A 1999 Merck & Co. study of its since-withdrawn painkiller Vioxx, touted to participating doctors and patients as meant to show whether Vioxx caused fewer stomach problems than another drug, was primarily a stealth marketing strategy, researchers report.

The true purpose was to get lots of doctors and patients in the habit of using Vioxx just in time for its launch, according to doctors who uncovered internal Merck memos discussing the strategy behind the study, called ADVANTAGE. They did so while reviewing roughly a million Merck documents for plaintiffs' lawyers preparing for trials in Vioxx lawsuits.

Drug companies are widely suspected of doing many such "seeding," or marketing studies, but there's been no "smoking gun" proving it before, according to the Annals of Internal Medicine, which published Merck's original report on ADVANTAGE in 2003 and will publish the new report Tuesday.

An accompanying editorial, co-authored by Annals editor Dr. Harold C. Sox, states the journal was not told the true purpose of ADVANTAGE, which compared Vioxx with an older, cheaper pain reliever, naproxen, when it published results indicating Vioxx was better tolerated.

Dr. Jonathan Edelman, head of scientific affairs at Merck Research Laboratories, said Monday "the ADVANTAGE study was primarily a scientific study" designed and executed by the company's clinical research unit and that any later use of data for marketing was a separate operation.

But Dr. Kevin P. Hill said he and colleagues, while working as paid consultants for lawyers representing plaintiffs who claimed Vioxx caused heart attacks or other harm, stumbled on documents indicating Merck's marketing division designed ADVANTAGE and handled the data collection and analysis.

Using funding from the Robert Wood Johnson Foundation's clinical scholars program, they searched further, uncovering items such as a memo from two top Merck executives nominating the study for an internal marketing award.

"The objectives were to provide product trial among a key physician group to accelerate uptake of Vioxx as the second entrant in a highly competitive new class," the memo states.

ADVANTAGE used about 600 family doctors new to clinical research, with each getting a stipend plus fees for recruiting a handful of patients each. Most clinical trials are run by a limited number of specialists at major teaching hospitals that each recruit hundreds of patients.

Vioxx came on the market in June 1999, after rival Pfizer Inc.'s Celebrex. Both makers claimed their drug caused less cramping, diarrhea and dangerous gastrointestinal bleeding than other pain relievers. Merck battled hard for market share, and Vioxx at its peak brought in about $2.5 billion a year.

Hill, now a staff psychiatrist at McLean Hospital in Massachusetts, said he and his colleagues found documents indicating "ADVANTAGE was marketing framed as scientific research," with an emphasis on how much Vioxx doctors in the study later prescribed.

"I don't think people would be willing to (risk side effects) if they knew that the aim of a clinical trial was to boost profits for a pharmaceutical company," Hill said.

The study's name implied it had a scientific purpose: ADVANTAGE, or Assessment of Differences between Vioxx and Naproxen to Ascertain Gastrointestinal Tolerability and Effectiveness.

But Hill said doctors participating in ADVANTAGE got a kit telling them how to talk to other doctors about Vioxx, and another Merck study running at the same time, called VIGOR, also examined how safe Vioxx was for people with gastrointestinal problems, so ADVANTAGE wasn't needed. VIGOR's results were published in 2000.

Merck spokesman Ron Rogers said Hill and his colleagues have been critics of Merck and just cherrypicked "some documents to support their thesis."

Dr. Bruce Psaty, a University of Washington epidemiologist, said Hill and his colleagues had disclosed their conflict of interest in their report but that the ADVANTAGE trial wasn't transparent about its purpose.

"I would think that at some level this is standard practice," he said of seeding studies, but they don't come to light except during the discovery phase of litigation.

Dr. Ross McKinney, director of Duke University's bioethics center, said seeding studies have been around for decades and usually are called postmarketing studies, meaning they're for drugs approved for sale. He said most never get published, but this one did because it addressed an important scientific question, stomach tolerability, even though it was a seeding trial. But McKinney said seeding studies make the public skeptical about enrolling in clinical studies.

"It's a serious violation of research ethics" and prevents patients from figuring out the risks and benefits of participating in the study, said Arthur Caplan, who heads University of Pennsylvania's medical ethics department.

Whitehouse Station, N.J.-based Merck pulled Vioxx off the market on Sept. 30, 2004, after its own research showed the then-blockbuster arthritis treatment doubled risk of heart attack and stroke. It has since reached a $4.85 billion settlement to end the bulk of personal injury suits over Vioxx, and the first payment checks are slated to go out later this month.

Sox, the Annals editor, and Dr. Drummond Rennie, the Journal of the American Medical Association's deputy editor, wrote that the institutional review boards required to protect patients in clinical studies should ask whether an experiment is a seeding study, particularly when there are obvious clues.

"Simply shining a bright light on their existence may have already sown the seeds of their destruction," they write.