Trust Govt More?

Way back in ’83, I was excited to read that medical expert systems, developed over the previous two decades of research, could often diagnosis as well as human doctors. I judged that artificial intelligence was ready for the big time, and left grad school for Silicon Valley in the hope of joining this exciting revolution.

Three decades later, we have over 10,000 times more computing power, yet medical expert systems are rarely used. Doctors say it takes just as long to enter patient info into a computer as it does to diagnose patients themselves, and medical licensing rules prevent selling such software to the public. Absent such licensing restrictions, expert systems might long ago have revolutionized medical practice.

Similar rules prevent cheaper medicine via nurses directly managing patients, even though randomized trials suggest nurses are just as effective. This all just shows what a strong lock doctors have on medicine right? Well, consider the example of school nurses.

Most states have special laws allowing school nurses to directly manage students as patients. True, school nurses can’t do everything docs can, but nurses who offered these same services to passersby at a shopping mall, without direct doc supervision, would violate medical licensing laws. Apparently, we like the comfort of knowing that medical help is onsite at school, but know that an onsite doctor would be very expensive, and so compromise with school nurses.

For soldiers, we similarly like the comfort of having medics available near each soldier, yet know that requiring medics to be full doctors would be very expensive. So we also relax our usual medical rules to let medics to care for soldiers without being doctors, or under their direct supervision. But we refuse to relax such rules elsewhere in society. Why do we allow the exceptions of school nurses and military doctors, but no other exceptions?

One obvious common element here is that most medics and school nurses are government employees. This seems to be part of a more general pattern, whereby we often relax regulations for the government. For example, the military is also not subject to OSHA rules on workplace safety, and the worst asbestos and hazardous waste sites have been on government property. Congress has also exempted itself from rules against workplace discrimination and stock insider trading.

So why are governments often held to lower standards?  Is it that they hold themselves to lower standards out of self-interest, and voters rarely notice?  Is it that voters like to punish and bring down evil alien corporations, but don’t feel as hostile to governments?

One might argue that we the public think we can trust our government more because we run it via democracy. But if so why don’t we then also trust ourselves to give government employees fair wages – why think they need unions to protect themselves from our exploiting them?

Relevant quotes:

Research conducted [before 1979] at the Stanford Medical School found MYCIN to propose an acceptable therapy in about 69% of cases, which was better than the performance of infectious disease experts who were judged using the same criteria. … MYCIN was never actually used in practice. (more)

A systematic evaluation [before 1982] of the capabilities of INTERNIST-I … on a series of 19 clinicopathological exercises (Case Records of the Massachusetts General Hospital) published in the Journal appeared qualitatively similar to that of the hospital clinicians but inferior to that of the case discussants. (more)

Licensed practical or vocational nurses (LPNs) … in 23 states [are] able to provide school health services, although some states limit their scope of practice and the need for RN or physician or chiropractor (MD) supervision depending on the nurse practice act of each state (more)

Advanced practice nurses are seeking to expand their scope of practice in at least 24 states. New legislature would legally allow nurses more freedom and authority to prescribe narcotics, and also to lead a patient care team, in states where doctors are now legally required to take the lead and supervise the team. Several states already allow prescribing and practice authority to advanced practice nurses. (more)

Added 11a: Stewart says: “psychologists have to be licensed in private practice, but in many states do not have to be licensed if they work in for a school board, prison, or hospital.”

GD Star Rating
Tagged as: , ,
Trackback URL:
  • Nathan Cook

    So why are governments often held to lower standards?

    The government’s interest in lower regulation of its activities is undiluted by any desire to raise barriers to entry, as it can directly legislate against competition.

  • Military medics are akin to civilian EMTs, and exist for the same reason: efficient provision of emergency first aid. Indeed, in their narrow specialties both are better at that type of medicine than a GP would be.

    Medicine does appear to suffer from a guilding effect, but nurse practitioners exist and are becoming more widespread, if anything.

    I’m sympathetic to your theory, but the military in general is one of those extraordinary services that even most libertarians think is a proper function of a government, and hard to either privatize or make OSHA compliant.

  • The biggest sinner seems to me to be physician associations and their lobbyists.

    Can we solve this coordination problem as a society? If we simply allowed generalist nurses to do all generalist physician work, how much would it improve healthcare efficiency?

    And where’s the fix for the specialist compensation problem?

  • AC

    As noted in follow-up letters, the study you cite has various weaknesses that prevent its findings from being generalized.
    Much more evidence than that is required to demonstrate the equal effectiveness of nurse practitioners and physicians in primary care.

  • Similar rules prevent cheaper medicine via nurses directly managing patients, even though randomized trials suggest nurses are just as effective.

    Just to be clear, the study Robin Hanson is linking here compares primary care provided by nurses as doctors. Further, the study only had enough trials to get p-values of order 5% for the most common conditions. So, in other words, nurses are as good as doctors at dealing with common conditions in family practice. (Anyone surprised?) On the other hand, if you’re worried that nurses are going to miss less common diseases, which is exactly where doctor expertise is going to be useful, then this study tells you nothing.

    Don’t get me wrong, I think people should have the choice to go to a nurse for primary care if they want to save money. But it’s certainly not obvious that the additional training of doctors isn’t worth the premium to have them see you.

