A quick Brome Pearl rant on the perils of confusing the biostats we use when we describe  the effects of screening.


Use mortality (all cause) not 5 year survival.


C