Bleyer and Welch begin their paper with a clear, and indisputable insight
In order for mammograms to deliver benefit, they must
1. Find more early cancers (confined to breast)
2. Leading to fewer advanced cancers (in the lymph-nodes or distant organs) among women over time
Does that make sense?
For screening to work: you gotta detect cancers that would otherwise spread, before they do. By acting upon them (surgery, radiation, chemo) you have to keep them from spreading-- that's the theory
Early goes up
Distant goes down
The answer is fall, and ideally proportionately to the rise in early cancers. Of course, distant/late cancer may not fall right away, perhaps after a lag-- corresponding to the time it would have taken them to spread.
But instead of noting a fall, Bleyer and Welch found this
Why does this matter?
Well, the goal of this entire thing is to catch and treat cancer before it spreads to distant sites, and to treat it SO IT DOESN"T SPREAD.
But Bleyer and Welch find no evidence that this has happened.
They also make this good point:
Right now you might be racking your brain for other explanations.
What if 2 trends are occurring that cancel each other out?
What if more cancer occurs for biology reasons, & screening is keeping it at bay?
Without mammograms, distant cancer would rise, you may contend
The word researchers use to describe these cancer labels is overdiagnosis-- being given a cancer diagnosis or label when, had the screening test not been performed, that label would NOT have been given in the woman's life
The author end with one of the most poignant paragraphs in the medical literature, a sober reflection on something the profession not only endorsed, but extolled;
Not just offered, but persuaded, badgered and pushed.