Tuesday, April 30, 2013

Medial Malpractice Claims - How Can They Be Reduced

Nicholas Bakalar (NY Times Science Section Apr 30) reports that a study published in the British Medical Journal Quality and Safety, concluded that diagnostic errors are the "most common cause of paid claims for malpractice."  Researchers reported that these claims resulted in 28.6 percent of malpractice payments, more than errors resulting from various forms of therapy, or the combined total of payments for errors in obstetrical care, medication error, monitoring mistakes, anesthesia complications combined.

I feel certain that some of these awards may have been made in cases where there, in fact, was no diagnostic error, or that the error played no significant role in the eventual outcome.  But I am also certain that there may have been other serious errors which never were brought to a lawyer's attention.  In any event, the number is what it is - and should be addressed.

Diagnosis is the result of algorithmic evaluation of data by the physician.  A physician begins his diagnostic journey with a detailed history and physical examination.  Progressing algorithmically, he then pursues additional data by requesting various diagnostic tests or procedures.  These may raise the possibility of a number of diagnostic probabilities leading to further, more involved diagnostic plans.  Eventually(and quickly one would hope) he arrives at the actual diagnosis, or the most likely diagnosis.

A computerized form of such appropriate diagnostic methodology will significantly reduce the possibility of physician diagnostic error.  Most physicians are not the "Dr. House," or some other amazing diagnostic genius, depicted of various TV dramas.  Physicians are human - they consistently make mistakes.  Development of this technology should significantly decrease diagnostic errors - especially where major diagnostic dilemmas exist.

But no technology will be able to correct improper input.  The physician who does not take an accurate history, or who does not detect an abnormal pulse, or who does not hear an abnormal heart sound, or who evaluates the size of a person's liver incorrectly will not input correct information.

 "Garbage in - Garbage out!"

Thursday, April 25, 2013

Gender Attributes - Is There a Distinction?

In "The Tangle of the Sexes" (NY Times Sunday Review, Apr 21) Carothers and Reis argue that men and women are "less different than we like to think."  The implication is that  the sexes belong to the same taxon (sort of like "species") and any noticeable difference between the sexes such as in emotional reactions, scientific abilities, social support manifestations, or intimacy are really just "dimensional" and that there are no real gender differences. We are all basically from the same species-specific gene pool and any distinctions are not sex-related.

The authors evaluated 122 attributes from more than 13,000 individuals and found that "one conclusion stood out:  instead of dividing into two groups, men and women overlapped considerably on attributes" such as those given above as examples.

But wait a minute - what have we really learned here.  We have learned that there is more overlap than there is distinction.  This, however, proves nothing about whether significant distinction exists in areas that do not overlap.  If we were to compare the attributes of rocket scientists with those of ordinary workers, I feel certain there would be considerable overlap in most of them.  It's not the overlap - it's the distinction that matters!  It's the distinction we notice - not the commonality.

Men and women are genetically distinct.  It is not reasonable to believe that the only difference this yields is one of gender.  Though the overlap of similarities is obviously huge, both physically and attitudinally, one cannot conclude that no distinctions of significance exist at the non-overlapping points on the curve.

Remember - 90% of human DNA has commonality with that of a chimpanzee.  It's the differences that that make us distinct - not the similarities.