"Junk" Science LO21690

Bill Braun (medprac@hlthsys.com)
Fri, 21 May 1999 12:51:38 -0500

Replying to LO21670 --

Winfried wrote:

>Thank you John, that's inspiring! I will take it as a case study and
>perform a brief analysis using (and learning about) cause-effect-logic
>(CEL) and the three steps of the scientific method (observation,
>speculation and falsification):
>
>TEACHING medicine uses CEL STATEMENTS:
>IF cause THEN effect !
>WORKING as a physician requires CEL QUESTIONS:
>IF effect THEN cause ?
>
>Statements are required to answer questions. If statements are not
>available, answering has to be done by means of speculation. In fact all
>statements are speculations with varying degrees of confirmation: From
>"often tested and never failed" to "not yet tested".
>
>I as a patient expect from a physician, that he has the full set of state
>of the art statements at hand, when he is doing his diagnosis business -
>answering the questions, asking me (or his diagnostic equipments including
>million $ investments available today) questions about other not mentioned
>effects (observable symptoms) to make even more sure about the causes -
>the illness.
>
>I definitely want to have as less as possible speculation in his
>diagnosis, when he starts his falsification business - the therapy. As a
>patient in therapy, I really prefer to confirm the existing system, making
>the therapy a success in curing my illness. (Although admittedly the
>epistemological value of the therapy with respect to improving the state
>of the art net of statements of cause-effect relations would be higher if
>the therapy failed. But as a patient, it is difficult to be proud of
>contributing to that, isn't it?)

If I have understood your comments, Winfried, I have some alternative
views of the dynamics under discussion. There are several aspects I'd like
to consider.

First, although implied, there was no reference to the level of
uncertainty inherent in making a clinical diagnosis. With a system as
complex as the body, the number of interactions that are taking place in
any one moment is mind boggling. There are complex interactions on an
intra-system basis and on an inter-system basis (system being circulatory,
nervous, digestive, etc.). The degree of difficulty, when presented with a
limited set of symptoms, is high. It is for this reason that physicians
begin with a differential diagnosis - a list of all the possibilities.
>From there they progress using a "rule out" line of thinking.

The rule out line of thinking permits the physician, through selective
"best guess" use of diagnostic testing, to eliminate body functions that
appear to be behaving in a normal fashion, all the while reducing the
variables down to a smaller set. The physician makes a diagnosis from what
is left.

Second, any physician will tell you however that calling the results of
their efforts a diagnosis is frequently an act of generosity. In many
cases they simply do not know. As they prescribe a treatment, they are in
fact continuing their quest for a diagnosis. If the patient responds well,
it may be good luck and the physician may still not know. If for example,
a given therapy is effective for a variety of ailments, a patient's
favorable response to the therapy does little to clarify the precise
problem.

To further complicate, approximately 70% to 80% of the therapies used by
physicians have no evidence to back them up. They have evolved over a
period of time by trial and error and represent the collective "best
guess".

If the patient does not respond, the failed therapy is similar to a
diagnostic test, and the physician can rule out many of the problems that
the therapy would have helped had the problem been one for which the
therapy is effective.

Except of course if the complex set of interactions is such that the
prescribed therapy ALONE was insufficient. With that we drift back into
the uncertainty with which the physician deals.

>Inducing illnesses (causes) form symptoms (effects) is of course only half
>of the business, namely diagnoses. Inducing cure (effect) from therapy
>(cause) the other.

In medicine, there is a fuzzy distinction between symptoms and problems.
Consider this set of relationships. I suffer an episode of chest pain
(symptom/effect), a response to Coronary Artery Disease (illness/cause).
The CAD (symptom/effect) is a response to high cholesterol (underlying
condition/cause). The high cholesterol (symptom) is a response to 1) fat
intake and 2) sedentary habits and 3) my body's metabolism (causes). My
body's metabolism (symptom) is a response to an inherited predisposition
[DNA], the cause.

This is often referred to as the natural history of a disease. I have
described it in the most cursory (and inadequate) way. What strikes me
however, is that during our upstream journey from the visible symptom
(chest pain), at about the high cholesterol point we slipped off the
simple cause and effect relationship into what I know as the
producer-product relationship. That is, it became impossible to explain
the effect as a response to one and only one cause. The interaction of
diet, exercise and body metabolism are all required to explain how I came
to have high cholesterol.

What to cautiously conclude? Could I suggest that when we use the term
cause we understand it to mean something that is both necessary and
sufficient? It seems to me that we tend to focus on the necessary part but
often let ourselves off the hook when it comes to the sufficient part.

Best regards,

Bill Braun

The Health Systems Group
- Physician Leadership Training
- Simulation Modeling for Healthcare
http://www.hlthsys.com

-- 

Bill Braun <medprac@hlthsys.com>

Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>