However, the conclusion has been hotly disputed by the organisers of the UK screening programme, who claim it saves the lives of more than a thousand women every year.

Critics argue that the methods used in the new study are fundamentally flawed, and the conclusions are not safe.

Alison Holts reports: "News ways of screening women"

The 10-year survey was carried out by scientists in Sweden, one of the first countries in the world to introduce a breast cancer screening programme.

The scientists found that there was an overall reduction of just 0.8% in deaths from breast cancer compared with the number that would have been expected without a screening programme.

The study, reported in the Swedish medical journal Lakartidningen, investigated the results of screening 600,000 women aged 50 to 69 between 1986 and 1996.

The tiny reduction in death rates contrasted sharply with the finding of the original clinical trials carried out in the country.

Those trials found cuts in the death rate of up to 30%, and precipitated the introduction of widespread screening in the UK in 1988.

BBC Science Correspondent Palab Ghosh: The research has been criticised

The new findings call into question whether the annual £34m cost of screening women aged 50 to 64 in the UK every three years is money well spent.

The Swedish researchers also warned that the screening programme had resulted in 100,000 women - one in six tested - being wrongly diagnosed as suffering from cancer.

Of these, 16,000 had undergone biopsy and more than 400 had surgery, including mastectomy.

UK review

Julietta Patnick is worried women will not go for screening

Last November an annual review of the UK programme showed a record number of 3,000 cancers were discovered at an early stage when the disease is most curable.

But it also showed that one woman in every 1,000 screened will develop breast cancer in the two years following an X-ray and that one in four cancers may have gone undetected at an earlier mammogram.

Supporters of breast screening estimate that the programme saves 1,250 lives a year in the UK.

They want more regular mammograms and the scheme to be extended to women up to the age of 69.

The BBC's Edward Sturton talks to Professor Baum and Julietta Patnick

Julietta Patnick, national co-ordinator of the NHS breast screening programme, said it was vital that women continued to attend for screening when invited.

She said 1995 figures showed death rates from breast cancer had fallen by 13% among women aged 50-64 since the screening programme was introduced.

The effectiveness of the programme was improving all the time, she said, and it was now estimated that death rates had been cut by 25%.

Ms Patnick admitted that it took time for a screening programme to start to save lives, but the UK system was now staffed by experienced experts and was subject to rigorous quality assurance monitoring.

She said: "Following a negative story about breast screening in the press last week we had women from all over the UK ringing up to cancel appointments.

"The worry is that this is going to cost lives. It makes me very sad because we do not know about many of those women on the end of the telephone line have the beginnings of a cancer that we might have found."

Screening 'sound'

Dr Michael Dixon, an honorary senior lecturer at the Edinburgh Breast Unit in Scotland said: "The science for mammographic screening for breast cancer is very sound.

"In fact, the new findings are easily explainable. Experience elsewhere has shown that there is a long lag between initiating a breast screening programme and reaching a level of detection that will reduce mortality."

The number of lives saved was linked to the experience of the radiologists carrying out the screening tests, he said.

One leading expert from the body which advises the Swedish government on breast cancer screening strongly criticised the new study in an e-mail to the BMJ published on the journal's website.

Professor Mans Rosen, deputy director general of the National Board of Health and Welfare in Stockholm, said: "The study ... suffers from too many methodological limitations and fallacies to be considered at all."

He said the researchers had not considered the fact that breast cancer incidence trends varied among the county councils studied and thereby influenced mortality trends.

Nor had they considered the need to have a long follow up to detect the effects of the programme.