involved subpoenas for confidential peer review information and illustrate how the journals and courts resolved issues, depending on the nature of the information sought.

The case of Block v Abbott Laboratories² reflects the types of information requested in a subpoena, the circumstances in which such a request might be made, and the responses of journals and authors on receiving such a subpoena. In Block v Abbott Laboratories, a litigant requested information and documents. The litigant requested "[a]ll documents relating to" a published article³ and letter to the editor,⁴ "including, without limitation, all correspondence, reviews, evaluations, written comments and proposed or actual revisions." The word "document" was defined "in the broadest sense of the term, . . . and shall include without limitation all writings and recordings of every kind regardless of the medium of storage, including all notes, summaries, correspondence, electronic mail, drafts, attachments to documents and all copies of documents that are not identical duplicates of the original, whether or not the original is in your possession, custody or control."

The journal resisted producing the documents, asserting that communications between authors and editors are confidential, as are communications among them and peer reviewers and staff, and that confidentiality is needed to ensure frank analysis and deliberations. The journal argued further that maintaining confidentiality serves to protect the public interest in the free flow of this discussion. The journal argued that the documents were privileged from disclosure by "among other privileges, the newsgathering and editorial privilege, the peer-review analysis privilege, and the self-evaluative privilege." The privileges the journal cited are not formally recognized, but reflect a strong public policy underpinning a news gatherer's right to cite confidential sources and the confidentiality typically afforded a peer review or self-evaluative process. The litigants did not oppose the motion to quash, and the court granted the journal's request to be relieved of the obligation to produce the documents.

In the same case, the litigants sought from the author an unpublished scholarly article that was in press at the same journal and "all materials relating to its submission for publication." In quashing the subpoena, the court noted,⁵ "the article is not finished, is subject to revision and should not be considered complete until it is published." Further, the court noted the article was "highly likely" to be published before the trial for which it was wanted. Also in the court's balance was the researcher's need to "preserve the integrity of his work, protect his intellectual property and safeguard his reputation and credibility" by controlling the distribution of his article before publication. The court wrote, "[The author's] interests cannot be protected unless he is able to retain control of the article until it is complete; and where the purpose of the article is publication, complete must be defined as published. Only in this way is [the author] able to preserve the integrity of his work, protect his intellectual property and safeguard his reputation and credibility. ‘Academicians engaged in pre-publication research should be accorded protection commensurate to that which the law provides for journalists.' Cusamano v. Microsoft Corp., 162 F.3.708, 713 (1^st Cir. 1998). . . . [T]he product of [the author's] efforts is fairly considered confidential."

The protection provided to reporters may extend to journals. In Cukier v American Medical Ass'n,⁶ an author had sought the identity of persons or entities who had made statements to the JAMA editors indicating that the author had a financial interest in the publication of the manuscript. The author indicated that he had no financial conflict of interest, but the manuscript was declined for publication. The author subsequently filed suit seeking the identity of the person(s) who had made such statements to JAMA. JAMA resisted providing such information, citing the Illinois Reporter's Privilege Act among other defenses. In support of this argument, JAMA submitted an editor's affidavit that described the confidentiality of its peer review process and JAMA's commitment to abide by the strict standard of peer review confidentiality of the International Committee of Medical Journal Editors.⁷ The court found that the JAMA editors were "reporters" within the meaning of the Reporter's Privilege Act.

A separate "scholar's privilege" is recognized by courts in some jurisdictions, typically in cases in which a study is incomplete or unpublished. Again, a balancing test is applied. In Solarex Corp. v Arco Solar, Inc.,⁸ Arco was a party in a patent infringement case and sought the identity of a scholarly journal's confidential peer reviewers and "all documents . . . relating to the submission." The authors of a scholarly manuscript had submitted it to a first journal. The editor of the first journal forwarded the manuscript to experts in the field, indicating the confidential nature of the document. One reviewer gave a positive review; the other gave a negative review. The editor forwarded the critical review to the authors and suggested revision. Rather than rewriting the article, the authors submitted the article to a second journal that published it. The litigants sought the identities of the first journal's peer reviewers to determine whether they had circulated the article to others such that the article would have been deemed to have been published under intellectual property law and would constitute "prior art" such that the relevant patent would be invalid. (The comments of the reviewer had already been produced because they had been provided to the author and thus were not deemed confidential.) The court declined to find a specific peer review privilege, but, when balancing the interests of the parties and society, the court noted that if independent reviewers believed their identities would be revealed, it was likely that they would not be as forthcoming in their criticisms. Thus, the court quashed the subpoena.

Journals often wish, for a variety of reasons, to resist subpoenas and requests for information arising from the peer review processes. First, complying with a subpoena is disruptive to the journal's activities and processes. Time and resources spent responding to a subpoena detract from the journal's other activities. Second, substantial costs can be incurred in responding to a subpoena, particularly a broad subpoena, for example, a request for all communications relating to the review of a manuscript, including the reviewers considered for such manuscript. In addition to the search and duplication costs, if documents are produced and the identities of reviewers and editors are revealed, litigants often will seek to depose the reviewers and editors. Providing testimony carries risks that are difficult to foretell, and depositions can involve significant legal expenses for the journal and its reviewers. Third, providing confidential information may violate the confidentiality obligations that a journal has assumed with submitting authors, and it may be deemed a waiver of confidentiality such that a journal will have difficulty resisting future requests for such information. Moreover, the perceived breach of trust may damage valuable relationships not only with the authors and reviewers involved in the case, but also with other current and potential authors and reviewers. Finally, a subpoena may be a form of harassment the litigants use against the journal or submitting authors who support a position contrary to the litigant's position.

If a subpoena or request for information arrives, a journal can take several steps to control what, if any, information will be released. On receiving such a request, the journal editors must recognize that, in general, when crafting subpoenas, attorneys are trained to ask for every document that might have some value for their case and often the attorneys do not know how broad their request is. Thus, editors can object to the scope and burdensomeness of responding to such a request and begin to negotiate with the parties who have served the subpoena.

Negotiation can be complex. Often, a recipient of a subpoena or request can convince a party to narrow the scope of the request, for example, to release the journal from providing all documents relating to the submission and to provide only the most relevant substantive communications or the reviews but not the identities of the reviewers. The litigants may allow the journal to redact irrelevant confidential information. Journals may require that the documents containing confidential information be destroyed or returned after they have been reviewed or the case has been concluded. Disclosure may be limited to experts or attorneys such that the confidentiality of the documents is preserved as much as possible. Finally, the journal may seek indemnification if the reviewers or authors sue the journal for breach of confidentiality; indemnification can include reimbursement for the cost of litigation and any award made against the journal for such breach. If such negotiations are unsuccessful, the journal can file a motion to quash to allow the court to determine the journal's need to comply with the subpoena.

A variety of business and legal reasons exist for journals to resist providing confidential communications in response to a subpoena or request for such communications. Journals that wish to preserve the confidentiality of their peer review and editorial processes have a variety of legal principles and techniques available to them to resist producing such confidential documents. In evaluating how to respond, the journal should consider not simply the case that prompted the subpoena or request, but also the long-term consequences for the journal and for the biomedical communications community of providing such information without resisting such a disclosure.