Re: Snake Oil

[John Denker <jsd@MONMOUTH.COM>]

At 06:08 AM 1/10/00 -0500, Robert W. Harris wrote:
> PLACEBOS WORK

Good point.

For those of you who don't have easy access to the NYT Magazine article
Robert cited, the abstract of Bruce Moseley's scientific report can be found at
  http://www.aaos.org/wordhtml/anmeet95/sciprog/319.htm
and reads:

> Five subjects were randomly assigned to a placebo
> arthroscopy group (three puncture wounds in the skin), three
> subjects were randomly assigned to the arthroscopic lavage
> group (joint arthroscopically lavaged with fluid but no
> debridement), and two subjects were randomized to standard
> arthroscopic debridement. Patients and physicians performing
> postoperative assessment remained blinded to treatment. One
> year after surgery, placebo arthroscopy patients had less
> severe knee pain, fewer hours of knee pain, and were less
> distressed because of their knee pain than preoperatively.
> Placebo patients remained satisfied with the operation one
> year postoperatively; four out of the five patients would
> recommend the procedure to family and friends and felt that
> they were helped by the operation, and the fifth patient was
> neutral.

To continue the discussion:

1) Placebos work -- more often than you might think.  They don't work with
100% of the patients or 100% of the complaints, but neither does anything else.

2) OTOH, the goal and standard of maintstream medical research is to find
drugs that work _better than a placebo_ and obviously placebos cannot meet
this standard.

3) Prescribing or selling placebos involves a certain amount of what might
be politely called "fibbing" and what might more precisely be called
"fraud" -- taking the patient's money under false pretenses.  This creates
a moral hazard for the practitioner:  if you fib about one thing, do you
fib about another?  Where do you draw the line?

4) Prescribing or selling _expensive_ placebos creates a _huge_ moral
hazard.  It might be argued that the price-tag strengthens the
psychological effect on the patient.  Well, fine... but we need to make
sure that the effect is concentrated on the patient not on the
practitioner.  There ought to be a strict rule that beyond a nominal
charge, any revenues from the sale of placebos must be given to
charity.  (This could be considered roughly analogous to the "excess
profits tax" imposed during wartime.)

5) IMHO, placebos are a worthy topic for scientific study.  For a number of
reasons, they are not an easy topic for scientific study.  For one thing,
typical scientists are not well trained to study fibbing;  they generally
exhibit a "just say no" reaction and run away.  Furthermore, assigning a
simple name like "the placebo effect" is a step in the wrong
direction.  There are presumably many complex effects operating in
parallel.  Understanding in detail how this works would be a major advance.