Looking for the Evidence in Medicine

An ambitious collaboration aims to review every trial of a medical 
treatment ever done to find out what works, what doesn't and where 
future research should be focused


    a gold mine - 1million randomized control trials of medical 
treatments that have been conducted over the last 50 years.  The bulk 
of scientific evidence on the ultimate value of treatments and 
interventions have been either forgotten or simply lost.  Drummond 
Rennie at UCSF (U of Cal, San Fran) are sifting through all that 
forgotten lore to recover its full value.

    Known as the Cochrane collaboration, after the last british 
epidemiologist archie cochrane the 2 year old collaboration is literally 
hand searching the world's literature to find and review all of the 
randomized control trials ever published together with whatever 
unpublished trials can be located - and publishing thte findings in an 
electronic form.  Updated regularly, these reviews should synthesize the 
latest state of knowledge about every available therapy or intervention
and give its implications for practice and research.


    The several thousand volunteers, coordinated by 10 centers who make up
the Cochrane collaboration expect their efforts to transform medical 
practice.  Unlike drugs and medical devices, which can't be sold unless 
their effectiveness has been documented, actual treatments are often 
dictated by authority, personal exerience and habit.  The cochrane 
researchers hope to bring in a new force for clinical decisionmaking : 
what oxford university epidemiologist David Sackett has described as 
"the conscientious,explicit, and judicious use4 of current best evidence."



    The collaboration shares that aim with a larger movement known as 
evidence-based medicine.  "The Cochrane Collaboration is an enterprise 
that rivals the Human Genome project in its potential implications for 
modern medicine" wrote David Naylor CEO of University of Toronto's 
Institute for clinical evaluative science, in "The Lancet" 4/95.


    "What the collaboration is trying to achieve is 100% laudable" says
U of Chicago epidemiologist john Bailar but it may end up with a lot of 
apparently strong,, precise results that are nowhere near as strong or 
precise as they appear to be.  Lapses of quality control or judgment could
easily slip into the enormous review process and they also question the 
underlying methodology, which relies heavily on a controversial 
statistical technique called meta analysis.

    Naylor adds that even if the reviews are reliable, medicine can't 
always be done by cookbook.  Good clinical practice will always involve 
some elements of inference and judgment leavened by experience.  They goal:
simply trying to help doctors do what they have always done: assess the
available evidence. "Its not that physicians have not been trying to be 
evidence-based in large measure " says donald Berwick CEO of Boston-based 
institute for health care improvement "It's just that the base of evidence
has gotten more and more complicated and now requires some other mechanism
to digest it"


<<Mining the literature>>
The Cochrane collaboration is offspring of project by oxford physician 
Iain Chalmers 1976 : a systematic review of all randomized clinical trials
having to do with pregnancy and childbirth.  Between 1978 and 1984 Chalmers
and Murray Enkin they organized a search of 70 journals dating back to 
1950 to identify all relevant trials in the area.  Chalmers wrote to 42000
obstetricians and pediatricians requesting information on any trials or 
data that may never have been published.


    The search turned up 3500 trials, out of which the researchers
selected for review only those that met specific criteria .  Trials only 
qualified if they used a formal randomization procedure or quasi-random 
method for dividing patients between the tratment group and control.


    The results published in 1989 "Effective care in pregnancy and 
childbirth included a number of eye-openers.  e.g. a clear message about
the value of giving corticosteroids to women about to give birth 
prematurely, to speed up maturation of the fetus.  The inexpensive simple 
treatment halved infant mortality.  "Effective Care" has constituted a 
sea change for us clinically.  Now all available on computer.


    Each center's primary purpose is to marshal volunteer review groups, 
which focus on specific conditions stroke or schizophrenia.  Reviewers 
each is responsible for studying a particular treatment or intervention, 
are committing themselves to a grueling task, and they sign on for life, 
says Chalmers.  "we spend a lot of time trying to dissuade people from 
getting involved, so that if they do they know with absolute certainty 
what they're getting into."  Those who attend workshops learn how to 
develop a review protocol a set of questions to be answered and an 
approach to answering them. the collaboration software forces you to take 
a systematic approache to doing the review."

   The favorable reception prompted UK's National Health Service (NHS) in 
1992 to fund a cochrane center, a year later, the NHS inititated the 
Cochrane collaboration - supported by centers in oxford; lyon france;
copenhagen denmark; amsterdam netherlands, milan italy baltimore; san 
antonio, san francisco adelaide, australia,, and hamilton ontario

    The review groups amass material for their reviews by hand searching 
all the journals in the world likely to publish randomized control trials 
on their particular subjects or having the journals perform the searches 
themselves.  The collaborators are trained to dig up unpublished data by 
every possible means. including pressuring pharmaceutical companies to 
release results on studies that they may have preferred to keep quiet


    In the past 2 years, nearly 100 systematic reviews have been completed,
covering conditions including stroke,, sbufertility, scchizophrenia, and 
parasitic diseases.  Completed reviews are being published in the Cochrane
database of systematic reviews available on CDROM by year end - on WWW,  
The Cochrane volunteers will watch the literature, updating the database 
and revising the original review when necessary - indeed, that's why the 
volunteers sign on for life.


