Version française Some observations on the study :
Breast cancer mortality in 500 000 women with early invasive breast cancer in England, 1993-2015: population based observational cohort study
Taylor C., McGale P., Probert J., Broggio J., Charman J., Darby S.C., Kerr A.J. et al. - DOI:

The British Medical Journal (BMJ) recently published a new study aimed at measuring the mortality of breast cancer and its evolution between 1993 and 2015.
Although this study does not reach any conclusions on the benefits of screening, screening proponents use it to promote it. This is a deceptive use, and here's why.

Summary of the study

First, a summary of the study.
The study includes slightly more than 500,000 English women diagnosed with breast cancer limited to the breast or axillary lymph nodes (no metastases) between 1993 and 2015.
The study aims to describe the evolution of non-metastatic breast cancer mortality over time and to estimate current mortality for groups of women with common characteristics (age range, tumor grade, presence/absence of estrogen receptors, HER2 status, cancer diagnosis by screening/diagnosis outside screening, ...).

The study confirms a decline in mortality over time. The 5-year case-fatality rate decreased from 14.4% for cancers diagnosed between 1993-1999 to 4.9% between 2010 - 2015.
Other study results include the mortality of cancers diagnosed by screening and of cancers not diagnosed by screening in women aged 50 to 64. Unsurprisingly (we'll see why in the next paragraph), the mortality is higher (around 1.5 times higher) for cancers not diagnosed by screening.

A decline in mortality over time, plus a lower mortality in the group of cancers diagnosed by screening, was all it took for screening zealots to seize on the study and proclaim the benefits of screening, despite the explicit warnings in the publication.

No conclusion on screening

First of all, it should be pointed out that the study's authors themselves consider that the observed reductions in mortality cannot be attributed to screening. They write that « increases in screening cannot solely explain the decreases in breast cancer mortality that we observed » and that « this observational study cannot determine the specific causes of these reductions in mortality. »

The authors of the study are justifiably cautious for the following reasons:

1. The problem of interval cancers
The groups of cancers diagnosed by screening and cancers not diagnosed by screening do not correspond to a partition between screened and unscreened women. Cancers not diagnosed by screening include interval cancers, i.e., cancers diagnosed between 2 screening cycles in women participating in screening. These interval cancers are screening failures, and it makes no sense to assess screening performance by attributing these failures, and the associated deaths, to unscreened women.
This problem is far from negligible: in France, interval cancers account for between 12 and 18% of all cancers, depending on the source. Moreover, these interval cancers often progress rapidly, with a poor prognosis.

2. The problem of overdiagnosis
Breast cancer mortality is the ratio of deaths due to breast cancer to the number of breast cancers. This ratio can decline for 2 reasons:
- either, for an equal number of cancers, deaths have decreased,
- or the number of cancers has increased without being paralleled by an increase in deaths.
Overdiagnosis, i.e., cancers found by screening but which would never have affected health if they had remained unknown, leads to the second possibility, an artificial increase in the number of cancers by definition without an increase in deaths.
Overdiagnosis, therefore, tends mathematically to lower the mortality rate and create the illusion of screening effectiveness, whereas, in reality, screening has not improved the prognosis of "real" cancers ("real" in the sense of cancers likely to impair health).
Again, this is not a minor problem since, according to studies, overdiagnosis may account for over 40% of cancers diagnosed by screening. INCa (French National Cancer Institute) admits that 10-20% of cancers diagnosed by screening are overdiagnosed.

3. The problem of group non-comparability
The cancer group diagnosed by screening and the cancer group not diagnosed by screening were not set up by randomization, and, as a result, the distribution of risk factors for death is probably not balanced between the groups.
In many cases, failure to respond to screening invitations reflects psycho-social problems or difficulties in accessing healthcare facilities, the consequences of which are not just limited to not accepting screening invitations but are also likely to affect cancer management and prognosis.
To appreciate the importance of psycho-social problems and access to care, see graph S12 of the study's supplementary data, partially reproduced below. This graph clearly shows disparities between regions, which most likely correspond to social inequality or disparities in access to care.
graphique S12

4. The problem of screening harms
Assessing the benefits of screening is not enough. Screening harms must also be considered and weighed up:
- stress due to false alarms
- overdiagnosis, with its psychological and social consequences, and the somatic consequences of the unnecessary treatment resulting from it
- radiation-induced cancers caused by repeated mammography.
The study does not provide any information on these aspects (nor was this the aim of the study). However, only after considering all of these factors - the reduction in the mortality and the cost of obtaining this reduction - can a judgment on the value of screening be made.
And it's up to each woman, and each woman alone, to make up her mind and decide, without any pressure from outside, whether or not she wants to be screened.

Référence of the study

Taylor C., McGale P., Probert J., Broggio J., Charman J., Darby S.C., Kerr A.J. et al.
Breast cancer mortality in 500 000 women with early invasive breast cancer in England, 1993-2015: population based observational cohort study


Dernière mise à jour le 27/06/2023