Biomarkers

[davis at wehi.edu.au]

Dan McLeod (dmcleod at hpb.hwc.ca) wrote:
>I'm brainstorming for ideas.  In the future, I anticipate making
>recommendations re: measuring biomarkers in blood samples from a large
>cross section of the normal population.

Dan, I'm not sure who you are or who you are working for, but this strikes 
me as a major public health endeavour, and as such is very very expensive.
You need the following criteria to be met for this to be effective:

1. Your marker must be sensitive and specific.  If you don't mind missing a
few cases then sensitivity is less important.  If you don't mind scaring a
few people needlessly then specificity is less important.

2. The disease that your marker is designed to detect must be treatable.
There is no point making a diagnosis in a screened population if there are 
no therapeutic options available.  Also, the treatment or intervention must
be demonstrably effective.  The controversy over treating people with 
higher-than-population-mean serum cholesterol levels is a case in point.

3. It must be cheap and quick.  Self-explanatory.  If it costs a million 
dollars to save one life by screening, you must ask if this money could be
better spent in other ways.

Unless you are part of a large government-funded organisation interested
in public health, I can't imagine why you would want to get involved in
something like this!

Ian Davis					davis at licre.ludwig.edu.au
Ludwig Institute for Cancer Research
Melbourne Tumour Biology Branch
Melbourne, Australia.