Print

Print


Hi Susan, Paul & Amy,

I know about blood letting, Benjamin Spock, HRT, pouring hot oil on wounds
and all those examples. I want to follow up of all the medical tragedies by
telling *why* these well-meaning were these experts getting it drastically
wrong. People can't believe blood letting existed for 2500 years. *Why *were
these well-meaning  medical doctors believed in blood letting even though
we clearly know people were dying.I want to tell them these experts weren't
dumb and we are prone to the make the same mistakes as they did.

If I can give examples of what could have wrong, I think I can better
explain why we need double blind, randomization and so forth. Just by
saying randomization, double blind and 'experts were wrong' will not get
the point across. We need some story telling.

So my question is besides cognitive biases and random variation what else
could be the reason why anecdotal observations go wrong? I got one more3)
the placebo effect. Hope you understand what I am saying.

Thank you so much
Anoop


On Sat, Jun 2, 2012 at 9:09 PM, Amy Price <[log in to unmask]> wrote:

> I agree and a lot of 'treatment' is marketing and word of mouth. If it
> works it will work consistently blinded and with controls. Anecedotes and
> experience are more fun and approachable than math and objectivity. The
> only reason I went in this field was because `I wanted true pictures for
> this I will even do the math even though I am no whiz. If people are not
> willing to pay the price to do research right they need to enjoy a
> different field. The price for doing resarch wrong is lives, people and
> money
> Amy
>
> From: Paul Elias <[log in to unmask]>
> Reply-To: Paul Elias <[log in to unmask]>
> Date: Saturday, June 2, 2012 9:48 PM
> To: <[log in to unmask]>
> Subject: Re: Why an Evidence Based Approach
>
> because in many instances, experts are clearly wrong...its that simple.
> they dont know the hell of what they are talking about and making
> conclusions that policy etc are based upon and the result is often
> morbidity and mortality...so some wise guy like Guyatt and Sackett and the
> like , maybe not the only thinkers on this, but decided enough of this and
> lets formalize and standardize the approach and put some teeth into
> it...people die in the thousands due to decisions based on so called
> experts wracked with credentials yet just heresay and anecdotes...is this
> due to the treatment or intervention or due to chance? to answer this, one
> needs an evidence based controlled systematic approach. yes, and large
> sample sizes and randomization to spread around potential confoudners and
> competing explanations...there remains so much uncertainty in all forms of
> research and inquiry...so many possible explanations...so we need to
> control or mitigate the uncertainty and at the least, quantify it and
> account for it in an evidence based approach.
>
> my 2 cents...
>
> * *
> **
>
>
>
> Best,
>
> Paul E. Alexander
> * *
>
>
>
> --- On *Sat, 6/2/12, Anoop Balachandran <[log in to unmask]>* wrote:
>
>
> From: Anoop Balachandran <[log in to unmask]>
> Subject: Why an Evidence Based Approach
> To: [log in to unmask]
> Received: Saturday, June 2, 2012, 11:03 PM
>
> I am trying to figure out why we need an evidence based approach and
> medical statistics . Or how best to explain it someone who is more
> convinced by anecdotes and personal experiences. Or why did things like
> blood letting lasted for almost 2500 years.
>
> Here are some of the major reasons I think:
>
> 1. Cognitive bias and heuristics - We easily make errors because that's
> how we are designed. This could be the form  confirmation bias,
> availibility heuristic,  representative heuristic and so forth.
>
> 2. Random Variability: Random variability is inherent in every biological
> response.  A few who smoked died at an early age while a few who smoked
> lived till 80 or 90. The only way to figure if there is an actual
> difference or if this is just due to random variation is by looking at
> large sample sizes and using proper statistics.
>
> Did I miss anything important?
>
> Thank you so much!
>
>