WILLIAM DAUBERT and JOYCE DAUBERT, Individually and As Guardians Ad Litem for JASON DAUBERT,
and ANITA DE YOUNG, Individually And As Guardian Ad Litem For ERIC SCHULLER,
Petitioners,

v.

MERRELL DOW PHARMACEURICALS, INC., Respondent,

ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT

BRIEF AMICI CURIAE OF
THE NEW ENGLAND JOURNAL OF MEDICINE,
JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION, AND
ANNALS OF INTERNAL MEDICINE
In support of respondent

I.  In Good Science, The Minimum Standard Of Credibility For Any Research, Requires Peer-Review And Publication.

"Good science" is a commonly accepted term used to describe the scientific community’s system of quality control which protects the community and those who rely upon it from unsubstantiated scientific analysis. It mandates that each proposition undergo a rigorous trilogy of publication, replication and verification before it is relied upon. Scientific advances occur, in part because dissemination of results by publication allows scientists to learn about and critique the reported research. In this way, scientists may expand upon what has already been achieved. Peer-review has long been utilized by technical journals to assure and improve the quality of the information they disseminate and is an essential prerequisite to publication.

Peer-review focuses on the methodology, analysis and results recited in an attempt to weed out error (e.g., from observer bias, improper statistical analysis, insensitive methods, carelessness and failure to recognize and control for confounding variables) and to assure that the text of an article is complete so that readers may judge the work for themselves. In this way, the peer-review process not only reduces the opportunity for reporting errors but assists in improving the quality of reporting errors but assists in improving the quality of the written report and impels scientist to raise the standards of research and writing. Its role in the publication process is nonetheless limited. Peer-review does not (and cannot) control what is selected by the editors of a journal for publication; nor can it guarantee that reported results are correct. Its role is to promote the promote the publication of well-conceived articles so that the most important review, the consideration of the reported results by the scientific community, may occur after publication.

Peer-review focuses on the methodology, analysis and results recited in an article in an attempt to weed out error(e.g. from observer bias, improper statistical analysis, insensitive methods, carelessness and failure to recognize and control for confounding variables) and to assure that the text of an article is complete so that readers may judge the work for themselves. In this way, the peer-review process not only reduces the opportunity for reporting errors but assists in improving the quality of the written report and impels scientists to raise the standards of research and writing. Its role in the publication process is nonetheless limited. Peer-review does not (and cannot) nonetheless limited. Peer-review does not (and cannot) control what is selected by the editors of a journal for publication; nor can it guarantee that reported results are correct. Its role is to promote the publication of well-conceived articles so that the most important review, the consideration of the reported results by the scientific community, may occur after publication.

Good science requires that a proposition be supported by experimental data, be reduced to writing, and be published after undergoing peer-review prior to any reliance thereon. The published article should include enough information about the design, methodology and analysis of the reported experiments or observations to permit other scientists to duplicate and verify the results as a basis for further study. Defense of a proposition before qualified peers – both in the form of peer-review and post-publication comments such as Letters to the Editor is the bedrock of the scientific credibility which justifies, inter alia, a court’s reliance upon the expression of expert scientific opinion in making findings of fact. Under certain circumstances, an expert’s opinion need not undergo peer-review unless the expert seeks to offer a view inconsistent with that previously published in peer-reviewed journals. Where the expert seeks to do so, she must submit the basis of the opinion for peer-review and publication before it is credible. Unlike the biases inherent in litigation -- which may skew an expert’s opinion or results to benefit one party over another -- good science is not primarily concerned with what the results may turn out to be, but rather with the accuracy and reliability of the experimental methodology and data by which they are reached.

In many circumstances, failure to seek peer-review and publication of research or results calls into question not only the ultimate conclusions reached, but also the methodology employed in conducting the research or analyzing the results. This is because unwillingness to subject one’s methodology to review – both during peer-review and by the scientific community after publication – stands as a telling admission that the standards of reliability associated with the peer-review process may not have been met.

Of course, it is unreasonable to expect all scientific analysis to be published in a peer-reviewed journal. However, for certain branches of science, such as biomedical science and, more particularly, epidemiology, bio-statistics and dysmorphology, and expert’s conclusions and opinions generally are not relied upon until they have been reduced to writing and published after review by a panel of experts. This review – a "cross-examination" of the methodology and analysis by those versed in the technical issues – either establishes the proposition as one which reasonably may be relied upon, or, in the alternative, suggests further work (e.g. revision of the article) thereby protecting the integrity of science.

