Mammography screening: the honest numbers per 1,000 women

Few screening tests have been argued over as fiercely as mammography. Supporters point to lives saved; critics point to women treated for cancers that would never have harmed them. Both sides are right about something. Mammography prevents some breast cancer deaths, and it also finds some cancers that did not need finding. The useful question is not whether screening "works", but how large each effect is — and that is best answered in absolute numbers.

What screening achieves

In 2012 an independent panel commissioned in the UK and chaired by Sir Michael Marmot reviewed 11 randomised trials. Inviting women to screening reduced breast cancer mortality by about 20% (relative risk 0.80). Translated to the UK: for every 10,000 women invited from age 50 for 20 years, about 43 breast cancer deaths are prevented and about 129 women are overdiagnosed — roughly three overdiagnosed women for each death prevented. The panel stressed that these estimates carry substantial uncertainty.

The German numbers, per 1,000 women

The official German decision aid, published by the Federal Joint Committee (G-BA) and based on reports from the Institute for Quality and Efficiency in Health Care (IQWiG), puts it this way for 1,000 women who take part regularly for 25 years:

• without screening about 24 die of breast cancer; with screening 16 to 21 — so 3 to 8 deaths are prevented;
• 11 to 15 women receive an overdiagnosis: a cancer that would never have caused symptoms in their lifetime, but is treated anyway;
• in each screening round, about 30 women are called back for further tests; 24 of them turn out not to have cancer. About half of the women who need a biopsy do not have breast cancer;
• about 2 in 1,000 are diagnosed with breast cancer in the two years between screenings despite a normal result (interval cancers).

Why overdiagnosis is the hardest harm to grasp

A false positive is frightening but ends with the all-clear. Overdiagnosis is different: the cancer is real under the microscope, yet it would have grown so slowly, or not at all, that it would never have troubled the woman. No individual can know whether her cancer was one of these, so every woman diagnosed is treated. Screening also makes survival statistics look better than they are, because diagnosis is simply moved earlier. We explain both effects in overdiagnosis and lead-time bias.

The German programme

The Mammographie-Screening-Programm invites women to a two-view X-ray mammogram every two years in specialised screening units; two doctors read every image independently. On 21 September 2023 the G-BA raised the upper age limit from 69 to 75; this took effect on 1 July 2024, and women aged 70–75 initially had to book appointments themselves rather than receiving invitations. The lower limit is now moving too. After the radiation-protection ordinance was changed in March 2026, the G-BA decided on 15 October 2026 to lower the starting age from 50 to 45. If the federal health ministry raises no objection, women aged 45–49 are expected to be able to book appointments from 1 August 2027, with personal invitations from January 2028.

Starting at 40? The US–Europe divide

In the US, the USPSTF has recommended biennial mammography from age 40 to 74 since April 2024 (grade B). Its own modelling shows the trade-off: starting at 40 rather than 50 averts about 1.3 additional breast cancer deaths per 1,000 women over a lifetime (about 1.8 for Black women, who die of breast cancer more often), but increases false positives by roughly 60% and adds about 2 overdiagnosed cancers per 1,000. The European Commission Initiative on Breast Cancer (ECIBC) suggests against screening at 40–44, suggests screening every two or three years at 45–49, recommends biennial screening at 50–69 and suggests it — preferably every three years — at 70–74.

Dense breasts

Dense breast tissue appears white on a mammogram, as tumours do, and it also raises breast cancer risk. The Dutch DENSE trial invited 40,373 women aged 50–75 with extremely dense breasts and a normal mammogram to an additional MRI scan. Interval cancers fell from 5.0 to 2.5 per 1,000 screenings. Among women who had the MRI, it found 16.5 cancers per 1,000, but also produced 79.8 false positives per 1,000. Whether this reduces deaths, or adds overdiagnosis, is not yet known. The ECIBC therefore suggests against routine supplementary MRI and leans towards tomosynthesis instead; the USPSTF calls the evidence insufficient. The German programme itself uses mammography only; the G-BA leaflet notes that it is unclear whether ultrasound or MRI offered outside it reduce breast cancer deaths.

Tomosynthesis: more cancers found — but which ones?

Digital breast tomosynthesis ("3D mammography") builds layered images of the breast. In the German TOSYMA trial, among almost 100,000 women aged 50–69, it detected 7.1 invasive cancers per 1,000 compared with 4.8 for standard digital mammography, with no increase in recalls. The open question is whether the extra cancers are dangerous ones caught earlier or additional overdiagnoses; interval cancer results are still to come. The ECIBC makes a conditional suggestion in favour of tomosynthesis, based on very low-certainty evidence.

High familial risk: a different pathway

Women with a BRCA1 or BRCA2 mutation, other relevant gene variants, or a strongly elevated risk from their family history need more than the standard programme. In Germany, the centres of the German Consortium for Hereditary Breast and Ovarian Cancer offer intensified surveillance: an annual contrast-enhanced breast MRI, digital mammography every one to two years from age 40, and ultrasound according to individual risk. If several relatives had breast or ovarian cancer, especially at a young age, ask about referral — see family history as a diagnostic tool.

What this means for you

• Aged 50–75 with average risk: taking part is a reasonable choice, and so is declining after weighing the numbers. Read the G-BA decision aid that comes with the invitation.
• Aged 45–49: the statutory option is expected from August 2027. Until then, discuss your personal risk with your gynaecologist.
• Told you have dense breasts: ask what that means for your risk and whether additional imaging would change anything — and what it would cost you in false alarms.
• Strong family history: ask for counselling at a specialised centre rather than relying on the standard programme.
• Any lump, skin change or nipple discharge needs prompt assessment, regardless of when your last mammogram was.

Women's preventive care has its own blind spots; we discuss them in the women's health gap.

This article is general information and not individual medical advice. Decisions about tests and treatment belong in a conversation with your own physician.