Fast forward to autumn 2004. A 69 year old woman—age 13 at menarche, age 22 at first birth, no family history of breast cancer, on oestrogen only hormone replacement therapy for 15 years—was referred to a surgeon because of discomfort in her left breast. As requested, she brought her most recent mammograms. On clinical examination neither patient nor surgeon could see or feel any abnormality in either breast. Bilateral diagnostic mammography showed a normal left breast but an obvious breast cancer in the right breast at 6 o'clock. Although it was smaller, the cancer was also clearly evident on previous mammograms.

Making this case some-what unusual was that the earlier films came from a screening examination done nine years earlier. At that time, a similarly placed right sided density had aroused suspicion, so many extra views were ordered; and a follow-up at six months was arranged, with more views recommended. Because the density had not changed after six months, the consultant surgeon and radiologist concluded that it was a false positive and reassured the patient. Surgery in November 2004 confirmed a 1.2 cm low grade, highly oestrogen positive adenocarcinoma with clear margins and a negative sentinel node.

I am that woman, and I was deputy director of the Canadian national breast screening study. Seven years from now will my circumstances be similar to those of Dr Roberts? Will my breast cancer be indolent? Or nasty? I don't know, but I suspect my prognosis is better than hers was. Should I be grateful that I lived for nine years unaware of my diagnosis? I am. Although women whose diagnoses have been delayed may take heart from my experiences, women with breast cancer confirmed by biopsy must not delay treatment.

Nevertheless like Dr Roberts I am compelled to reflect on what my experience has to say about the effectiveness of breast screening. How much of the benefit claimed for screening resides in the detection of cancers that ultimately will not prove lethal?

In my mind I can hear a chorus of voices from those who promote screening:

In a personal communication Don Berry, an expert in breast cancer statistics, described three possible outcomes of screening. One is that “the cat is already out of the bag” when breast cancer is detected by screening. Most deaths from breast cancer detected after screening fall into this category. The reductions in breast cancer mortality attributed to screening have never even remotely approached the 50% shown in methodologically sound trials. A signal failure of screening is that most women with screen detected cancers who would have died had they not been screened will still die from their breast cancers (Annals of Internal Medicine 2002:137: 363-5. [PubMed]). The second outcome is that death is prevented because screening succeeds in detecting lethal cancers before they have disseminated systemically. This represents the reported reduction in breast cancer mortality as a result of screening of as much as 30%. The third is detection of breast cancers that will never prove lethal. In this category fall not only most ductal carcinomas in situ currently being diagnosed but also my kind of tumour, which after nine years was still relatively small and (probably) node negative.

Dr Roberts asked, “What's the use of breast screening?” and “Might screening actually be detrimental?” She urged, as I have recently (Journal of the National Cancer Institute 2003;l95: 1508-11.), that women be “completely informed” before they accept screening, even though the information might “not be what they want to hear.” She concluded that “a rethink” about screening is needed.