[timr]

No, this isn’t true, and the article makes the same fundamental mistake.

While certainly deaths are a more reliable indicator than diagnoses, you always expect to see an increase in “deaths from X” when you more aggressively screen for X. The intervention of cancer treatment comes with serious risks, and screening sometimes finds cancers that would otherwise never be a problem.

We’re talking about very subtle differences in population-level trends, so these kinds of errors matter.