II.  THE PEER-REVIEWED PROCESS

The publication and peer-review process begins when an article is received by a journal. Articles are usually submitted without solicitation, and a journal’s editor pre-screens them to determine if minimum criteria for publication are satisfied. (For example, the editor examines whether the subject matter is relevant to the journal’s readership and whether the article is a repetition of prior publications.)

The vast majority of the articles received by a journal which have not been initially rejected by the screening process are then submitted to a panel of outside experts selected on the basis of the paper’s content. Typically, a journal maintains its own roster of qualified reviewers, often listing thousands of experts, and selects at least two for each article. The identities of the reviewers are not required to be disclosed to the author under the peer-review system used by most journals, since it is believed that this frees the reviewers to be as critical and candid as is appropriate without fear of retribution or bias. In addition to the outside reviewers, many journals, such as the Annals of Internal medicine, also use in-house associate and technical editors who are expert in the relevant fields to review an article.

Once the outside reviewers have completed their assessment, typically within three weeks, they send their two part written reviews to the journal’s editorial staff. In the first part, each reviewer provides identifying and background information and an opinion on whether the article should be published, the priority for publication and the article’s content, and a copy of this part is usually provided to the author.

The reviewers’ principal function is to identify any analytical problems in an article, as well as inconsistencies, ambiguities or gaps in the stated facts or analyses. In this way, they try to make sure that an author neither overstates nor understates the work or results. However, the reviewers "are consultants, not arbiters of publishability. Peer review is not a straw vote among referees but rather a source of information and technical advice to the editors, who should make the final decision." In advising an editor, the reviewers will indicate whether the article should be published, needs revision and then should be re-reviewed, or should be rejected. Typically, an editor will require an author to revise the article notwithstanding the results of peer0review; and the final decision on whether to publish an article is the editor’s, often based upon a number of factors beyond the article itself such as space limitations, timeliness and priority.

As previously discussed, peer-review and publication does not end the scientific inquiry. Indeed, the mere act of publication frequently provokes scientific critique of the article by readers, who often submit Letters to the Editor commenting on the strengths or weaknesses of the work and its implications. Authors almost always respond in the same column with additional comments expanding upon or explaining a point in the article. In some instances a letter can lead to corrections or retractions, which then become part of the available scientific literature for further study.

The question before the court is not the role of peer-review and publication in scientific research – science itself has already determined its paramount importance. The Court must determine the role of publication in peer-reviewed journals, as part of good science under the rules of evidence.

At issue is whether, under the Federal Rules of Evidence, a court may exclude from evidence, or find insufficient, an expert’s opinion if it contradicts an existing body of scientific work published in peer-reviewed journals and others in the field would not recognize it as having been validly reached, e.g., by good science employing the checks of publication subject to peer-review. As recognized by the Ninth Circuit as well as the First, Fifth and D.C. Circuits, a court should look to whether the proposition being set out by an expert is in accordance with published scientific data whenever available – data developed by good science, utilizing publication and peer-review standards which exist independent of the biases inherent in litigation – before finding the opinion admissible or sufficient to establish a prima facie case. It is elementary that this inquiry should not be governed simply by the ultimate conclusion reached by the expert, but by the methodology and analyses used in reaching it in the context of any already published results.

Equally important to good science, and therefore the rules of evidence, is that the converse is not true: Results which have survived the peer-review process are not immune from further review or attack. Peer-review does not guarantee that results are error free. Publishing research results in a peer-reviewed journal does not end the research inquiry; it merely advances the knowledge with which research is conducted. The reported results become a building block upon which further research is based; and subsequent research, such as a replication of or attack on the initial results, forms the basis of continuing study once published in a peer-reviewed journal.

Subsequent work may seek to disprove the prior analysis or inquire into the researcher’s methods or results. However, the new research must also be conducted under the rigors of good science before it may be relied upon. Moreover, research to prove a new proposition or refute earlier work – and thereby invalidate both the work and anything that builds upon it – must also be peer-reviewed, although it must be scrutinized more rigorously than confirmatory research because of the inherent difficulties that would result if incorrect conclusions were relied upon.

The goal of publication and peer-review, as part of good science, is to prevent premature reliance on certain propositions by the scientific community before the supporting evidence has been reviewed by a panel of expert peers and published for all to see; the goal of the rules of evidence, as part of due process, is to prevent premature reliance on certain propositions by a fact finder before it has been found probative, relevant and reliable. Neither guarantees the accuracy of the information or assesses its weight in finding the truth. But both establish a threshold level of reliability. In either situation it would be imprudent to rely upon a proposition before, at a minimum, it has been passed upon for plausibility.