Two articles and an editorial in the April 16 issue of the Journal of the American Medical Association (JAMA) describe articles ghostwritten by employees of Merck & Company or by outside writers hired by Merck. Then Merck had well-known academic physicians, who again may have had nothing to do with the research, sign on as primary authors.

The ghostwritten articles concerned refecoxib (Vioxx®), a COX-2, selective, nonsteroidal, anti-inflammatory drug that Merck voluntarily withdrew from the market in September 2004 after studies found that the drug may increase the risk of heart attacks and strokes.

Using documents made public during the lengthy Vioxx litigation, the authors of the JAMA articles were able to demonstrate Merck's sleight-of-hand in manipulating the authorship of some of the Vioxx publications.

It's important to emphasize right now that Merck, with its Vioxx troubles, is by no means the only pharmaceutical company today who's playing fast and loose with the authors' names that are attached to research reports about profitable drugs.

Here's how it works. A pharmaceutical company designs and begins a clinical trial, collects the results, and carries out a statistical analysis of the data. The drug company's own employees, or medical writers of for-profit information companies, then prepare a draft describing the study and the results. Last, the pharmaceutical company invites bonafide investigators with suitable academic affiliations to step up and sign on as the article's primary "authors."

The academicians who accept the invitation to "author" such clinical-trial or review articles may have had little to do with either the study itself or the review, and yet they may be paid by the company to present themselves as the authors of the article.

Why, exactly, are these practices wrong? First, because every author listed at the head of a scientific article is expected to take responsibility for the accuracy of its results.

Sure, some of the academic "authors" of these articles may review them carefully and insist on changes, but the available evidence that surfaced during the VIOXX litigation indicates that others of these "authors" allowed their names to be used with little or no examination of the article.

And even if several academics do review the article carefully, if they were not directly involved in the actual research they will be unlikely to notice if a study's results have been skewed in such a way that favors the company's product.

Conclusions regarding the significance of a trial's results depend heavily on the statistical methods applied in the data analysis. (Remember the oft-quoted statement by Mark Twain, "There are three kinds of lies: lies, damned lies, and statistics.") A drug company could easily alter the data it has collected by conjuring up reasons to throw out unfavorable results; or, it could manipulate the analysis by choosing the statistical method that will present the results in the best light.

Even if the drug company has not massaged the data in either of these unacceptable ways, an "author" who's uninvolved in the trial will have no way of knowing whether the written report is accurate. It's hard enough for honest and compulsively precise clinical investigators to avoid such pitfalls in their own trials.

Even though the frequency of ghostwritten authorship is uncertain, questions need to be raised about the legitimacy of any studies and promotional activities spawned in the first place by pharmaceutical companies.

The trust of the medical community in any articles about such drug-company sponsored studies will be tough to restore now. And such trust-building efforts will require the cooperation of clinical investigators, journal editors, peer reviewers of articles, and medical educators.