When GlaxoSmithKline P.L.C. marketers looked for doctors to promote the antidepressant Paxil, they called the project CASPPER.

The name was more than just an offbeat tribute to the friendly cartoon ghost. It was a wink and a nod to "ghostwriting," a questionable practice in which scientists put their names on research written by someone else, usually a writer paid by a drugmaker.

Ghostwriting critics say it disguises marketing material as scientific research.

Charges of ghostwriting have been lobbed against many companies in recent years, including Glaxo, Wyeth, AstraZeneca P.L.C., and Merck & Co. Inc., often arising in lawsuits from consumers claiming a drug hurt them.

The CASPPER documents are expected to play a role in a Philadelphia Common Pleas Court case being tried now that was filed by a Bensalem woman who says that taking Paxil during her pregnancy caused her son's heart defects.

Studies on ghostwriting have suggested that anywhere from 8 percent to 75 percent of articles in medical journals may involve the practice.

Dan Carlat, a Tufts University professor and frequent drug-industry critic, said he believed ghostwriting was "rampant" and represented just one way that pharmaceutical companies swayed research to boost sales.

"It's very subtle and nuanced . . . so it's very hard for a doctor not at the top of the field to detect," Carlat said.

Ghostwriting harms patients, said Adriane Fugh-Berman, a doctor and associate professor at Georgetown University and a paid witness in litigation involving complaints over Wyeth's hormone drugs, Premarin and Prempro.

Documents released in connection with 8,000 lawsuits filed against Wyeth over Premarin and Prempro show that the company, which employs several thousand people in Collegeville, paid a medical-writing firm to produce articles from 1998 to 2005 that allegedly downplayed the risks of hormone treatment and emphasized benefits.

In 2002, researchers stopped a landmark federal study after finding that menopausal women who took certain hormones had an increased risk of breast cancer, heart disease, and stroke.

"Ghostwriting was used to sway physician opinions to favor hormone use as disease prevention long after that was a scientifically defensible position," Fugh-Berman said.

Joseph Camardo, Wyeth's senior vice president of global medical affairs, called Fugh-Berman's view "baseless." Medical opinion evolves, he said, and the papers in question reflected scientific understanding then. He also said the authors alone controlled the content and writing of the papers, though some did have outside help paid for by Wyeth.

"It's really being misrepresented as something we wrote and paid for that said what we thought it should say," Camardo said. "But the authors were not paid, and the authors had the final say."

Like many companies, Wyeth has changed its policies as standards have evolved, and now discloses names of all contributors.

At Glaxo, employees hoped CASPPER, which stood for Case Study Publications for Peer Review, would create 50 articles in 2000 to "benefit the sales force by expanding the database of published data to support Paxil," a program brochure says.

Glaxo says only a handful of papers may have actually resulted from the CASPPER effort.

At the time, Paxil faced stiff competition from Pfizer Inc.'s Zoloft and Eli Lilly & Co.'s Prozac.

The CASPPER document urged Glaxo sales representatives to ask doctors to publish case studies of positive experiences with Paxil.

The Glaxo sales reps would then offer "editorial assistance" if the doctor is busy or needs writing help. The assistance could include "developing a topic" and "writing a first draft." The brochure later clarifies that "the development of a topic ultimately rests on the physician."

Hamid Moussavian, of New York, and Mark Miller, of Pittsburgh, two doctors who wrote positive articles connected to the CASPPER program, both said the ideas were theirs and they alone wrote them. Robert Lehman, a Baltimore doctor who also wrote a positive Paxil article associated with CASPPER, did not return a call seeking comment.

Leemon McHenry, a research consultant for Baum Hedlund, the Los Angeles law firm that discovered the CASPPER documents, said the program itself was problematic, whether or not authors had ghostwriters.

"What's important is the intention here," McHenry said, "and the intention is to bias the literature by cherry-picking these doctors."

Glaxo spokesman Kevin Colgan said the intent of the program was to "improve understanding of the benefits and risks of the product." He also said authors were heavily involved.

Robert Norris, president of Complete Healthcare Communications Inc., the Chadds Ford medical-writing company that worked on CASPPER with Glaxo, defended it as ethical.

"The name of the program? In hindsight, it was a horrible idea," he said, adding that "the title was whimsical, not sinister. At that time, around 2000, there were no negative ethical implications associated with the term ghostwriter."

Back then, writers who offered editorial assistance, including literature review and style editing, did not always get their names on publications, Norris said.

Standards have shifted as other controversies have surfaced. In some cases, drug companies asked scientists to put their names on papers they had not written. When those scientists refused, they sometimes saw very similar papers published under other scientists' names.

In 2003, editors of the Annals of Internal Medicine, a Philadelphia-based journal, learned from an article in the New York Times that a paper the journal had published on Vioxx had been written by people paid by Merck.

As such stories became public, some medical-journal editors started demanding disclosure of all contributors.

Annals of Internal Medicine editor Christine Laine said she wanted medical-research centers to forbid ghostwriting.

"If you're an academic researcher, the currency in that world, the thing that gets you promoted, is not just the quality of work but the number of publications, so the lure of being able to get another publication without a lot of personal investment is really tempting," Laine said.

She and others say consistent acknowledgment of all contributors, including any payments from drug companies, also would help.

"We don't think the writing assistance per se is bad," Laine said. "It's when it's not disclosed that it's a problem."