Study Finds Breast Cancer Screening Should Begin Around Age 40

Blake Cady MD Professor Emeritus of Surgery Brown Interview with:
Blake Cady MD
Professor of Surgery (emeritus) at Harvard Medical School
Partners HealthCare, Harvard Medical
School institutions, Boston

Medical Research: What are the main findings of this study?

Dr. Cady:  Our findings support mammography screening, and our data is consistent
with the randomized trials. Breast cancer screening with mammography is the most extensively researched screening method ever studied. Only one  “randomized” trial failed to show reduced mortality, (Canadian NCSS studies),  and there were major flaws in its design and execution that negate their results, as noted in multiple critical publications (volunteers, not geographic assignment, palpable masses detected at examination assigned to “screening” arm, large contamination bias (control group got screened anyway), and very poor quality of mammography). Yet it is this NCSS study that is cited by critics and the press.  “Failure Analyses” look backward from death, rather than forward from assignment in randomized trials. The concept of failure studies is well established as noted in recent reports of air-bag failures in cars, and many industrial studies. Seat belt prevention of deaths was discovered by police recording injuries and deaths in crashes after the fact – a failure analysis – not by randomized clinical trials. In breast cancer, failure analyses have advantages of little cost, early results, simplicity, and convenience, compared to randomized trials. Since our results support findings from randomized clinical trials (RCT), they can be accepted as reliable and accurate.

Our findings show that about 71% of deaths from breast cancer occur in the  approximately 20% of our patients not in regular screening programs, while only 29% of deaths occur in the 80% of women who were regularly screened by mammography. By extrapolation, women regularly screened have only about a 5% breast cancer mortality, but women not screened have close to a 50% mortality.

(This is my extrapolation from our data, not direct data from our “Failure Analysis”)

Medical Research: Were any of the findings surprising?

Dr. Cady:  The most surprising finding was that the median age of women dying OF breast cancer was 49 years at initial breast cancer diagnosis, while women dying NOT of  breast cancer were 72 years of age at initial breast cancer diagnosis. This is consistent with competing mortalities in older women (heart, lung, other cancers, etc:) and the fact that young women are frequently not screened, and have few competing fatal illnesses.

Medical Research: What should patients and health care providers take away from this report?

Dr. Cady:  Take-away message is that screening for breast cancer should begin about the age of 40 or so and be yearly, at least in women in their 40’s and 50’s.

Interestingly, the USPSTF recommendations ignore their own published models, because of their subjective emphasis on “harms”, which are minor, while ignoring large mortality benefits. Their own models (not actual data) show a further 70% reduction in mortality by yearly screening versus their recommended screening every 2 years! Furthermore, Life-Years gained are also improved, in their own models, by more than 70%, versus their recommended every 2 years.

Medical Research: What further research do you recommend as a result of this study?

Dr. Cady:  Further research should involve more “Failure Analyses” by other institutions.

Since they involve few costs (our study was by volunteers only), and simplicity, they can be undertaken wherever good records, over a prolonged time, are available.

We did a far smaller previous “Failure Analysis” which came to similar conclusions.

It will be virtually impossible to do any more RCTs, and confirmatory data might be more easily achieved by other “Failure Analyses”; perhaps  they will disagree!


A failure analysis of invasive breast cancer: Most deaths from disease occur in women not regularly screened
Webb ML1, Cady B, Michaelson JS, Bush DM, Calvillo KZ, Kopans DB, Smith BL.
Cancer. 2014 Sep 15;120(18):2839-46. doi: 10.1002/cncr.28199. Epub 2013 Sep 9.