In an age of for-profit clinical research, this is the new face of scandal. Pharmaceutical companies promote their drugs with pseudo-studies that have little if any scientific merit, and patients naïvely sign up, unaware of the ways in which they are being used. Nobody really knows how often companies conduct such trials, but they appear with alarming regularity in pharmaceutical marketing documents. In the marketing plan for the antidepressant Lexapro for the 2004 fiscal year, Forest Laboratories described 102 Phase IV trials — the classification under which seeding trials fall — in a section labeled “Marketing Tactics.”

Oversight bodies like the Food and Drug Administration generally don’t view seeding trials as research scandals: seeding trials are not illegal, and the drugs in question have already received F.D.A. approval. But even after particularly egregious seeding trials have been exposed, the F.D.A. has not issued sanctions. Take the notorious Advantage study, a seeding trial of the pain reliever Vioxx conducted by Merck.

According to a 2008 report in the Annals of Internal Medicine, litigation documents show that the Advantage study was conceived and managed by Merck’s marketing department. Three subjects died in the Advantage trial; five more subjects experienced heart attacks. Oversight bodies should treat the Advantage study as a violation of research ethics.

How can studies that endanger human subjects attract so little scrutiny? Forty years ago, when most clinical research took place in academic settings, the main dangers to research subjects came in service to genuine scientific aims. A large regulatory apparatus was developed to protect human subjects from the ambitions of overweening academic researchers. In the early 1990s, however, pharmaceutical companies realized that it was faster and less expensive to conduct trials in the private sector, where the driving force is not knowledge, but profit. And the regulatory apparatus designed for the old era has proved woefully inadequate for the new one.

More NYT