Index Home About Blog
From: B. Harris)
Subject: Re: Science and Polypharmacy
Date: 13 Aug 1998 11:08:58 GMT

> wrote:
>>The standard of care in "Conventional Medicine" very often involves
>>putting patients on combinations of drugs. The vast majority of these
>>combinations of drugs (where 3 or more drugs are involved) have never
>>been studied at all, let alone in double-blind trials; yet it is part of
>>the STANDARD of CARE to prescribe these multiple-drug combinations .
>>It is scientifically impossible  to predict the side effects or
>>clinical effects of a combination of drugs without studying that
>>PARTICULAR combination of drugs in TEST subjects.

    It is scientifically impossible to predict the weather a day from
now also, if by "predict" you mean with 100% acuracy.  But you can make
a pretty good probabilistic guess.

    And because drugs work by means of receptors on cells, and
generally ignore each other in function (with exceptions that are often
well known), the truth of the matter is that drugs generally do
opperate independently.   And you can predict that they will, baring
well known metabolic interactions (same P450 metabolism enzyme, for
instance, or a synergistic mechanism of action).  A high fraction of
the time you'll be right.  Occasionally, you'll be wrong.

From: B. Harris)
Subject: Re: Alternative lies about Rx drugs (was: NEJM)
Date: 11 Oct 1998 12:23:04 GMT

In <> writes:

>"Responses to the following would be of  interest :
>It is common practice in "Conventional Medicine" to put patients on
>combinations of drugs. The vast majority of these combinations of drugs
>(especially where 3 or more drugs are involved) have never been studied
>at all, let alone in double-blind trials ( with the exception of
>Oncology/AIDS treatment, where the toxicity of the drugs demands study);
>yet it is frequent practice to prescribe these multiple-drug
>It is well accepted in Pharmacology that it is scientifically impossible
>to *accurately* predict the side effects or clinical effects of a
>combination of drugs without studying that *particular* combination of
>drugs in *test* subjects.

    Actually, it isn't possible to accurately predict the effects and
side effects of any combination of drugs, even when you do study it in
test subjects.  The reason being that when you give the drugs, it isn't
to the same subjects as were tested, and people vary widely in their
reaction to drugs.

    Not being able to predict the effect of a drug with perfect
accuracy is not a new thing in pharmacology-- it is the usual thing.
The only question is whether or not the presence of another drug in a
patient induces greater variability in that patient's reaction to a new
drug, than you would get from other factors.  Factors such as whether
the patient was young or old, male or female, thin or fat, malnourished
or  well nourished, ill or well, Anglo or Asian, fed or fasting, etc,
etc.   Or even whether or not you had orange of grapefruit juice for
breakfast.  So far as I can tell, the general answer to this question
is that it does not.

   By and large, drugs work by way of protein receptors on cells, and
these are very specific (which is what allows humans to live and work
stably in so many different environments and conditions and eating so
diets so varied in content).  If a drug works on one receptor, its
effects on others will probably be small.  The general rule, therefore,
is that drugs work independently and additively.  Exceptions are drugs
with the same mechanisms of action, or action on the same physiologic
system.  The other main exception occurs in drug-drug interactions
through drug metabolism.  Both of these can be broadly predicted
beforehand, and watched for.  For example, the FDA now requires that
all new drugs be evaluated for liver metabolic degradation paths, so
that conflicting metabolism problems between drugs can be avoided
before they occur.

> Knowledge of the pharmacologic profiles of the
>individual drugs in question does not in any way
>assure *accurate* prediction of the side effects of combinations of
>those drugs, especially when they have different mechanisms of action,
>which is very common because polypharmacy is most often prescribed to
>patients with "multiple illnesses".

   On the contrary, drugs with widely varying mechanisms of action are
LEAST likely to interfere and interact with each other.

                                     Steve Harris, M.D.

Index Home About Blog