Thursday, May 2, 2013

Degrees of Freedom




Why are Docs afraid of the technological innovations?” he asked. There was some degree of defiance in the verbiage.

    That set me thinking. Are we as physicians afraid of innovation? Are we afraid that the smart-phones now have apps that help individuals take self responsibility? My immediate answer was “No, of course not! No doctor worth his salt would be afraid. From time immemorial, doctors have embraced innovative tools to help against maladies. Why should we be afraid?”

But then, it got me thinking.

“Prediction is very difficult, especially if it is about the future.” ~ Niels Bohr



     Using a portion to describe the whole is a monumental task. It is based on many assumptions. The largest one of these assumptions is the belief that every portion is identical to the whole. But nature is not like that. Is it? Even worms exposed to similar circumstances have different genomic structural activity. So to buy into a concept that has its main pillar of support made of fudge is not too firm a foundation to bank on. Is it?

     We are progressively evolving into the linear algorithms of Boolean Logic. If this, then that mind set. We are taking a sampling of the population and deriving methodologies to render therapies. Is that an advance in medicine? After all it is as all men and women inspired by mastering public health seem to determine that what’s good for the public interest is good for the individual. And now we are relegating those human inspired tools to the linear logic of machines. Yes, yes, they claim to have the fuzzy logic and the Artificial Intelligence of a rat or a mouse, or a monkey. But do these machines have the antero-lateral prefrontal cortex? If they don't, can we invoke a verisimilitude of the same in their machine - structured software driven un-empathic hard drives? 

Can we?

     Yet the larger argument diffuses the one above easily. Even the statisticians who develop these wonderful equations to satisfy societal issues consider the possibility of outliers in the field. Due deference is accorded to these few percentages of men and women that don’t fit the mold of the many. And even looking at the classic Bell curve, one finds that 95% of the “majority” nicely fits between the two bookends of 2 standard deviations. And each deviation amounts to roughly 34% on either side of the mean. So the populace in between the first and second deviation is roughly 14% on wither side and they obviously would have variances in their makeup with some discord to the mean populace, now wouldn’t they?



(Here x is the individual data, u (mu) is the population mean and n is the total number, all this results in (sigma) the standard deviation).

      Now given this degree of error even sampling formulas have allowed for some degree of freedom in the form of (N-1) rather than (N). Which basically means that even the statisticians realize there is some error associated with the formula and they further acknowledge that the sampling error albeit small still remains even when allowing for the (N-1) in the equation.

So when someone says that we as doctors are afraid of the algorithms defined by guidelines and mandates that define the best form of therapy for a patient, aren’t they missing something? When One of One becomes One of Many, is the treatment rendered appropriate?

Let me ask a question…

     A patient aged 44 comes for a physical examination. The ensuing examination is normal. The patient asks for a PSA test, because he has heard that it detects prostate cancer and since he has never had it before, he wants to stay on top of the “game.” The results of the PSA fall in the normal range at 2.0ng/L. But then he states that some difficulty with ejaculation after sex but it has resolved. What would you tell him? “Everything is okay?” “Your PSA is normal?” “Come back in a year?”
If you did bat it off based on the guidelines and the algorithms on the computer, what would be the consequences to the patient when a year later he was discovered to have advanced prostate cancer? Was the population-based scenario treatment was none better then quackery?

     Is there then some room in the robotic devices, the computer hard drives loaded with 1 and 0s that give the doctors the degrees of freedom to practice the art of medicine on a One on One basis? Or is that passé? 

From, "e Pluribus Unum" we are morphing into "Ex Uno Plures!" 

Just some food for thought!

No comments:

Post a Comment