Hunches vs "Hard" Facts

Steve recently said:

Likewise, a great scientist has what they call "physical intuition." He "just feels" that he’s onto the right explanation, long before he can prove it. Same thing with a great mathematician.

Here's another example to illustrate Steve's point:

Every clinician with some experience develops a sixth sense about many of his or her patients. I have countless times been at a bedside when all the "numbers" were heading in the wrong direction, when the patient's vital signs were flickering dangerously, and when statistics would tell us that the patient's chances for recovery were virtually nil. EBM would have counseled one to stop. When a physician has known a patient for some time and many statistics would suggest "throwing in the towel," the physician's hunch that this patient will make it must also, in my opinion, be thrown into the equation. It is a very uncomfortable feeling to go against what is considered "standard of care" and not to throw in the towel when overwhelming evidence points to a patient's imminent demise. But it is more uncomfortable not to do so than to see a patient come to irreversible harm.

Sometimes we are fooled or swamped by wishful thinking. But sometimes we are not, and despite all evidence the patient lives. We generally call this a "hunch," but I believe that it is no less "scientific" than are the "numbers" or the x-ray. This "hunch" of ours is made up of many small things that we cannot really dissect apart: our particular experience with that disease, our having known the patient and his/her will to live and ability to fight against adversity, our knowing a patient over the years and knowing how s/he has reacted to chemical changes (within limits not all patients react the same). We have come to know how they look when confronted with some other life crisis -- how they look when all is lost and how they look when they are still struggling with a chance to win. Persons are assuredly created equal; that is, they ought to have the same chances to pursue their interests and learn to use their talents. But just as assuredly, because individuals have different abilities and talents, we do not live or die equally.

These hunches (or intuitions, if you prefer) are highly individualistic and are not something mysterious or occult. For me, they are uncomfortable because they often truly transcend reason. They consist of our integrating a large number of facts very quickly and sometimes quite substantially -- but they are not sent to us by heaven. Rather, they are a product of our reasoning from many facts that we know and which together form what Stewart Hampshire would call a compost heap[4] -- something that can no longer be separated into its component parts but one which has formed a substance all its own. One cannot practice by hunches alone; that would obviously be insane. But one ought not to reject hunches outright and merely because they deviate from the EBM of the day.

This kind of "hunch" does not lend itself to EBM, which consists of a collection of numbers allegedly having the same implication for every member within a specific group. Not only is this inhuman, but it is also truly bad patient care and is, therefore, ethically suspect. I am not suggesting that hunches necessarily should be acted upon; what I am saying is that they must figure into the equation at least to the extent that they are positive; ie, if the "numbers" tell me that this patient is moribund and I have a hunch that this may not be so, following that hunch (even when wrong) is unlikely to harm the patient. Unfortunately, what lurks behind much of EBM and rigid protocols has more to do with material profit and loss than too many would like to admit.

The obverse is very likely to occur also. If we are to use EBM humanely and efficiently, we would be compelled, I think, to allow physicians within reason to "play their hunches." Physicians who operate according to their hunches most of the time obviously (and with good reason) should be suspect -- suspect, not guilty -- of practicing poor medicine until the individual circumstances are investigated. Statistics are very helpful and sometimes critical to our understanding of a case. But we should never forget that statistics apply to groups of people and not to unique, identified lives -- namely, to a particular patient in the sickbed.