Wednesday, October 26, 2005

read book now




BOOK Reviews


Discussion of Jessica Mulligan's Internet and Health report.




read book now




First, a word from me:

Many years ago, back in CP/M days when the most advanced machine I had was a luggable called Adelle the Otrona, I was invited to speak to an organization of computer using physicians that met in Vail, Colorado (alas, in summer: no snow, no skiing). One of the fellow speakers was a physician educator name, I believe, Larry Weed; I will never forget him although I may get his name wrong. He had interesting if odd ball ideas about physician education.

At the conference I saw in action a program called Tieresius: you fed it information about age, sex, medical history, and symptoms, and it put out a diagnosis. A very good diagnosis. It was right far more often than beginning residents, and held its own with experienced physicians; and if you threw in some odd disorders like tropical diseases, it often caught things that the internist didn't. It was, of course, a glorified check list, but it was built around an expert system program, and it did an impressive job.

Another program took the diagnosis and recommended treatments, and when you told it your recommend treatment it made comments and asked questions about possible side effects and synergies.

About half the physicians at the conference were enthusiastic to get the program and others like it. One young physician told me frankly that he loved internal medicine, and considered diagnosis the best part of his job, and if a machine could do his job he would contemplate suicide. I don't know if he meant it, but he may have. I did a good part of a column on this, and I ought to go find it. Alas, it would exist on 8" disks now if it exists at all, and I have no way to read 8" floppies. I discover I have in fact shut down the last machine that could read 5 1/4" for that matter. I really need to install a 5 1/4" drive in one or another of these machines. If I can find one.

I have heard little about the society I spoke to, nor about the program Tieresius; I would like to know more.

Jerry Pournelle


I just read Jessica Mulligan's piece. She mentioned a ruling in Texas regarding divorce and wills. It turns out she chose a poor instance for her example. That ruling actually a symptom of a deeper problem in Texas law.

In Texas, the divorce and will laws are hopelessly convoluted. It is literally a full-time job to be aware of all the pitfalls. Trying to simplify either process for the layman with a piece of software is simply a bad idea in this state. I haven't bothered to look it up, but it wouldn't surprise me to learn that portions of those laws reside in the Texas constitution. Our state constitution is a mess with literally hundreds of amendments.

I'm a programmer, so I have little patience for this sort of nonsense, but that's how it is.

On the subject of the article itself. I imagine doctors are often confronted with patients who notice one symptom in themselves that matches something they've heard of before. It sounded like Ms. Mulligan was being reasonable, but as she noted she ended up in a confrontation with the first doctor. The Internet reference might have made a difference, but it sounded like this doctor had a lot of experience pushing through mistaken preconceptions. He probably wasn't going to listen, anyway.

I think a good attitude to take in a situation like this is, "I'm an intelligent person and I'd like to make an informed decision about my health. Obviously you've noted some symptoms or contra-indications that I've overlooked. Please take a moment to educate me on this one very small portion of medicine." Most doctors I've met love to share their knowledge with people who are actively wanting to learn. If your doctor doesn't want to spend that time with you then you should very likely seek another doctor. Medicine is not an exact science. The more you can learn about your specific condition and what symptoms you should keep watch for, the better off you'll be. You need to be an active participant in the team that's working on your condition. They are the experts. But that doesn't mean you can't learn enough to make valuable contributions.

Drake Christensen []


Some comments on things I thought passed over too lightly.

Many people today use the emergency room as a primary care provider and as a gateway to specialists. This alone can create a standard of care issue as well as cost individuals and society more than necessary (cf. emergency room base charges versus doc'in a box). Although the article has a profoundly valid point, using a trauma center to treat a chronic condition which has reached critical stage may be the greater problem for society.

In at least one study, if I recall correctly, old style AI expert systems could match physicians diagnostic skills but after the test, the physicians improved. Compare this with Gary Kasparov's suggestion that chess with short time limits be played by players who have access to personal computers and especially their own customized analysis and data bases. There will be a shakeout and in the long run society will be better off. Of course as has been noted elsewhere, in the long run we are all dead.

