Explanation of our study on mastectomies in France carried out by Cancer Rose

October 17, 2017

Dr. Cécile Bour, MD

Dr. Jean Doubovetzky, MD

 Dr. Vincent Robert, MD


Our collective has conducted a study (see news release) to verify the assertion of defenders of systematic screening that, after implementation of generalized screening, there would be a decrease of surgical practices. As this postulate has not been verified in France, we are relying on the PMSI (program for the medicalization of information systems) which records hospital stays in an exhaustive manner. The findings of the article to be published in the October issue of the journal Médecine, and in open access by following this link, show that the de-escalating of surgical procedures has not occurred:

Study in Médecine/oct 2017

Or here : Researchgate

Our study on our website

Following the feedback and questions received from readers, we provide below some explanations to clarify the most frequently asked questions.



 First of all, contrary to what our opponents say, the PMSI (program for the medicalization of information systems) is completely reliable. The rating errors that could occur here and there are not significant, in fact the quotation of mastectomy procedures does not change, the surgeons know them very well. If there were errors, they would be made in both directions. It should be noted in this regard that the figures put forward in studies other than ours on surgical procedures come from the same data source that surgeons use to report their data. In France, there is no other data base for epidemiological and statistical analyses due to the lack of a national cancer registry, an information system that exists in other countries. Any “cheating” on the quotations is highly visible, leads to severe sanctions by the national insurance fund against fraudsters, and is not credible, as the figures for total mastectomy procedures, which are more remunerative, would then be at the expense of those for partial mastectomies, but this is not the case, as all procedures are on the increase. Over the last four years, an average of 19,966 total mastectomies have been performed annually, compared with 18,351 annually in the four years preceding the generalization of organized screening (2000-2003), an increase of 8.8%.

Indeed, at the same time, the number of breast cancers diagnosed each year has increased. But, even in relation to this figure, the numbers don’t add up.

In 2012, there are still practiced 4 total mastectomy procedures for 10 new cancers, as in the year 2000. And there are 15 partial mastectomy procedures for 10 new cancers compared to less than 13 in 2000.

In other words, the number of partial mastectomies is increasing faster than the incidence of invasive cancers. And the number of total mastectomies is increasing in parallel with the number of invasive breast cancers.

Under these circumstances, we should speak of a therapeutic escalation, not of a de-escalation.


Role of individual screening

Indeed, some patients have recourse to individual screening; the changes in breast cancer surgery that we have noted cannot be explained by the screening coverage that has remained stable (organized and individual screening added) in recent years.

On the other hand, while the lack of decrease in screening coverage may explain the lack of decrease in total mastectomies, it does not exonerate screening from any responsibility for the significant increase in the number of lumpectomies and partial mastectomies. This increase in the number of surgery procedures depends less on the number of women screened than on the sensitivity and specificity of the screening. With technical progress, double reading of mammograms, the switch to numerical mammography, and an improvement in the skills of radiologists, the sensitivity of mammography is steadily improving.

Therefore, it is rather the performance of screening in detecting increasingly small tumors that is responsible for the increase in the number of surgery procedures, as it leads to an increase in over-diagnosis, and consequently over-treatment.

Once again, the truth is that an unfulfilled promise was made to women, by announcing that the generalization of organized screening would result in “lighter” treatments.


On the issue of total mastectomies

Click to enlarge

L’attribut alt de cette image est vide, son nom de fichier est Diapositive2-4.jpeg.

Some are satisfied with the decrease in the ratio of total mastectomies to partial mastectomies, or the ratio of total mastectomies to “total procedures”. This apparent improvement is only due to the fact that partial mastectomies are increasing significantly compared to total mastectomies, which are also increasing, but to a lesser extent. However, this is not a good indicator of a de-escalating treatment. It would only be gratifying if the number of total mastectomies were reduced. Unfortunately, this is not the case.

The annual number of total mastectomies is not decreasing, neither the number of total mastectomies relative to the incidence of invasive cancers. How can these results be explained?

    – The re-intervention rate (partial mastectomies complemented afterwards) is only 3% and cannot account for the data; see page. 53 of the report: “Improving the quality of the healthcare system and controlling expenses: Health Insurance proposal for 2015” Report to the Minister in charge of Social Security and to the Parliament on the evolution of Health Insurance expenses and revenues for 2015 (law of August 13, 2014).

    – The recommendations requesting that conservative surgery be favored whenever possible are perhaps not followed, in such a way that the intended benefit of screening is cancelled out.

