yearly rate of change by linear regression gives the slope of the least-squares line of best fit to a graph of attribute values against elapsed years. For example, given the following case fragment:
14 Sep 08 30 Sep 08 05 Jan 09 07 Apr 09 01 Jul 09 10 Jul 09 13 Oct 09 ABC 71.2 72.09 70.67 90.29 114.26 246.20
The following condition could be added:
yearly rate of change of PSA by linear regression > 100