A watchdog group this week urged the National Institutes of Health to put an end to “ghostwriting,” alleging that scientific publications by some federally funded researchers were actually written by drug and marketing companies — including one book co-written by Stanford psychiatry professor Alan Schatzberg.

The Washington group, Project on Government Oversight, bases its allegations on documents made public in a lawsuit against GlaxoSmithKline related to the drug Paxil, as The New York Times first reported on Tuesday.

The documents include a draft of the book “Recognition and Treatment of Psychiatric Disorders: A Psychopharmacology Handbook for Primary Care,” which was published in 1999 by American Psychiatric Publishing, Inc. (APPI), a branch of the American Psychiatric Association (APA). The draft was “developed” by two writers for the medical publishing company Scientific Therapeutics Information (STI) of New Jersey under an “educational grant” from drug company SmithKline Beecham Pharmaceuticals (now GlaxoSmithKline), according to the document. Schatzberg and Charles Nemeroff, at the time chairman of psychiatry at Emory University, are the book’s authors, the draft said.

Also among the documents was a 1997 letter from STI editor Sally Laden to Nemeroff describing the company’s timeline for completing the book. The timeline included several dates when drafts of the book would go to the co-authors and the “sponsor” for comments, “sign-off” and “final approval.”

“You and Alan [Schatzberg] are in good hands with Diane [Coniglio, an STI writer],” Laden wrote to Nemeroff.

Schatzberg is the former president of the APA, which issued a statement on Tuesday condemning the ghostwriting allegations as untrue.

“Unrestricted” grants, such as the one SmithKline Beecham awarded for the book, “support specific projects” but do not allow companies to control books’ content, the APA said.

“From our perspective, timeline and details in the letter were never approved by APPI [the publisher] or the authors,” the statement added.

“This type of editorial assistance was quite common, especially the use of editorial experts to compile and check facts in books on pharmacology,” said Ron McMillen, chief executive of APPI, in the statement. “To say the book was ghostwritten is not true.” He added the authors and the publisher did not sign a contract until two years after the letter from Laden, and that the book was peer-reviewed by eight physicians.

Schatzberg, Nemeroff and STI did not return requests for comment.

Schatzberg, the chairman of psychiatry and behavioral sciences at Stanford for 19 years before stepping down this summer, faced congressional scrutiny in 2008 when Sen. Chuck Grassley (R-Iowa) raised questions about Schatzberg’s stock holdings in a drug company related to the professor’s federally funded research. Stanford attorneys at the time said the University took “appropriate steps” to prevent a conflict of interest. Schatzberg ultimately resigned as principal investigator for the grant.

Since then, the Stanford School of Medicine has revised the ethics policy it implemented in 2006, called the Stanford Industry Interactions Policy. Earlier this year it prohibited adjunct clinical faculty — volunteer teachers who often work in private practices — from giving paid speeches written by drug and medical-device companies.

The policy also forbids faculty, students and staff from engaging in ghostwriting: “in other words, individuals may not publish articles under their own names that are written in whole or material part by industry employees,” it says.

As medical school spokesman Paul Costello pointed out, Stanford’s ghostwriting policy would not have been in effect in 1999, when the book was published. He added that Schatzberg and the APA “strongly deny that the manuscript was ghostwritten.”

On Tuesday, doctors spoke critically to The Times about SmithKline Beecham’s alleged influence on the book. And members of the watchdog group, including former Grassley staffer Paul Thacker, pointed to other examples of what they called ghostwriting among federally funded researchers.

But Schatzberg told The Times that’s not what happened here. SmithKline Beecham was not involved in the book’s content and the 1997 letter was “a theoretical proposal that bears little, if any relationship to what actually happened,” he said in an e-mail, adding: “An unrestricted grant does not give the company any right of sign-off on content and in fact they had no sign-off in content.”