Hastings Center Report

The vision most people used to have of medical research is that professors gain funding from universities or government agencies, go to the lab for careful observation and experiment, and after some years publish their hard-won results in scholarly, peer-reviewed medical journals. This is probably a simplified story that never played out quite as envisioned. As a result of commercial interests in the outcome of medical research, we are now even further from it. Pharmaceutical companies' marketing departments have developed a novel strategy. The old vision allowed costly failures when reality did not cooperate with the company's plans for its new molecular entities. With the rising price of bringing a new drug to the market, a method that guaranteed success was needed. Instead of fitting the conclusion to the evidence, the industry's strategy now is to "defend the molecules": Select the data that promote the drugs, file away the results that are unfavorable, and then buy just the right academic credentials to sign on to ghostwritten articles produced by the marketing department staff or by public relations agencies employed by the company. Publish these articles in the most respected [End Page 17] medical journals. Complete the circle of evidence by citing the published ghostwritten journal articles in "Dear Doctor" letters and other promotional materials, as if they were independent verification of the effectiveness and safety of the drugs.

To some extent, academic medicine has long involved ghostwriting. Scientific writers have often written up the results of studies. But while this practice deprives the authors of the credit they rightly deserve, it does not threaten to undermine the scientific integrity of medical research. What is relatively new on the scene is industry-sponsored ghostwriting. Drug promotional materials masquerade as legitimate scientific research, and competition between the pharmaceutical companies for the market share is fought out using "key opinion leaders" who are paid to sign on to the ghostwritten articles.

What follows is an account of one such case. The documents that build the case were filed into the public record by GlaxoSmithKline's lawyers on June 23, 2004 (docket numbers 290 and 291) in the In re Paxil Products Liability Litigation MDL 1574. They were then used in an ABC Primetime program on antidepressants that aired on December 9, 2004.¹ The author is a research consultant for Baum-Hedlund, the Los Angeles-based law firm that filed the suit on behalf of individuals suffering from withdrawal symptoms upon discontinuing Paxil, which is one of the antidepressants known as selective serotonin reuptake inhibitors (SSRIs).

As it turns out, the differences in efficacy among the SSRIs are minimal, so their side effects profiles can make all the difference in how doctors prescribe them. In the mid-1990s, Eli Lilly, which manufactures Prozac (another of the SSRIs), began attacking Paxil, manufactured by GlaxoSmithKline—at that time, SmithKline Beecham—on just these grounds. In 1996-97, Lilly sponsored conferences and journal supplements focusing attention on the side effects caused by Paxil.² Paxil has a short "half-life"—the time after taking a drug at which half of the dose remains in the body—and therefore a short wash-out period when compared to Prozac. Paxil's half-life is twenty-one to twenty-four hours; Prozac's four to sixteen days. Because of the short half-life, a patient who abruptly ceases therapy with Paxil tends to have more severe withdrawal symptoms.

Lilly's strategy was to play up the problems associated with Paxil's short half-life. One key paper sponsored by Lilly was Jerrold F. Rosenbaum and John Zajecka's "Clinical Management of Antidepressant Discontinuation," presented at a closed symposium titled "SSRI Discontinuation Events" and later published in a supplement of the Journal of Clinical Psychiatry. ³ In this short paper, Rosenbaum and Zajecka stressed the importance of gradually tapering all SSRIs except fluoxetine (Prozac). They also suggested substituting fluoxetine in cases where symptoms of "discontinuation"—as they now called withdrawal—persevere even when the SSRI is tapered slowly.⁴ Both citations for this claim refer to cases of paroxetine withdrawal...