    – Or total mastectomies are performed for non-invasive tumors (notably CIS). These procedures represent an over-treatment associated with over-diagnosis. They would make lose the benefit of a general trend towards more conservative surgery for invasive cancers.

When the progression is considered without even considering the time scale, the general picture is one of an increasing trend, almost linear with a fairly high random variability.

We do not note any definite break in this linear trend and it would be impossible to locate the year when screening under invitation became generalized if the years on the x-axis were not indicated (2004). (This can be confirmed by a Davies test). Therefore, the trend in the annual number of total mastectomies has  not been modified by the generalization of organized screening. No reduction in this rate can be claimed.

concerning total mastectomies

L’attribut alt de cette image est vide, son nom de fichier est Graph-2.jpeg.

In the statistics of cancer incidence presented by INCA (French National institute of Cancer) site, only the figures concerning invasive cancers are reported, since cancers in situ (CIS) correspond to a separate entity, wrongly referred to as ‘cancer’ and not considered as ‘real’ cancer, and they are not taken into account.

It has been shown that surgery does not improve the prognosis of in situ cancers. This is why we have studied the ratio of the number of mastectomies to the incidence of invasive cancers and not invasive + cancers in situ.

The number of mastectomies (of all types) is greater than the number of new cancers. It therefore seems that in situ cancers are surgically operated on “in doubt”, not only by partial mastectomy, but also sometimes by total mastectomy.


Our observation is as follows: for every 1,000 invasive cancers, 213 more surgical operations were performed  in 2012 compared to 2000 (a).

This is an excess of 10,387 interventions compared to what the incidence of invasive cancers indicates (b).

Another method of calculation can be used: in 2012 there are 71,916 interventions compared to 53,876 in 2000. There are therefore 18,040 additional interventions in 2012 (c). Of these 18,040 additional interventions, 7,663 can be explained by a rise in the incidence of invasive cancers (d).

The remaining 18,040-7,663 = 10,377 interventions cannot be explained by the rise in invasive cancers. To the nearest rounding errors, the 10,387 given by the other method of calculation are included.

Re-interventions have a limited part to play, since they account for just 3% of mastectomies. Our hypothesis is therefore that, for the most part, these 10,377 additional procedures are attributable to over-diagnosis leading to over-treatment.

a) (ratio of total acts in 2012 year x 1000) – (ratio of total acts in 2000 year x 1000) = (1.475 x 1000) – (1.262 x 1000) = 213

b) 213 x 2012 year incidence = 213 x 48,763 = 10,387

c) 71.916 – 53.876 = 18.040

d) 2012 year incidence x ratio acts/incidence year 2000 = number of acts related to the increase in incidence between year 2000 and year 2012 = (48,763 x 1.262) – 53,876 = 7,663



There is a statistically significant decrease in the proportion of total mastectomies (p < 0.00001 in Spearman’s rank correlation test). However, this decrease in the proportion of total mastectomies is not synonymous with a lighter surgical procedure. Indeed, as shown in the graph below, the decrease in the part of total mastectomies is not due to a decrease in total mastectomies but to a greater increase in partial mastectomies than in total mastectomies.

click on the image to enlarge :

L’attribut alt de cette image est vide, son nom de fichier est Diapositive1-4.jpeg.


The increase in total mastectomies could be attributed to the increase and aging of the female population.

To test whether this hypothesis stands true, the annual number of mastectomies can be related to the annual number of new cases of breast cancer.

A mastectomy is performed because there is cancer and not because of being a woman.

In summary, two arguments allow us to claim that screening has not lead to a de-escalating in surgical procedure of breast cancer.

1. The trend towards an increase in the annual number of total mastectomies has not changed as a result of the generalization of the screening under invitation.

2. The number of total mastectomies per 1,000 new invasive breast cancers has not been decreased due to the generalization of screening under invitation.

Thanks to Dr Vincent Robert for all these analyses.


Our study was presented at the congress of the French Society of Breast Senology and Pathology, November 2017 in Lille.

Here is the presentation :  SFSPM Lille PC



diaporama SFSPM Lille PC

🛈 Nous sommes un collectif de professionnels de la santé, rassemblés en association. Nous agissons et fonctionnons sans publicité, sans conflit d’intérêt, sans subvention. Merci de soutenir notre action sur HelloAsso.
🛈 We are an French non-profit organization of health care professionals. We act our activity without advertising, conflict of interest, subsidies. Thank you to support our activity on HelloAsso.
Scroll to Top