Thread by @MichaelBaum11

Screening revisited.
For every complex problem there is an answer that is clear, simple and wrong. (H.L. Mencken, American Journalist 1880-1956)

Screening fails because it is based on a flawed assumption that the kinetics of growth are linear or log linear as illustrated here.

The logical inconsistencies in this model are multiple for example:
1. Recurrences can occur between 1-25 years with a peak hazard ratio at 2 years after surgery and a second wider peak 7-9 years.
2. The amplitude of the first peak reflects prognostic factors.

In populations screened the incidence of DCIS rises from 1-20% with no corresponding fall in "later stages". Also DCIS is often multifocal demanding a mastectomy yet multifocal invasive cancers are very rare.

The good news is that adjuvant systemic therapy has lead to close on a 50% reduction in mortality since the overview in 1985. Furthermore the fall in mortality benefits the <50s who are not invited for screening.

In contrast the over-diagnosis associated with screening mean that many patients suffer the toxic side effects of over-treatment as a result we have a zero sum game where the rare fatality from treatment cancels out the 1/2,000 cancer deaths avoided.

We can explain these logical inconsistencies by using a different mathematical model of a complex system following the rules of chaos theory. This allows for long periods of latency both for the primary and its metastases.
What kick starts them to progress is the next challenge

Finally to help you understand what this means, consider the weather forecast. The weather and the climate are chaotic systems that are very sensitive to initial conditions like the fluttering of a butterfly's wing.