Notice the Texas case ruling is more to the effect that furnishing do-it-yourself legal software is a problem than that users pro se are practicing law without a license; there is generally no license required to practice pro se.

Based on my own experience, in and out of court, there is a real problem here. The proposed solutions can nevertheless be wrong and wrong headed (and violate fundamental principles of freedom). Some sources of do-it yourself tax material for instance, libraries and post offices, mostly have signs advising the patrons no tax advice is available. There are individuals and businesses who are stretching the do-it-yourself forms from assistance in filling out the forms to advice. Do it yourself divorces are fine for people whose lifestyle is college roommates with sex privileges but there are often tax consequences and other issues many of the soon to be divorced are not even sophisticated enough to ask the software about.

This is not a new issue. In the pre-computer days there were, and may still be, people who sincerely believed that attorneys hid from the public the form books that were all that is necessary for a practice. In fact free law libraries are pretty common as well as mandatory for both the general public and convicted felons who have a right to access legal material.

For people whose lifestyle is roommates with sex privileges an attorney mediated divorce need be no more expensive than buying the software and cheaper when time is accounted for. Similarly in states which have adopted the Uniform Probate Code writing a will and passing probate can be equally simple and cheap. Hugh Dacy of How to Avoid Probate said in his introduction that in states which had the Uniform Probate Code there is no reason to avoid probate. Contrast this with eastern states where (machine politicians) the state has systems in place to use the courts and probate to extract money (some would even say graft) from the citizens the system exists to serve (check e.g. the records on a certain vice presidential candidate's spouse and probate for a description of the system in New York).

The real distinction between writing a well and estate planning should be acknowledged. Estate planning can often extend to include life style and retirement planning (or vice versa if you will) and can be well worth the cost. A simple will of the everything to my wife, and if she is not around then to the L5 society or its successors is cheap enough with or without the software, from books or from attorneys. Again it is the lifestyle planning that costs.

Finally as the AMA expands its own use of the internet (e.g. the HCFA 1500 form used for payment claims moving to electronic claim processing) for everything from continuing medical education to tracking referrals in managed care programs there will indeed be changes. Consider the medical privacy issues alone.

I suggest it is not the making of cloistered data available that has changed the distribution of rewards, but the changing nature/cost of access. At least in the legal field the information has always been available in libraries at no cost but it took costly training to access it. Just as Dr. Pournelle could once profit by knowing how to use tables and more than 3 scales on a slide rule so once knowing how to Shepardize was a paying skill. People will indeed insist on society paying for the expensive education on using a slide rule in a calculator society. Indeed remembering that the data is often good to only 3 significant places anyway is worth learning, but at what cost?

I have no useful suggestions except to say that sophomore syndrome will always be a problem and dealing with wizards means always remembering thou art crunchy and go well with ketchup.

Clark E. Myers
e-mail at:
I wouldn't Spam filter you!


Subject: Jessica Mulligan's Unpleasant Adventure

I’m a few years out from my days in the Emergency Room, but I still remember well the terror deep in the heart of every Wizard’s Apprentice. The intern in the ER can be likened to a front-line combat Second Lieutenant—a young person of some ability thrust into a situation which demands more than he can possibly give. Such insecurity can easily provoke rash, self-protective rigidity when his authority is challenged. It is wrong for the witness or victim of such behavior by an Apprentice to ascribe it to any attempt by Wizards to maintain a "knowledge monopoly."

I am surprised at times by my colleagues’ hostility to the patient who arrives in the office or hospital equipped with voluminous printouts of obscure information obtained via the Internet; they see only that such a person is likely to challenge them and require considerable extra time. I am always happy to have the benefit of whatever research has been done for me by the party most interested in a productive outcome. Rarely, however, is there much benefit. If diagnosis and treatment required no more than sifting the knowledge base, then our most able practitioners would be the third-year medical students who routinely amuse their elders by perfectly applying book learning to reach completely unrealistic conclusions.

