Joe Flower had this to say in response to criticism of comparative effectiveness research in the comment thread on his post ("Fear and Loathing Over the Stimulus Bill") examining the backlash against the Health IT provisions in the stimulus package …

"No one that I can see is saying that doctors and patients should not
be allowed to choose the best treatment. But there need to be some
bounds, some incentives to pay attention to what the best evidence
shows. Why don't doctors do bloodletting, as they routinely did 200
years ago? Why don't they whip out every kid's tonsils, as they did
when I was a kid? What happened to the idea that radical mastectomy was
the gold standard treatment for breast cancer? What about the fad for
high-dose chemo and bone marrow transplants for breast cancer, all
before studies showed little benefit and great risk? What about all the
routine things that are still done that have shown little benefit in
studies (like routine episiotomies, brain bypass surgery for patients
with warning signs of stroke, or HRT to prevent a second heart attack
in women)? What about the hundreds of thousands of spinal fusions still
being done, not for tumors or spinal fractures or congenital problems,
for which the surgery shows great benefit, but for chronic back pain,
for which repeated studies show little long-term advantage over
non-surgical techniques?

What do we do with such information? Do we just shrug our shoulders
and do nothing about it? Do we wonder whether such over-treatment with
unproven or even disproven therapies has anything to do with the fact
that we spend roughly twice as much per capita as every other major,
medically modern economy, whether socialized or mixed, for worse
outcomes, and still can't seem to afford to offer even basic care to
all Americans?"