Legal action by Pfizer against the New England Journal of Medicine (NEJM) has triggered a debate about confidentiality and accountability in the way medical publications review and publish pharmaceutical research.

The issue has arisen in relation to product liability lawsuits filed in the US against Pfizer over cardiovascular and other adverse effects attributed to the company’s COX-2 inhibitor analgesics Celebrex (celecoxib) and Bextra (valdecoxib).

In April 2005, Bextra was taken off the market in North America and Europe due to concerns about its cardiovascular safety. That echoed the fate of Merck & Co’s COX-2 inhibitor Vioxx (refecoxib) the previous year. However, Celebrex managed to survive regulatory scrutiny of its potential side-effects and remains on the market.

Writing in the journal Science, editor-in-chief Donald Kennedy notes that Pfizer is pressuring the NEJM for access to confidential reviews of papers the company claims are being used as evidence by some Celebrex and Bextra plaintiffs to bolster their case. According to Kennedy, Pfizer served the NEJM with a number of subpoenas to this effect and, when the journal refused to comply, followed up with a motion to be heard in a US District Court in Massachusetts.

Kennedy’s take on these actions is that Pfizer is engaged in a “fishing expedition” in the hope of finding something to bolster its defences. He quotes as follows from the company’s motion: “Scientific journals such as NEJM may have received manuscripts that contain exonerating data for Celebrex and Bextra which would be relevant for Pfizer’s causation defence.”

Kennedy also quotes Pfizer as arguing that the “public has no interest in protecting the editorial process of a scientific journal”. He sees this as an “assault” on an editorial system that enables journals to maintain a network of reviewers who “willingly provide, without compensation, confidential and candid evaluations of the work of others”. Given that “all of us in scientific publishing depend on reviewers, we’d better try to keep them at it, happy, and secure”, Kennedy comments.

He sees the issue essentially as a conflict between competing interests. “One is the public’s interest in a fair system of evaluating and publishing scientific work – one that offers high confidence in, though not an absolute guarantee of, the quality of the product,” Kennedy writes. “Pfizer dismisses this with a wave of the hand, a strangely inconsistent position given the enthusiasm with which it and other drug companies seek to have their own research validated by the very system of scientific publication that Pfizer’s motion decries and would undermine.”

On the other side, Kennedy continues, sits “a private interest in gaining information that might protect a corporate defendant against a plaintiff’s attack. Without questioning the legitimacy of the latter, it is surely fair to ask whether fulfilling that need should trump the public interest.”

In Kennedy’s view, the “prospective weight” of what Pfizer hopes to discover through access to confidential files does not justify the “prospective damage to the public interest” of overriding a system intrinsic to journal integrity. If “efforts of this kind were to succeed, the sad day might come when Science would have to add a firm caveat emptor to its instructions for peer reviewers”, he warns.

Pfizer issued a statement noting that subpoenas “are a routine part of fact-gathering in any litigation by both plaintiffs and defendants”. Indeed, the company adds, “in this litigation, both parties served subpoenas on a number of authors and/or medical and scientific journals”.

While some commentators have castigated Pfizer for what they regard as bullying tactics, others insist the case is not so clear-cut. They argue that the confidentiality principle is a front for a peer- and editorial-review system that is non-transparent, biased or conflicted, politically motivated and by nature incestuous. As one blogger asked: “If the peer review process is truly fair, objective and motivated by “pure” science, then what is there to fear about full disclosure on both sides?”