On behalf of the entire medical community I apologize to Ms. Mulligan for whatever diagnostic and therapeutic errors were made in her management. She seems to have been victimized by an intern’s normally reliable rule of thumb that informed self-diagnosis is usually wrong. I agree with her conclusion that when the first interview went badly she should have sought another opinion before leaving. I welcome, without optimism, the prospect of an Expert System which could reduce the workload faced by the average doctor, but will oppose on ethical, not economic, grounds any attempt to encourage lay persons, less able than that intern, to diagnose and treat themselves.

-Tim Herbst


Re: Internet and Medecine

Jerry -

Excuse me? I have as much dislike of medical arrogance as anyone, but are we seriously suggesting that a half hour spent surfing the ‘net is in any way equivalent to a full medical education?

Surely any Emergency room in these wired times gets at least a couple of dozen people each day who have already "diagnosed" themselves based on either TV or the Internet. Do we really expect trained professionals to waste time and resources chasing leads that their expertise and knowledge tells them are unlikely?

Even though many doctors have far, far too high an opinion of themselves, surely most of them are smart enough that they can be allowed to let their opinions override those of the general public that appear in their examining rooms.



Barry Rueger

If you read into anything said here that I or any sane person considers surfing the Internet equivalent to a full medical education, then one of two statement is true: you don't read closely, or I and the rest of us write badly.

For myself, it has been half a lifetime since I attempted any kind of practice in psychology, and even then I restricted my attentions to bright young people not doing well in school. (I even helped a few.) But one thing I did learn: while some people have no clues as to what is wrong with them, others have a very great deal of insight. I would myself think that true in the case of medicine: I would also think that any physician ought to learn the difference between a garrulous old woman or an opinionated ass and an intelligent, educated person, who in spending half an hour contemplating her condition has spent more time than the physician in question has: and I would consider her opinion a valid input, not to be refuted out of hand without even a moment's thought. But that is my arrogant opinion.

Penultimately, NO: when it comes down to it, no. I do not accept the notion that the final decision about me is to be made by an "expert" of any kind; not until a judge and jury have deprived me of my liberty. No, I do not believe there is any expertise competent in a sole opinion and snap judgment to override my views about myself. That may be a failing, but there is it. I believe the law also has something about informed consent, so perhaps I am not entirely out of line in that view?

And finally: I am quite convinced that a physician plus a good diagnostics program like Tieresius (I believe I have that right: there was one program that took case histories and asked questions physicians might not routinely think of, and another that took the information and made diagnoses, and since it has been 20 years or so since I encountered them I may have the names wrong) -- I believe that a physician plus a good expert system program is likely to be correct more often than either alone.



Jessica Mulligan says:

Interesting discussion! As always, Chaos Manor readers turn out to be intelligent and thoughtful.

Drake Christensen has some good points, but I think he misses the mark in one place. I have no doubt Texas law is convoluted and needs simplifying; I lived in Houston for a year in the early 90s and was astounded at what the State legislature had done to the place. The importance of the ruling, however, rests in both the judge's reasoning and that fact that it was done in a Federal District court.

The reasoning by the judge was not that the convoluted nature of Texas law required human intervention, but that in his opinion the software constituted practicing law without a license. I suspect the several thousand divorces filed without an attorney's help (4,000 in Houston last year alone, by one report) had something to do with this. If this isn't blatant protection of lawyers by another lawyer, I don't know what is. And because it was done in District court, the ruling applies not only to Texas, but also to every state within that court District (assuming the judgment passes appeal in the 5th Circuit Court). This makes the ruling a precedent, which can now be cited in similar cases throughout the District's jurisdiction.

On ERs: I agree with Clark Myers' point regarding using ERs as primary care facilities. It was my situation that I recently moved here to LA and have no primary care physician yet (partially my own fault for laziness and partially extensive business travel over the past 2 months). My choices were to call around and try to make an appointment with someone, or go to the ER. Being alarmed at the color of my skin and eyeballs, I chose the ER.

His point about being crunchy food for wizards is well said, too.

It was good to read Tim Herbst's openness to patient information as a physician. I'm sure there are plenty doctors like him out there. I may have just had the (bad) luck of the draw. I do wonder, though, what kind of reaction I'd have drawn from the MD if I'd told a story such as, "You know, I was talking with my Mom this morning and she said these were exactly my Dad's symptoms the day he was diagnosed with gall stones."

