Re: Evidence Based Medicine

[Tuan V Nguyen <t.nguyen@garvan.unsw.edu.au>]

Hi anh Nam and friends,

	Have some time to discuss with you on this interesting 
topic.

	About 100 years ago, a French mathematician by the name 
of Pierre Louis, who had developed a technique which he 
called "numerical method" of appraisal of medical treatment. 
At the time, Claude Bernard, arguably the father of modern 
experimental medicine, opposed the concept; he went on to 
urge doctors to reject Louis' methods. He said that 
"mathematicians average too much and reason the phenomena as 
they construct in their mind." Ironically, 100 years later, 
Bernard's disciples have almost completely abandoned his 
advice. EVERY medical publication these days has statistical 
methods as a credence of research.

	I see the controversy surrounding evidence-based 
medicine (EBM) now as synonymous to the above contention. 
Histrory repeats itself!


>I see EBM just as a new scientific paradigm or just a new methodology,
>new model that has emerged in medicine science over the last decades
>basing on clinical observation (the outcome, the side effect of a
>method, a technique, a drug et.). 

	To put it in a plain language: EBM is a new SCHOOL of 
thought in the appraisal of medical treatments. About ten 
years ago, when the medical literature was (and is still now) 
bewildered with discrepancies in findings from different drug 
studies, they started thinking of a new way to combine 
evidence from various studies. This process is called META 
ANALYSIS. This is essentially a technique of synthesizing & 
weighting all available data, a systematic review of papers, 
if you like. 

	Initially, it was rejected by many people. Those who 
opposed meta analysis argued that you can not mix orange and 
apple, which is a fair argument. However, if you can evaluate 
the evidences which were derived from a homogeneous way, then 
the result of such an evaluation is worth mentioning and 
learning. It is used very extensively now. I myself have 
written and co-authored two papers on this. One in Annal of 
Internal Medicine and one in Journal of Bone & Mineral 
Research.


>It is absolutely not a new concept for
>medicine generally. Maybe I have misunderstood by your words of "old
>day" and "these day" so I thought you saw it as a new concept for
>medicine generally. 

	EBM is not a new IDEA, but it is a new WAY of appraisal 
of treatments, of thinking about medicine. It recognises the 
uncertainty of medicine and of diagnosis, which was not 
universally acceptable from a physiological viewpoint. It can 
be viewed as a way of accumulating data to update knowledge, 
however, the accumulation is done in a systematic way.

>In fact, Mechanism-based medicine still has a very
>very important role in medicine science. 

	I think you mean the physiology-based teaching of 
medicine. This is still the way medicine is taught. This is 
OK from an individual level. However, EBM is concerned with 
public health. What makes EBM controversial these days is 
that people who favour EBM want to apply EBM to individual 
patients.

	The dilemma can be put it this way. Consider the case 
where you, a doctor, sees a patient with bone loss; now you 
have two available treatments, say, oestrogen and calcitriol. 
While the former is more expensive than the latter, they 
produce the same effect (on the evidence from meta analysis). 
Which one would you prescribe to the patient? Well, from a 
EBM point of view, you SHOULD prescribe calcitriol because it 
is cheaper. Then, the question is: is that prescription wise? 
because the evidence that you based your decision on was an 
AVERAGE; that is some patients may respond to the treatment, 
some may not. So, your patient may or may not receive the 
optimal treatment. The headache part is how to make use of 
this EBM thing. The other day, anh Huy (UK) has made an 
excellent case for the use of EBM.


>The teachers in medical
>schools still need mechanism-based medicine (maybe black-white... as you
>wrote) to teach their students before mentioning to any new method like
>evidenced-based medicine. 

	Oh yes, there is nothing wrong with the physiology-
based teaching. After all, a medical student must be 
accustomed to the physiology, underlying mechanisms of the 
human body. BUT, medical students must and should learn 
natural science, i.e. physics, maths, chemistry etc as 
doctors these days spend a lot of time to interpret results 
from biochemical tests or scan figures. Once a doctor can 
understand the basic principles of these tests, he/she can 
confidently deal with patients. I have seen a lot of doctors 
(not med students) who can not read nor interpret data from a 
bone scan, can not understand the physical principle behind 
ultrasound, etc. I see this as a failure of the classical 
teaching of medicine.

	However, at certain stage, the student will be a 
doctor, he/she will have to face the dilemma that I have just 
posed earlier. You see, knowledge is power, diagnosis is a 
matter of probability. There is no way to eliminate mistakes, 
but there are ways to minimise mistakes and to live with 
intelligent decisions.

	Personally, I do not like to see EBM taught in medical 
schools now. Instead, I would like to see them to learn how 
to read a research paper first. I also would like them to 
have some basic ideas about scientific methods. You see, a 
doctor should not just be a physician, but should also be a 
researcher.


>An evidence-based medicine approach could only
>differentiate between solid evidence observed in long-time clinical
>application and evidence that is based on a presumed mechanism. 

	The problem with EBM is that it does not deal with 
mechanism or physiology. After reviewing 100 papers or more, 
it simply says Yes/No. Now, whether that answer makes sense 
or not is dependent on how WE understand the mechanism of the 
treatment. Imagine that a review of treatment A says that it 
works better in patients with X condition (which is, let us 
assume that, physiologically nonsense), then it can be either 
(i) the review or (ii) the assumption of the mechanism is 
wrong. back to square one!!!


> Of course, nowadays, EM has been elucidated partly under the light of
>modern science but it is absolutely not an experimental medicine
>originally. You can not bring " no^`i xo^ng ca?m" to a westerners, they
>will prefer and believe on Aspirin while a lots of Vietnamese will
>prefer "Xo^ng ca?m", "ca.o gio'" no need to known any its scientific
>proof, their mother said it is good, so it is good, that's all. Maybe
>evidence-based medicine here..???

	Yes, may be. :-) But this is based on PERCEPTION rather 
than EVIDENCE. The mother has a perception that it is OK, but 
there is no evidence for it. Sometimes, perception works 
better than evidence. 

	More than 2500 years ago Bhudda said that women had 
lower bone mass than men, because - he wrote - women give 
birth and lose a lot of blood. Nobody in the West pays 
attention to what he has said (because Bhudda gave no 
evidence - well, he may not have one!) That was purely based 
on perception, yet it turns out to be correct by modern 
medical evidence. However, he was wrong in the second 
reasoning, women had low bone mass because they lose 
oestrogen, not blood! So, perception sometimes works, but it 
is not always!


>I didn't know that. The "revolution" can happen at any time in any field
>of science, however, which criteria people base on to call a change as a
>revolution is difficult to known.

	It is a matter of terminology. If you have invented 
something that changes the way people think, then that can be 
said to be a revolution. What I have said (written) is that 
the American has changed the way people view medicine - from 
a black/white discipline to a probable discipline. I mean, 
doctor can be wrong, not because the doctor has not been 
trained properly, but because the doctor has not been 
equipped with sound knowledge.


>After 50 years, anything can be out of date and become wrong regardless
>with or without EBM. But I know a lots of bublications that have been
>recognised as a canon, a criterion, like the Pythagoras' theorem they
>migh be eternally true.

	Sure. Medicine as other applied science is based on 
experiments, so things that are now viewed as correct can be 
wrong in the next 20 or 50 years. That is precisely the 
reason why we need EBM.

	Make no mistake: I do not advocate for the EBM, I 
simply say that EBM must be applied to an individual with 
care and thoughtful consideration.


	Tuan