The journals, bibles of the profession, have huge influence on which drugs doctors prescribe and the treatment hospitals provide. But The Observer has uncovered evidence that many articles written by so-called independent academics may have been penned by writers working for agencies which receive huge sums from drug companies to plug their products.

Estimates suggest that almost half of all articles published in journals are by ghostwriters. While doctors who have put their names to the papers can be paid handsomely for 'lending' their reputations, the ghostwriters remain hidden. They, and the involvement of the pharmaceutical firms, are rarely revealed.

These papers endorsing certain drugs are paraded in front of GPs as independent research to persuade them to prescribe the drugs.

In February the New England Journal of Medicine was forced to retract an article published last year by doctors from Imperial College in London and the National Heart Institute on treating a type of heart problem. It emerged that several of the listed authors had little or nothing to do with the research. The deception was revealed only when German cardiologist Dr Hubert Seggewiss, one of the eight listed authors, called the editor of the journal to say he had never seen any version of the paper.

An article published last February in the Journal of Alimentary Pharmacology , which specialises in stomach disorders, involved a medical writer working for drug giant AstraZeneca - a fact that was not revealed by the author.

The article, by a German doctor, acknowledged the 'contribution' of Dr Madeline Frame, but did not admit that she was a senior medical writer for AstraZeneca. The article essentially supported the use of a drug called Omeprazole - which is manufactured by AstraZeneca - for gastric ulcers, despite suggestions that it gave rise to more adverse reactions than similar drugs.

Few within the industry are brave enough to break cover. However, Susanna Rees, an editorial assistant with a medical writing agency until 2002, was so concerned about what she witnessed that she posted a letter on the British Medical Journal website.

'Medical writing agencies go to great lengths to disguise the fact that the papers they ghostwrite and submit to journals and conferences are ghostwritten on behalf of pharmaceutical companies and not by the named authors,' she wrote. 'There is a relatively high success rate for ghostwritten submissions - not outstanding, but consistent.'

Rees said part of her job had been to ensure that any article that was submitted electronically would give no clues as to the origin of the research.

'One standard procedure I have used states that before a paper is submitted to a journal electronically or on disc, the editorial assistant must open the file properties of the Word document manuscript and remove the names of the medical writing agency or agency ghostwriter or pharmaceutical company and replace these with the name and institution of the person who has been invited by the pharmaceutical drug company (or the agency acting on its behalf) to be named as lead author, but who may have had no actual input into the paper,' she wrote.

When contacted, Rees declined to give any details. 'I signed a confidentiality agreement and am unable to comment,' she said.

A medical writer who has worked for a number of agencies did not want to be identified for fear he would not get any work again.

'It is true that sometimes a drug company will pay a medical writer to write a review article supporting a particular drug,' he said. 'This will mean using all published information to write an article explaining the benefits of a particular treatment.

'A recognised doctor will then be found to put his or her name to it and it will be submitted to a journal without anybody knowing that a ghostwriter or a drug company is behind it. I agree this is probably unethical, but all the firms are at it.'

One field where ghostwriting is becoming an increasing problem is psychiatry.

Dr David Healy, of the University of Wales, was doing research on the possible dangers of anti-depressants, when a drug manufacturer's representative emailed him with an offer of help.

The email, seen by The Observer, said: 'In order to reduce your workload to a minimum, we have had our ghostwriter produce a first draft based on your published work. I attach it here.'

The article was a 12-page review paper ready to be presented at an forthcoming conference. Healy's name appeared as the sole author, even though he had never seen a single word of it before. But he was unhappy with the glowing review of the drug in question, so he suggested some changes.

The company replied, saying he had missed some 'commercially important' points. In the end, the ghostwritten paper appeared at the conference and in a psychiatric journal in its original form - under another doctor's name.

Healy says such deception is becoming more frequent. 'I believe 50 per cent of articles on drugs in the major medical journals are not written in a way that the average person would expect them to be... the evidence I have seen would suggest there are grounds to think a significant proportion of the articles in journals such as the New England Journal of Medicine, the British Medical Journal and the Lancet may be written with help from medical writing agencies,' he said. 'They are no more than infomercials paid for by drug firms.'

In the United States a legal case brought against drug firm Pfizer turned up internal company documents showing that it employed a New York medical writing agency. One document analyses articles about the anti-depressant Zoloft. Some of the articles lacked only one thing: a doctor's name. In the margin the agency had put the initials TBD, which Healy assumes means 'to be determined'.

Dr Richard Smith, editor of the British Journal of Medicine, admitted ghostwriting was a 'very big problem' .

'We are being hoodwinked by the drug companies. The articles come in with doctors' names on them and we often find some of them have little or no idea about what they have written,' he said.

'When we find out, we reject the paper, but it is very difficult. In a sense, we have brought it on ourselves by insisting that any involvement by a drug company should be made explicit. They have just found ways to get round this and go undercover.'