Regarding Barry Reuger's points: No, I don't consider 30 minutes of data browsing the equivalent to a medical degree. However, when several professional sites consistently present me with the same two possible diagnoses for my symptom set, I take notice… and so should a doctor, I think.

Interestingly, after I was admitted to the hospital on my second visit, I was given a pamphlet that explained in detail my right to ask questions and make informed decisions on my treatment. If I had been given that pamphlet on being admitted to the ER on the first visit, I probably would have stuck to my guns with Doc #1.

Jessica Mulligan



I suppose I ought to wade in here.

Three topics for my reply:

1) What should have happened with Jessica Mulligan and why it didn’t.

2) Diagnostic programs

3) Medicine and the Internet

1) What should have happened with Jessica Mulligan and why it didn’t.

Yellow patients who are vomiting shouldn’t go home, in my view. The differential diagnosis includes (but isn’t limited to) gallstones, hepatitis (of all sorts), various and sundry malignancies, a bunch of infections and poisonings, and a few oddball congenital diseases.

Lets talk for a moment about how doctors work.

First thing we do is to take a "history of present illness", which is basically a list of questions to narrow the range of diagnostic possibilities. Ok, lady comes in yellow, vomiting, with right upper quadrant pain. No fevers. We ask a series of questions about those findings. For example, we ask if she had had pain over the past several days. Do certain types of food make it better. How long does it start after eating? Had it before, etc, etc, etc.


We then proceed to take a "past medical history". For example, is she hypertensive, does she take medications, over the counter medications, herbal medications, etc. We ask about allergies, family history, a "review of systems". We examine her, order and analyze tests and make a diagnosis.

In JM’s case and if the story she is telling is accurate, diagnosis was pretty simple. It shouldn’t have been missed.

Please note, however that this doctor business is hard to do, all of the above is horribly complex and takes many, many, many years to learn.

I suspect that she got an intern or resident who was still in the earliest stages of the learning curve. Perhaps he was uncomfortable with her self diagnosis from the internet, perhaps he was hurried, perhaps arrogant, perhaps insecure, perhaps sleep deprived, perhaps just plain stupid. In any event, he was wrong.

Some caveats however. Would be interesting to hear the other side of the encounter, no? Suppose the young MD got the following history. I’ve never had abdominal pain until today and it was very mild, I’ve had no fevers, no chills, feel well, no nausea, I felt a little sick this morning and had to vomit, but I feel great now. I don’t have any risk factors for hepatitis. I’m not taking any medications. He then examines her and finds a normal exam except for very mild jaundice. Lab tests come back with a tiny elevation in bilirubin (the stuff that causes jaundice). She has no evidence of hepatitis. She probably has gall stones in the mildest state. Lots of people do and for most of them it doesn’t bother them. Now what should he do? Hospitalize for $1500/day? Perhaps he should draw a hepatitis panel, advise her to see her more experienced family physician in the next few days to have a gallbladder ultrasound, and avoid fatty foods and encounters that could lead to the spread of hepatitis? (I’m not suggesting that is what he should have done, just trying to think the way he might have been thinking.)

2) Diagnostic programs

I use these a bit and like them. They aren’t going to replace me any time soon. The hard part isn’t making the diagnosis most of the time. The hard part is getting good information to put into the program.

3) Medicine and the internet.

This is a complex subject. There is so much junk available on the internet that finding accurate information is very hard. Putting it in context is harder still. For example, JM decided that she had gall stones not hepatitis because of the location of the pain and the absence of chills and fever. Gall stone pain and the pain caused by hepatitis can be and often are very similar. Gall stones, if infected, almost always cause chills and fever.

JM’s references are fairly good, but I could find much better information for her on the internet. Any good medicine textbook could help even more. My point, well the internet has a ways to go to be able to provide expert information.

Mark Huth []

Thank you, Dr. Huth, for a well thought out exposition.


