NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/sci/med/23072

Internet Message Format | 1992-12-25 | 8.8 KB

Xref: sparky sci.med:23072 talk.politics.medicine:442
Newsgroups: sci.med,talk.politics.medicine
Path: sparky!uunet!spool.mu.edu!news.nd.edu!mentor.cc.purdue.edu!pop.stat.purdue.edu!hrubin
From: [e-mail] (Herman Rubin)
Subject: Re: Reliability in medicine and engineering  (was: ... incomes)
Message-ID: <[e-mail]>
Sender: [e-mail] (USENET News)
Organization: Purdue University Statistics Department
References: <[e-mail]>
Date: Fri, 25 Dec 1992 17:27:09 GMT
Lines: 146

In article <[e-mail]> [e-mail] (David Nye) writes:
>[reply to [e-mail]]
>
>>>The big question of how to cut costs without cutting quality depends on
>>>what you mean by quality.  If we continue to insist that quality means
>>>unlimited access when we want it to all health care resources, then it
>>>can't be done.  If instead we decide that we really just want reasonable
>>>care for everyone, we can have it for much less than we currently spend.
>>>It will mean that some die who could have been saved.  It will mean
>>>rationing, such as no CPR for patients over 60 in an out-of-hospital
>>>arrest, (but less than 5% of those patients make it back, and at an
>>>exorbitant cost).  It will mean either eliminating medical malpractice
>>>or never finding against the physician when he follows standard approved
>>>algorithms for diagnosis and treatment (since nothing short of these
>>>will stop physicians from practicing defensive medicine).  It will mean
>>>that the insurance and legal industries will be cut out, which will put
>>>many lawyers and insurance industry people out of work.  It will mean
>>>that some physicians will lose their houses, and fewer qualified people
>>>will be attracted to medicine.

>>This is a typical totalitarian bureaucratic attitude toward what is
>>reasonable, and it means poor care and little progress in any situation
>>which requires thinking and innovation.  If the rationers decide that
>>only 2% of our national income can be spent on medicine, that is what
>>will happen.

>I'm not saying that I like any of this.  As a doctor, I would much
>prefer to practice medicine the way it is, if only someone would pay for
>all the patients who currently can't afford insurance.  This of course
>will never happen.  I think the changes I have outlined are unavoidable
>and that the current system is doomed to collapse under its own weight
>in a few years.  The US being what it is, I'm sure it will always be
>possible for those with money to buy premium care.  For the rest, I
>think the quality of medicine will suffer to some extent as it does
>already under Medicare and Medicaid, but at least it will be accessible
>to all.

A major part of the problem is the "insurance" now being peddled.  It
is totally against the idea of insurance; insurance should not, and
normally does not, cover expected costs.  But the real problem is the
proliferation of the desire for medical care in a proliferating population,
where many of the members have little or no financial responsibility.
We cannot avoid the tragedy of the commons except by rationing, with
all its undesirable aspects.

There is a realistic limit to charity.  We do not have an absolute moral
obligation to provide reasonable medical care for every American any
more than we have such an obligation to provide it for every other
human being, and that is clearly out of the question.  Having the
government do it is worse; at least with private charity the giver
has some control over where the money goes.  I do not consider an
elected government as giving any measure of that kind of control.

Also, it is by no means clear that anyone except the extremely rich
will have access to good medicine.  A good example of what may very
well happen is our miseducational system.  When we were at one of the
largest universities, the people at the local school system (they were
very unusual in even considering this) recommended that our son go to
an academic private school.  Suffice it to say that there was none
which would not have required relocating.  Right now, very few are
getting a reasonable education; this is one of the reasons why the
medical students cannot think.

>I share your concern about the loss of thinkers in medicine.  The
>average GPA of applicants to medical school has fallen almost every year
>for over a decade.  This of course doesn't mean that we aren't getting
>many good people, but some that would have been great physicians are
>choosing other careers.  It could get a lot worse, depending on what
>happens politically.  Add to this the increasing complexity of medicine.

As long as doctors are paid for 'procedures', it will get worse.  The
idea that a doctor in a given specialty is entitled to X dollars for
seeing a patient in a "routine" office visit, no matter how good, must
be abandoned.  It is this, as much as anything else, which drives out
the good people.  Medical pay should vary as much as the pay of
big-league baseball players, at least.

Forget the GPA; it measures nothing useful.  What is worse is that the
curriculum on which that GPA is based has eliminated most of the thinking
which used to be there a half century ago, replaced by rote and routine.
It is doubtful that most university graduates with high GPAs can think.

>The only way I can see for doctors of the future to be able to cope with
>all this information is with computers.  The medical record will be
>entirely electronic before long.  It already may be cheaper than paper.
>Once the history and exam are in a computer, a decision support system
>will use natural language processing techniques to extract the pertinent
>clinical information (or perhaps the traditional record will be replaced
>with a database format).  The computer will then present the doctor with
>a differential diagnosis and suggestions for further workup and
>treatment generated from those algorithms I mentioned earlier.

At this point, the non-thinking doctor is no longer needed.  But the
computers cannot think; they can only do what has been programmed into
them.  We MAY eventually have some sort of non-routine intelligence
available for computers, but it will not come from the type of work
going on now.  Predictability is antithetic to thinking.  But it is
correct to say that the routine diagnosis, etc., will in the not too
distant future be done more and more by machines.  It will be the job
of the human doctor to decide when to ignore the machine; those who
just follow the machine are the ones not needed, except for those
operations where human skill is still needed, and except as handholders.
Neither of these needs a full medical knowledge.  A single word of
caution is not enough; people have too much of a tendency to believe
anything which comes out of a computer, and the physicians I have known
do not seem to be exceptions.

                                All of
>the elements of this system are already developed, although natural
>language processing techmology still needs more work.  The Internist
>project at U. Pitt. with which Gordon has been involved has been
>downsized and is available to run on a PC or Mac (Randy Miller's QMR).
>It does a surprisingly good job on a limited domain (internal medicine).

I am quite suspicious of this.  Part of my suspicion is based on a talk
given by an MD at a symposium on influence diagrams, in which exactly
this problem was discussed.  I doubt that enough is known to handle
situations in which a half-dozen or more simulataneous conditions are
not unusual, and where there are over 500 recognized diseases which
usually do occur in combination.

>The reason everyone isn't using it yet is because of the time that it
>takes to enter the clinical information into the computer.  If it were
>already in the computer and the expert system's advice were available
>for free, everyone would use it.

The official definition of "expert" is anyone who has more knowledge than
the "person in the street."  I believe that a paramedic, or even someone
who has had a pre-med course, might qualify.

>I suspect the quality of medicine as practiced by the average
>practitioner would be markedly better under such a system.  perhaps it
>would even be reasonable to have nurses provide all community-based
>primary care.  Most physicians would be involved in research, sorting
>out the information needed to update the expert system and doing basic
>research.

There is another category needed, which I keep harping about, and which
is more and more ignored; the thinker, the one who can see that the
computer program, or conventional wisdom, has ignored much that is
relevant, and that a "crazy", or at least unusual, diagnosis is needed.
--
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
Phone: [phone]
[e-mail] (Internet, bitnet)
{purdue,pur-ee}!snap.stat!hrubin(UUCP)