‘Five year survival rates can mislead’ say Gerd Gigerenzer and Odette Wegwarth in the bmj this week and go on to explain why an increase in survival from 44% to 82% is not evidence that screening saves lives, beginning with ‘lead time bias’.
(If a cancer is found by screening when there are no symptoms it is described by patients and screening enthusiasts as having been found ‘early’ and the early finding to have ‘saved my life’. But no-one can know how long it has been there. And this ‘early’ finding can simply extend the time a person is a diagnosed cancer patient; the time from diagnosis to death is longer than it would have been if the cancer had been found later when symptomatic. So, this does not show screening saves lives, it only appears to. This explanation is not as comforting or welcome as ‘screening saves lives’ so is often ignored.) “This point has been made many times before. But the message has not yet reached every doctor.”