Let's assume there were 150 "cancer treatment breakthrough" stories in that time (about "every other month"). Then it is a reasonable guess that any "cancer tre

Let's assume there were 150 "cancer treatment breakthrough" stories in that time (about "every other month"). Then it is a reasonable guess that any "cancer treatment breakthrough" improves survival rate by 0.1 percent points, far below the effect size that could be reliably demonstrated in an academic study. Which is great, but not what the layman reader would take from such a headline. "Breakthrough" implies "you have to update your mental model" - but no layman has to think differently about cancer treatment after reading this story.