How is the Seer Data on PC developed?

Since the post on PC incidence and deaths created some discussions, it got me thinking of how the numbers are collected. If anyone know the exact method of collection please let us know.

Deaths are pretty easy as all death certificates with cause of death are recorded; there may be some errors as to the cause of deaths, but I assume the numbers are fairly accurate.

The newly Dxed cases are where I'm questioning the accuracy.
How does this get recorded? does the doctor that does the DX actually send something in? Are the numbers gathered from pathologists who read the slides or doctors that do treatments? How do upgrades and downgrades in Gleason score get recorded without double counting. Do second opinions get recorded?
Or is the number of new incidences developed from a sampling of some sort.

We use these numbers to measure if our screening methods and treatments are working, but we never question the method of collection and just assume that they are a reflection of the real world.

JT

JohnT,
On the info on the graphic representation of cases and deaths it, SEER, stated that the current number is based on cases and deaths from 2007-2011 so it would seem that they go a bit into the past. I'll try and read the full report, under SEER Prostate Cancer, to see if I can find out more . . . Q