Subject: Jessica Mulligan's Internet and Health report 11f

From: WD Lindberg []

Dear Jerry,

I work in the power and chemical process industry. For years, I was a field engineer working on power generation equipment. In that job, one of the primary tasks was diagnosing ailing machinery, electrical equipment, electrical / electronic circuits or computer programs. Along the way, I discovered I have a talent for it. It is one of those talents, where the human brain’s abilities in pattern recognition allow one to draw a conclusion from what appears to be incomplete data.

Diagnosis consists of:

1) Something wrong happens which usually presents some symptoms after a precipitating event.

2) Diagnostician reviews presented symptoms and sometimes forces the event to occur again so evolution of symptoms can be directly observed.

3) Diagnostician draws a conclusion.


4) Diagnostician does more testing and research (go to 2).

A good diagnostician will appear to go effortlessly from step 2 to step 3 (conclusion). Anyone using a rigorous diagnostic technique will be able to narrow down the list of options given enough time. Jumping to a conclusion or the shotgun technique (fixing things at random until the symptoms disappear): at best wastes time at worst it is dangerous to personnel and equipment.

Even a good diagnostician uses tools: troubleshooting guides, checklists or expert systems. A good diagnostician knows when it is too soon to draw a conclusion. Even good diagnosticians screw this up sometimes. But please note above: "Anyone with rigorous diagnostic technique will be able to narrow down the list of options given enough time". This includes research on the internet or consulting a system like Tieresius. Time is the only difference between the 2 cases. The expert can draw a correct conclusion quickly under the pressure of emergency conditions.

The human body is a complex system. It has mechanical parts, electrical systems, chemical systems and something similar to computer programs. An internist will use the same basic technique as I do. The good ones will get the right answer with little apparent effort from what appears to be incomplete data. A satisfactory one will use tests and other input to bolster and verify a hypothesis. The unsatisfactory ones cause harm to the patient.

Even the good ones when faced with ambiguous answers would review the list of applicable tests (blood tests, ultra sound, etc.) to determine if there was a short duration non-invasive test that could quickly differentiate the ambiguity. I think ultra-sound fits that description and a 2 day lab test does not.

I apply this same reasoning method when I go to the doctor. What appears to be a nave question from a "layman" can often prove very insightful (particularly when backed up with some research). A good expert listens to those questions (its hard to be that good) and while explaining the expert’s answer quite often finds the real answer.

Of course you have heard of the definition of an expert: X meaning experimental and a spurt is a drip under pressure.


WD Lindberg []



A very interesting discussion so far. I'd like to toss in my .02 worth. First a little context. I was employed in various capacities in the medical industry for a little over a decade. My immediate family is filled with doctors, nurses and hospital administrators, the topics Ms. Mulligan brings up are the subject of many dinner table conversations. Her unfortunate experience is tragically common.

I see two entirely separate issues intertwined in this discussion, and I will address them separately.

First is the issue of individual medical practice vis--vis the patient (the Arrogant Wizard dilemma). Medical practitioners, whether doctors, nurses, administrators or any other such classification, suffer from the same rates of incompetence and fallibility as the rest of us humans. The ratio of morons to masters is the same amongst doctors as it is among car salesmen. The medical field is far more complex, naturally, and it takes far more brain power to become a doctor than a car salesman. However, of all those smart enough to become doctors, few are actually smart enough to excel at being doctors. The difference is that the application of medicine is such that only the most gifted practitioners are capable of actually being effective. A medical practitioner who is only moderately competent is just not good enough, and in many cases harmful.

To it’s credit, the medical field tries very hard to weed out the moderately competent from the educational system, and later from the practice. However, the people leading these winnowing efforts are themselves human, and most of them are only moderately competent at identifying those in need of winnowing. Run this cycle over a few decades and you arrive at the frightening scenario Ms. Mulligan describes.

Second is the issue of Healthcare delivery as a system. The organization and administration of healthcare delivery is particularly complex. Like any business endeavor, it doesn’t take much incompetence or corruption to bog down the whole process into a maze of regulations and procedures whose original purpose was to prevent error and graft, but which most often result in delays and inefficiencies.

