On Tue, 28 Feb 2006 17:02:06 -0800 Rick Froman wrote:
>
>My point in including the secondary source was not to comment on the
>original research findings but to comment on the spin put on them by the
>person quoted by the secondary source.

I admit to not not really understanding what your intent was, which
is why I provided a couple of different interpretations of what you
have intended to say.  The point was unclear.

>I am sorry if she was misquoted
>but I have nothing against her personally. I used the quote only to make
>a point. I don't remember for sure but I don't believe she had any part
>in the research being discussed. If I was taking exception to the
>results of the study itself, I should have cited the primary source.

Dr. Dimilia said that "if it worked for her" then it doesn't matter if
it doesn't work in the overall group.  One can interpret this as "spin"
or the recognition that medications have different effects in different
populations, that is, there is a medication by group interaction.  Indeed,
the combination of glocosamine and chondritin (G+C) was the only
effective treatment for people with moderate-to-severe pain from knee
osteoarthritis -- even Celebrex didn't "statistically significantly'
reduce pain (though one can argue whether a result with p< .06
should be taken as being truly non-significant without a replication
on a larger sample; by the same token, the "marginaly significant"
result of G+C in the overall group [p< .09] deserves the same
opportunity instead of being dismissed as ineffective).

>My point was that you are missing the point of nomothetic research

Actually, given that we are referring to the post you made before
I made any comment, your point couldn't be referring to something
that I would say in response to your post.

>if you say, "it worked for some and it didn't work for others". It's like
>saying if half the sample did worse on the treatment and half the sample
>did better (for an overall "no difference" outcome) that that is a great
>outcome for half of the group and a bad outcome for the other half. That
>would ignore the effects of chance which is what the nomothetic approach
>is all about.

I'm willing to give you the benefit of the doubt and not assume that
you're claiming that all human beings constitute a homogeneous population,
free of subgroups whose genetic and/or environmental experiences might
moderate their drug response.  We know that some groups of individuals
may show a response to a particular drug while other groups may show
a different pattern of response or no response at all.  Although traditional
groupings such as male-female, ethnic-racial, etc., might be one way of
thinking about these groups/population, the situation is probably much
more complicated -- the following abstract provides some indication of
this point:

Authors
Wood AJ.

Institution
Division of Clinical Pharmacology, Vanderbilt University, Nashville,
Tennessee 37232-6602, USA.

Title
Ethnic differences in drug disposition and response. [Review] [11 refs]

Source
Therapeutic Drug Monitoring. 20(5):525-6, 1998 Oct.

Abstract
Interindividual variability in drug response is a well-recognized problem
resulting in both undertreatment and overtreatment of individuals receiving
similar doses of drugs, with the potential for lack of therapeutic effect
and for drug toxicity. The potential that interethnic differences might
contribute to such interindividual variability in drug response has recently
been recognized. Such interethnic differences in drug response may be due to
either altered drug disposition or altered drug sensitivity among races at
similar drug concentrations. In turn, such racial differences in drug
disposition may be related to genetic or environmental factors, which are
often difficult to separate.

Perhaps the most extreme version of this situation is represented by the
drug "BiDil" which is targeted towards African-Americans.  The following
is an excerpt from a news article in Nature:

|The expected approval in the United States of the first drug targeted to a
|specific racial group is sparking debate about the future of 'personalized'
|medicine.
|
|Enthusiasts predict a future in which people are given genetic tests to
help
|choose the drug to which they will respond best. But some experts worry
|about the precedent of accepting race as a crude marker for underlying
|biological differences - which could still leave many individuals being
treated
|with drugs that don't work well for them.
|
|Last week, an advisory committee to the Food and Drug Administration
|(FDA) recommended that BiDil, a drug for congestive heart failure, should
|be approved for sale with a label designating African Americans as the
|target population. The FDA usually follows the advice of its experts and is
|expected to make its decision by 23 June.
[Wadman M. Racially targeted drug set for approval. Nature. Vol. 435(7045)
(pp 1008-1009), 2005. Date of Publication: 23 JUN 2005]

And:

Authors
Meadows M.

Title
FDA approves heart drug for black patients.

Source
FDA Consumer. 39(5):8-9, 2005 Sep-Oct.

Abstract
The Food and Drug Administration's approval in June 2005 of a heart failure
drug aimed at black patients marks the first time that the agency has
approved a drug for a specific racial group. When added to standard heart
failure therapy, BiDil dramatically reduces death and hospitalization in
blacks.

The key idea here, however, is that there are different groups of
individuals whose metabolism of drugs varies, probably for genetic
reasons though not always on such obvious bases as sex or racial
group membership.  This means that drug levels that seem to work
fine for one group (e.g., young white males) might not be appropriate
for other groups, either because the other group's metabolism
reduces the effects of the drug or increases its effect (i.e., producing
an overdose effect).  "Personalized drugs" of the future may be able
to take into account an individual's specific genetic and biological
status and adjust drug dosage appropriately.  The "Main Effect
of Drug" view is an out-dated perspective which has been replaced
by a "Drug-by-Population interaction" perspective".

>If 1000 people start to flip coins, after one flip approx. 500 will have
>flipped heads. On the next flip it is down to 250. Then 125, 62 or so,
>then 31 and about 15 and then about 8 then 4 then 2 then there may be
>one person who has flipped heads all 10 times. Instead of believing that
>person has the power to flip heads at will, I believe that, given that
>situation, one of those thousand people is likely to flip heads 10 times
>and it has nothing to do with their head flipping ability (supernatural
>or otherwise). By chance, some will seem to do better in any trial. What
>we need to be able to do in research is separate chance from the actual
>effect.

Golly, and here I thought that setting alpha=.05 would mean that such a
result would occur only 5% of the time on a purely chance basis.  Of
course, replication of a pattern of results, such as G+C being beneficial
for people with moderate-severe knee pain, would be directly relevant
to establishing whether G+C is in fact effective.

I won't bother to point out that your argument also holds against the
use of Celebrex which, though providing a statistically significant effect
in the total sample did not provide a statistically significant effect to
people with moderate-to-severe, a group that one might have thought
*should* have shown some if not the most benefit from a bona fide
"drug" (but watch out for those side effects)..

>Of course, there could be individual differences in response to
>a treatment and those can be isolated in idiographic research.

I'm not exactly sure I understand your usage of "idiographic" in the
context of drug evaluation.  If an anti-depressant medication is
effective in reducing depressive symptoms in women but not in men,
is that an "idiographic" result?  How about if the drug works for
middle-class white women but not for lower-SES women of color?
Is there are some definitive reference population that defines what
a "true drug response" is and variation from this standard by other
groups is "idiographic"?

>I would
>also expect, if a number of people have such an effect, to see some kind
>of noticeable outcome in nomothetic research, too.

Do you mean like what the FDA has done for BiDil, namely, targeting its
use for African-Americans and, more generally, the "personalized medicine"
approach?  Perhaps the nomothetic-idiographic framework isn't such
a good one in thinking about drug metabolism and response.

-Mike Palij
[EMAIL PROTECTED]
New York University

>Rick
>Dr. Rick Froman
>Professor of Psychology
>John Brown University



---
You are currently subscribed to tips as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to