Writing in How to Survive Medical Treatment (Century, 1987), Stephen Fulder says of any kind of mass screening: "For every case of a disease that is caught early there is a case of a disease that is treated when it would have got better by itself, and a case of a healthy person treated or biopsied because the screening test gave a wrong result."

That is exactly the line taken by Johannes Schmidt of the Practice for Family Medicine and Clinical Epidemiology in Switzerland. Writing in The Lancet (28 March 1992), he says that post morten findings show that many small cancers detected by mammography would have remained dormant if left undiscovered and uninterfered with. Mass breast screening is 10 times as likely to pick up a clinically benign cancer than it is to prevent a cancer death, he says.

The detection of dormant cases is particularly significant because the conventional (and unnecessary) treatment that invariably follows can be so devastating to the patient: radiotherapy, chemotherapy, long term treatment with powerful drugs which themselves have been shown to cause other cancers. It is no coincidence that the Bristol Cancer Help Centre among others has built its reputation on developing a programme simply to help cancer diagnosed patients survive their treatment.

Although mass screening in the UK is at an early stage, the experience of other countries suggests that mammograms have a high rate of inaccuracy. In Canada, during the first four years of an eight year trial, some 70 per cent of test results were unacceptable. Only in the last two years of the trial were more than half the tests up to the required standard. (The Lancet, 13 July 1991). There is also evidence from Finland that the reasonably high level of accuracy obtained during initial screening trials may not be repeated in a national programme, in which sensitivity of mammography may be 25-50 per cent lower (JAMA, 22-29 July 1992).

Besides the pointless exposure to radiation, routine screening may make women less alert to warning signs and delay seeking advice. Daniel Kopans of Massachusetts General Hospital, who has written a published critique of the above mentioned Canadian study, says: "Women may derive a false sense of security having had a negative screen for breast cancer.(The Lancet, 17 August 1991).

"Women who participate in screening should be reminded that a negative screen does not eliminate the possibility of cancer," says Kopans. "They should bring any new changes in their breasts to their doctor's attention despite a recent negative mammogram." In other words mammograms are emphatically not an alternative to routine self examination and awareness of the breasts.

This gross level of inaccuracy may be one reason why mass screening for breast cancer by mammography hasn't made much difference to survival rates. As Switzerland's Johannes Schmidt pointed out in his letter to The Lancet: "We should not overlook the finding that breast cancer mortality has remained unchanged for decades despite huge efforts to improve early detection and local treatment." Writing in The Lancet recently (24 October 1992), Kopans and others confirm the prevailing view that the death rate from breast cancer remains unchanged.

Where a malignant cancer is present, detection by mammogram, rather than physical examination, can help spread the disease. Animal studies have shown that if a tumour is manipulated, the tumour cell spread to other parts of the body can increase by up to 80 per cent (K Smatchlo et al, Ultrasound Med Biol, 1979 5: 45-49). In a letter to The Lancet (11 July 1992), Drs D Watmough and K Quan speculate that excessive levels of force ("as much compression as the women could tolerate") used in mammograms during an earlier study in Malmo, Sweden, might explain the findings by I Andersson et al (BMJ 1988; 297: 943-48) that 29 per cent more deaths occurred in the screened group than among the controls for women under 55 in the first seven years of follow up.