Marilynn Larkin is a self-employed writer and contributing editor to the Lancet, one of the most prestigious medical journals in Britain. While looking for work earlier in her career she agreed to ghostwrite a scientific paper for a medical communications agency retained by a drug company. It was her first and last incursion into that world.

"First I had to sign all kinds of forms not to tell anyone I was doing this," she said. "They gave you an outline, then provided tons of references you knew you had to use.

"I discovered I didn't like this kind of work. After I sent it, I got the whole thing back from the company with a sample from another company which read like PR writing. It was just a really straight sell. I said, 'I'm sorry, I can't do this.'"

The article was destined to be published in a medical journal supplement under the name of a research scientist. Ms Larkin says there are several kinds of scientists who will agree to have articles ghostwritten. "One is the person who has been around for centuries and is like a figurehead in the field. By that time, they don't care. They will take the money or pretend they didn't know they were taking the money for that reason.

"Or there are scientists who don't realise what is happening. They think they are getting help with writing and don't realise when they get feedback that the drug company is also going to review their work in most cases. There's no question that the drug company has the last say."

Many ghostwritten articles go into industry-sponsored supplements, she says, but many are published in the less prestigious medical journals themselves and "some of the ghostwritten articles go into top-tier journals", she says. "It is a pervasive problem with the whole medical publishing system."

In Ms Larkin's experience, the doctors who are supposed to check and approve the articles that go out under their names are not always assiduous about it. She once helped to edit a supplement based on talks given by research scientists at a symposium. "The writers did go back to the researchers to check the quotes for accuracy. More than half of the scientists did not get back to them ... They can't be bothered to look at it or they don't care. I couldn't believe it."