Steve...See my comments within your text...ron

cross-over problems and analysis based on assigned to test group/
presumed to test rather that actually tested/actually did not test

Most men make decisions before understanding their disease, before
getting a complete diagnosis.  A PSA test, a positive biopsy and off
to treatment.  Unfortunate, but the real world.

Google for the Montreal analogue of the Tyrol study.  Also the
following link has some information

http://www.medscape.com/viewarticle/545957?src=mp

The men in Tyrol who did not test and the men in the rest of Austria
who did test.

Nowhere near 15 years.  What is the time midpoint of the men covered
in the Tyrol study?  1997?  When did the anaysis that you read
complete so that they could analyze the data, write a manuscript and
go through the review process?  2005?

That's 8 years elapsed.  Not enough time for most men with early-stage
disease (the majority of what testing finds) to die.  So why the large
reduction in mortality?  Opponents argue the elapsed time is too short
for PSA testing to be the key factor in the observed mortality
reduction.

Maybe some men? Probably 20-40%, I've read some estimates as high as
70-80%.  That's a big cost of doing business.

That's the problem, there is no solution available today.

Hi Ralph...My intent, when I first posted to this thread, was to make
clear that most of those who argue against PSA testing do so based on
data (data that has the same quirks and flaws as data from any other
non-controlled experiment) rather than ignorance.  When I respond to
some of your points below, keep in mind that I am trying to discover
the truth by having a dialogue.  I am not trying to defend their
position.  There is a difference.

There are two points in your message that I would like to respond to.
You wrote...

The Medscape link I posted above contains the following:
-------------------------------------------------------------------------------------------------
"it is disappointing that the authors of the current review found only
2 prospective trials of sufficient quality for evaluation. One study
randomized 46,193 men in Quebec to a prostate screening group or usual
care.[3] Screening was completed with a DRE and PSA, and the main
study outcome was prostate cancer at 11 years.

Although this trial demonstrated a 62% reduction in mortality related
to prostate cancer in the screening vs usual care groups, the current
review questions this result. The statistical analysis was not by
intention-to-screen (ie, data are evaluated for each individual
according to randomized assignment, regardless of whether he received
the screening intervention or not). The question of data analysis is a
significant issue in this trial in which only 23.6% of men randomized
to prostate cancer screening completed the screening protocol.
Moreover, 7.3% of subjects in the usual-care group received prostate
cancer screening. When the authors of the current review performed an
analysis of the study's data using the intention-to-screen principle,
the relative risk for death from prostate cancer in the screening
group relative to the usual care cohort was 1.01.

The other study evaluated in the review randomized men between the
ages of 50 and 69 in one Swedish town to prostate cancer screening or
usual care.[4] Compliance with screening was superior in this trial
compared with the Quebec study, although rates of crossover prostate
cancer screening in the usual-care group were unclear. The follow-up
period was 15 years, and data analysis was completed by the intention-
to-screen principle. While prostate cancer screening was associated
with a higher rate of finding all prostate cancer and, particularly,
localized tumors, the prostate-cancer-screening and usual-care groups
had similar rates of prostate cancer-specific survival and total
survival.

Combining data from the 2 studies resulted in a relative risk for
prostate cancer-specific death of 1.01 associated with prostate cancer
screening vs usual care. This result should make clinicians and
patients think twice about routine prostate cancer screening as
currently practiced, especially in light of what the Quebec and
Swedish studies failed to assess: quality-of-life data, costs of care,
and the overall potential harms of screening. These factors will have
to be weighed against the benefit of greater cancer detection rates at
earlier stages of prostate cancer."
-----------------------------------------------------------------------------------------------------

With only a 1.01 adjusted risk ratio for PCa mortality between the two
test groups, this analysis calls into question the claimed PCa
mortality advantage attendant screening.  The analysis leading to the
"1.01" may well be in error.  While I am unable to judge whether the
analysis is correct or incorrect,  I have come across this point in
several different papers.  I don't think that the different camps are
trying to mislead, rather I think these are honest scientific
differences of opinion around a difficult question.

The second point concerns overtreatment.  You wrote...

As I stated earlier in the thread, I have seen estimates of
overtreatment ranging from 20 to 80%.  Klotz, a highly respected
researcher has said, "At least two studies have attempted to model the
rate of diagnosing clinically insignificant disease, suggesting that
it ranges from 30% to 84%.4-5  The current incidence-to-mortality
ratio of about 7:1 suggests that the higher (84%) figure is more
likely."  One of the early papers I came across that influenced my
thinking in this area was Moul's study (Journal of Clinical Oncology,
Vol 21, No 21 (November 1), 2003: pp 4001-4008) of 300+ men who chose
AS/WW as their first method of treatment.  Moul follows these men as
they either stay with the AS/WW program or drop out to seek radical
intervention.  He uses the standard K-M statistical analysis to
project the drop-out rate over time.  He predicts that at 10 years,
20-25% of the men would remain in the AS/WW group; the rest having
gone on to seek radical treatment.  It is important to note that the
men in Moul's study had PSA <20, GS<7, TNM<T3; considerably more
lenient than the Epstein criteria, so I view the 20-25% as a floor.

Ralph, all I am saying is that because of PSA testing the pendulum has
swung from undertreatment to overtreatment.  While I don't consider
myself to be in the anti-testing camp, I feel that they ask fair
questions and bring data to the conversation that seems to support
their position.  Perhaps in a few years better controlled experiments
will tell us whether PSA testing saves lives or not.  But even then
the core problem will not be solved.  The real problem arises when we
realize that there is no happy or satisfactory middle ground available
today.  Men should not die from PCa, nor should men be overtreated.
Better testing methodology is needed.  We certainly have the means to
identify most men with PCa, but we can't always tell which of them
needs treatment...ron