The classic example is the physician who spends an inordinate amount of time with each patient. The definition of "inordinate" is a slippery slope, because in the practice of medicine each patient is truly unique, and any attempt at obtaining a numerical average is foolish. Nevertheless, in any large hospital or HMO you will find several physicians who out of incompetence or sloth spend time with the patient beyond that which is therapeutic. Have enough time wasted by enough physicians and it will affect the bottom line. The most common solution is for administrators to simply command a time-per-patient limit. Over the decades one HMO I have experience with chiseled that down to seven minutes per patient. A sad joke making the rounds at this facility is that the only patients who get appropriate care are the ones in the morgue, because they don’t need any.

The solutions to the healthcare crisis are a tangled chaos in detail, but quite simple to describe. Get rid of the morons and clean up the corruption.

In the end, the most worrisome result of the many shortcomings of the healthcare system and its practitioners is that consumers like Ms. Mulligan even consider doing their own medical research. Whether that research is helpful or not is irrelevant, the fact that consumers no longer trust the healthcare system is the disease that needs addressing. The Internet is just a symptom.

  Mike Kelleghan


I’m enjoying your site immensely, and am a satisfied subscriber.

As a physician for almost 30 years, I wanted to briefly respond to the article by Jessica Mulligan on her treatment in an ER by a young physician, and the possible implications of the general public having access to unsifted information.

First, people have always come to physicians with their own opinions, sometimes expressed, sometimes not. And often their impression of the physician derives from how they think the physician deals with the puzzle that they bring, whether it’s a physical, mental, or emotional problem. Having the Internet doesn’t change that, it does make MUCH MORE information available to many more people. And that information is much easier to access than it used to be. But for my whole professional life I’ve had people come in with long lists of questions. I suspect the questions are raised partly to get information to help them make decisions, and partly to grade me on my knowledge base. The patients are basically shopping, and putting up trial balloons to find out if they can trust me.

Some physicians express concern about these information seekers. Those same physicians may be more comfortable with the tradition of the physician as the high priest,; giving orthodoxy that must be followed closely or face the consequences. I suspect I was trained to function in that way. But a funny thing happened along the way. I found that I could learn a great deal from my patients. In fact, some of my patients were proverbial "professors of medicine", and added greatly to my education. Initially I suspect I was a little frightened by those patients with their long "lists", but I learned that if I listened carefully and respectfully, and answered each question as clearly as possible, I usually gained the trust of the patient. And the quality of the information that patients brought me 25 years ago wasn’t much worse than what patients bring me today from their research on the net. Because there is so much freely available information, much of it requiring an advanced degree and significant experience to interpret for its application to the individual case, the availability of that information doesn’t always lead to a better informed group of patients. But it does provide a great opportunity for the physician and patient to learn to know and trust each other. And from that trust, the opportunity for a "healing" relationship in the best sense of the term may develop.

It’s difficult to know exactly what happened in Ms. Mulligan’s case. As another respondent suggested, it would be helpful to hear the physician’s side of the conversation. It may have been a combination of things: personalities, fatigue, anxiety, apprehension, misunderstanding, and pain all come to mind. Many physicians, not all of them young, have yet to come to the understanding that you can "learn a lot just by listening." Some young physicians (and older ones as well) are insecure in their knowledge base, and may become defensive when challenged by a patient with a sheaf of research in their hand and a slightly aggressive attitude. But I would submit this whole scenario has little to do with the internet, and a lot to do with the vagaries of human interaction. I have some patients who tell me they don’t want to look things up on the internet, because they don’t understand what they find, they look to me to interpret what is current and make appropriate recommendations. I have others who always bring in printouts from web sites or magazine articles to share, and we discuss what they’ve found. I have found that approaching these patients with an attitude of "let’s find out what we both can learn from this information" is a good way to defuse the situation and reassure the patient that their doctor takes their quest for information seriously.

Best wishes on getting your server/web site stabilized. I’m in the process of setting up a Linux server at home, maybe just to say that I’ve done it. <G> Your experiences are invaluable in showing things "not to do," though I’m afraid I’ve done almost everything one shouldn’t do in/on computers at one time or another. Thanks.

Jack Jacobson





birdline.gif (1428 bytes)