HW 8, 5d

HW 8, 5d

by Jean Digitale -

Our understanding was that survival would appear better in screened subjects than unscreened subjects with both lead time bias and overdiagnosis. This is evident in the beginning of this graph. Can you provide clarification on how to differentiate them with increasing years from diagnosis? We can't find anything in the lecture or the chapter, so feel free to point us to something we missed.

I know that survival time should not be confused with calendar time (e.g. screened and unscreened subjects could die on the same calendar date, but lead time bias could be present when looking at survival time).

Re: HW 8, 5d

by Kerstin Kolodzie -

Hi Jean,

Patients diagnosed with Pseudodisease have a especially benign course of the "disease". That means at a late stage when all other true disease patients died already, they are still alive and do well. The group only consists of Pseudodisease participants now and survival would be very similar to the survival in the general population.

Lead time bias increases the survival time by moving forward the date of diagnosis without affecting the date of death. Over time, survival in the screened group tends to run more parallel to the un-screened group because the advantage of the head start diminishes. There might be a difference between groups, but the slopes of the survival curves would be similar.

Does that help?

Kerstin