As my urologist told me at one point, "this is not rocket science".  If
you have Excel and you know how to use it, you should be able to graph
the measurements and have it graph a least squares line approximating
the data.  You can then look at the line and estimate the doubling time
close enough for medical purposes.  If you just plot the points on
paper, you can also eyeball it and sketch in a line and estimate the
doubling time that way.

PSA(n) =  PSA(0) x 2^n
where "n" is the number of doubling times. Divide through by PSA(0),
[PSA(n)]/[PSA(0)] = 2^n
Take the log of both sides:
Log ([PSA(n)]/[PSA(0)]) = n Log (2).
solve for n,
n = [Log ( [PSA(n)]/[PSA(0)]) / [Log (2)]
and divide the actual time interval between when PSA(0) was measured and when
PSA(n) was measured, by n, to find the doubling time.
My PSA was 12 in December 2002 and 34 in August 2003, 8 months.  34/12 is
2.83333333  The log of 2 is 0.301029995 and the log of 2.833333333 is
0.45229767 Divide 0.45 by 0.30 and get n = 1.5 then divide 8 months by 1.5
getting a doubling time of 5.3 months, very bad.  My PSA is back down to 12,
very good.

 The real PSA doesn't obey the exponential  law that well- Using six readings
means that you are using a least mean squares fit to the exponential equation,
and that means that you are assuming that the measurements are in error.  If
you want to see where the rate increases changed, use a least mean squares fit
to a polynomial power series,

I have used a calculator that computes PSA doubling time for a maximum
of six psa measurements. See: http://kevin.phys.unm.edu/psa/ 
I wonder if there are online calculators that can use a higher number of
readings.

==========
to the newsgroup.......

i'll pass along the facts against the psa doubling calculator.

i used my dad's psa numbers for a three year period who was doing
watchful waiting that they had him on.  they were as follows:

every 6 months, the psa reading were taken.

1st psa - 6.0
2nd psa - 8.0
3rd psa - 12.0
4th psa - 20
5th psa - 90
6th psa - 288

after that, they put him on lupron shots

so, i plugged in the first four values - which were a two year actual
readings.  the calculator came back and said that it would take up to
the year of 2016 to reach the level of 4000.

when i plugged in all six values - the calculator came back and said
that it would reach the 4000 level by the year 2006.  

there is a far cry between 18 months from now and the year 2016.   if
someone was to place much faith in this, i think it could cause very
serious consequences, affecting their quality as well as how long they
lived based on this instrument.  

although, it had a disclaimer notice on it,  there are many variables
that i don't see take into play.  for example, and this is just one
example.  a gleason of 7.  we know that a 4 +3, is more aggressive than
a gleason of 3 +4, yet both are a gleason of 7.  there was nothing in
the calculator that would reflect even the slightest change on the scale
for this.  and there are many other examples i could come up with.  

my opinion, it may be a working model in time, but it needs refinement.

~ curtis

knowledge is power - growing old is mandatory - growing wise is optional    
"Many more men die with prostate cancer than of it. Growing old is
invariably fatal. Prostate cancer is only sometimes so."