   While some of the reviews give no more than a qualitative conclusion, 
others will go a step further and do what's called a meta analyasis when
the literature search yields enough sufficiently similar studies.  In 
meta-analysis, a single quantitative conclusion is extracted 
by synthesizing the data of many different trials.

You get the data from the original authors , go over the data in detail, 
do a lot of to and fro getting rid of errors, then you get them involved 
in the interpretation both of their own study and the overview.


   At the very least, both critics and enthusiasts of the Cochrane 
collaboration say the reviews should provice important information about 
what isn't known in medicine .  Or, as Naylor puts it, " The wonderful 
thing about the systematic approach of reviewing available evidence across
every imaginable aspect of clinical practice is that it will, one hopes, 
nail down those areas where we just don't know what we're doing" whether 
blood thinning or clot busting drugs are beneficial in acute strokes.
    /*---------------------------------------------------------------------*/

Tilting aginst a Major theory of movement control.

Hiroaki Gomi and Mitsui Kawato at ATR Human information processing
research labs in Kyoto, Japan.  Use a semicircular steel apparatus with 
complex system of levers and shafts supporting an armrest to test the 
reigning theory about how the brain controls voluntary movements of the 
arm and other limbs.

The theory known as "The equilibrium point hypothesis" 

 implies the brain does not compute all the forces required to move the 
limb from one place to another, but simply launches the limb, depending
on reflexes and the intrinsic elasticity of the muscles to get it to its 
destination.  That's possible, the theory holds, because that intrinsic 
springiness makes opposing muscles seek an equilibrium , or balance 
whenever the brain perturbs the system by setting the limb in motion.  
Evidence shows the brain is in fact in control throughout a movement.
doing all the calculation necessary to figure out what muscles to move 
and when.


    /*---------------------------------------------------------------------*/

Physicists trap photons and count them one by one.

     []  ..........   []   .   .   .   .   . (   )   .   .  [ ]
    Oven          Excitation                  /\       Detector
                  Source                   photon    
                                            source

counting quanta
Atoms in an excited state, file through a microwave cavity to detect
the number of photons.
    /*---------------------------------------------------------------------*/
Taxonomy helper

biodiversity experts have talked up a storm about cataloging all 1.5 
million of the world's known species in a single internet database.
Now a plucky group of US government biologists appears to be making a 
start at least for north american species.  Their interagency taxonomy 
information system (ITIS) which will go fully on-line this month, aims 
to put all the tens of thousands of known north american flora nad
fauna on the WWW.


    The project will bring order to the sometimes unsystematic world of 
systematics.  To southerners, a catfish is a fresh water fish, while in 
new englance the term applied to certain marine fish.  They're crying out 
for stabilityin nomenclature.

    Created by the EPA et al, ITIS is starting out with a data set of 
128000 species from the national oceanographic data center that is being 
checked for quality by taxonomists.  To handle inconsistencies in 
nomenclature, ITIS will register a species under the various names by
which it is known, but when given a query will point to user toward the
most widely accepted scientific name, reflecting the latest taxonomic 
consensus.  It will provide references and the anmes of experts on that
species.


    Amphibian expert david wake at U of Cal,, berkeley, agrees that such 
a resource will be a boon if it does, indeed, prove to be sufficiently 
malleable to reflect differenes of opinion.  The ITIS URL is 
<a href="http://www.itis.usda.gov/itis/>itis url</a>

    /*---------------------------------------------------------------------*/
p51 Table of CRP, SAP, MBP,
LBP, sCD14,C3...

    /*---------------------------------------------------------------------*/
Imprint lithography with 25nanometer resolution

SY CHou

A high throughput lithographic method with 25nanometer resolution and 
smooth vertical sidewalls is proposed and demonstrated.  The technique 
uses compression molding to create a thickness contrast pattern in a thin 
resist film carried on a substrate, followed by anisotropic etching to 
transfer the patteern through the entire resis thickness.  Metal patterns 
with a feature size of 25nanometers and a period of 70nanometers were 
fabricated with the use of resist templates created by imprint lithography
in combination with a lift-off process. With further development, 
imprint lithography should allow fabriation of sub-10-nanometer structures
and may become a commecially viable technique for manufacturing integrated
circuits and other nanodevices.

    /*---------------------------------------------------------------------*/