Re: Phuong Lien & Evidence-based medicine

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

Dear anh Huy and friends,

	Your comments remind me of an article I wrote in the 
local paper  here last year on the issue of Evidence-based 
Medicine (EBM), which I am sure you are aware of. In fact, 
yesterday, the issue is on again in the national newspaper. 

	For those of you who are not familiar with this 
concept, let me provide you a brief summary. In the old days, 
medicine was taught by experts in a black-and-white manner; 
that is, when the expert says X causes disease Y, you just 
have to listen to him/her. Full stop! These fellows were also 
very powerful in setting the research direction, and of 
course, funding. These days, medicine, just like any other 
discipline, is a matter of knowledge and probability. And, 
knowledge is power. EBM is an idea of combining all available 
knowledge (as published in medical journals) to arrive at the 
best possible treatment. The statistical process of combining 
evidences from published papers is called META ANALYSIS. 
Thus, if you want to know whether mobile phone is DEFINITELY 
associated with cancer, you can go to search for, say, 1000 
papers (some of which conclude yes, other conclude no) in 
medical literature and combine the results to arrive at the 
answer of yes/no.

	Back to anh Huy's comments. While there is no problem 
whatsoever on the use of synthesis of research evidence to 
search for the best treatment in public health, there is 
problem in applying this concept to an individual. On the 
average, the survival rate or successful rate of a treatement 
may be 80%. However it is an average; there are patients who 
do better than that and there are patients who had worse 
response.

	The complication is more evident when we consider the 
genetic make-up of an individual. It can be said that all 
diseases have, either in small or large part, genetic 
precipitation (and Diem Quynh will hit me for this). There is 
no reason to believe that the treatment will work for all 
patients in all genetic backgrounds. In fact, there have been 
evidence showing that the efficacy of treatment is dependent 
on the individual's interaction between genetic and 
environmental factors. I mean, if once keeps smoking while 
he/she has cancer of the lung, then treatment will not bring 
a good result at all.


>Dear friends,
>
>Thank you, anh Son and other friends for the diagnostics of Phuong
>Lien.
>
>Given the best current therapy, "standard-risk" childhood ALL 
>(Acute Lymphoblastic Leukaemia) patients have a >70% long term
>survival rate. The rate for the best prognostic group is around 80%.

	I guess you have to tell us how long the "survival" is? 
10 yrs or 20 yrs or more.

>I don't work in ALL, but a researcher in my dept says that 
>Phuong Lien's white cell count is favourable and she has at least an
>75% chance of long term survival.

	This is a great news for her family. 


>Disclaimer: please remember the meaning and limitation of
>statistics and don't use the above statistics irresponsibly.
>
>Huy

	Wise comment!



		Tuan Nguyen