  • I’m cutting and pasting this from an upcoming blog post as it is relevent…

    While individuals have the (albeit unrecognized) right to utilize any healthcare provider they choose, whether it be a physician, nurse, witch-doctor, shaman, etc., one has to feel sympathy for the residency-trained physicians being supplanted by ‘doctor nurses’. The dilution of academic titles is a triviality compared to the growing equivalence hospitals are drawing between “doctor doctors” and “doctor nurses”. The only conclusion that one can draw from comparing even low quality MD-written articles and DNP/NP-written articles is that to even call the degrees by the same name is an insult to physicians (and scientists, mathematicians, engineers, etc.). There is simply a dearth of intellectual honesty and rigor in advanced nursing degrees, which represents a disservice to medical science and patients alike.

    • I think you missed the ongoing Prof. Hanson thread where he claims that about half of all medical expenditure is wasteful, and so if a benign witch doctor is the cheapest option for that half, we’re all better off (I get that the mechanics are probably a little more complex than that, but this is the context).

  • stewart

    A point that’s missed in the ‘gov’t vs private’ issue is that in government settings, there is an assumption and role of oversight, while in private settings, the regulations assume the individual is practicing independently, without oversight, and thus their training and expertise is held to ahigher standard. That’s why, for example psychologists have to be licensed in private practice, but in many states do not have to be licensed if they work in for a school board, prison, or hospital. Ryan’s points also seem valid; expertise in a narrow field vs. high (but not necessarily expert) skills in a broader field.

  • Robert Koslover

    “…medical licensing rules prevent selling such software to the public.”

    Wow. Now that sounds like a rule/law that darn-well ought to be changed! I’m not a physician, but (to the best of my knowledge) nothing presently stops me from buying medical textbooks, except for: (1) their relatively high prices, and (2) my lack of interest in them. Why should medical diagnostic software be any different?

    Robin, perhaps if you were to lead a public revolt on this, we the people could secure our rights to purchase (presuming we can afford it) and use any/all such medical software as we see fit?

  • Vilhelm S

    I looked up the “MYCIN outperforms specialist doctors” paper (, and the result seems a little weaker than I expected it to be from that one-sentence summary.

    The way they set up the study was to go through existing patient charts, selecting a bunch of cases and preparing case summaries from them. “The summary included the history, physical examination, laboratory data, and the hospital course before the therapeutic intervention”. So this sounds like a paragraph or two of text. Then they had MYCIN and human doctors recommend treatments based on the summary, and judged the recommendations.

    I’d be a bit concerned that this sound quite different from how doctors usually do diagnoses, by interacting directly with the patients and ordering up labs as needed. (See for an amusing anecdote about “spot diagnosis”). So the study is not really showing that MYCIN outperforms the doctors in their usual working style.

  • Rudd-O

    People trust the government more, simply because they map their abused of childhood into society, with corporations as the big brother and government as the daddy who supposedly protects them from the big brother. Since in our society the vast majority of people have been raised abusively to believe that the rules daddy and mommy makes up for the children do not apply to daddy and mommy, that hypocrisy carries over to the political realm.

    Politicians know this at some level, which is why they present their violations of self ownership and property rights as “protection”. Anyone with their eyes and ears open enough can also notice this, since the entire language of the political realm is peppered with the euphemistic language used in parent child relationships.

    Root causes. The root cause is child abuse. For as long as we continue looking the other way, or attacking those who pinpoint this elementary truth, we will continue dicking around with nonanswers.

    • BZ

      Child abuse is the cause of Statism. Ergo, Libertarians come from good parents. Didn’t see that coming.

  • Vladimir Edwin Tchepes

    Interesting train of thought, maybe a warning about where things will eventually go!

    Some issues with a bearing:

    1) Much medical peer reviewed literature is wrong. (See work of Ioannidis, his PLoS articles are mathematical, but worth reading the implications are far reaching.)

    2) With a little Internet research many consumers can predict with great precision what a doctor will decide and prescribe. In other words many people may be on a par with their physician, though typically much slower. Given the wait to get an appointment they will sometimes be able to self-diagnose more quickly.

    3) Although Expert System technology is not much recognised in this era, there are parts of it that have gone a long way beyond Mycin and Internist.

    4) Iatrogenic problems (doctor caused) are common but often unrecognised. (Some research suggests that more than half of western populations have fairly broken digestive ecologies, from anti-biotics.)

    5) The Internet has taken over some functions, like banking, fairly convincingly. Some vested interests are actively prosecuting those who offer smart services on-line (i.e. Legal profession shutting down sites that outperform and are cheaper than “normal” legal services.)

    Given the above, non-emergency medicine, looks vulnerable in the fullness of time.

  • Patrick L

    If a corporation cheats, or tries to get exception to a law, or otherwise tries to avoid their duties, there are a lot of forces to stop them.

    What happens when a government official does the same thing? One case that comes to mind is the Area 51 lawsuit, where employees sued to figure out what exactly they had been poisoned with, and the government was able to side-step the whole process of responsibility by simply declaring a lack of standing. To my knowledge there isn’t a single other entity other than the United States government that could get away with that.

    A regulation regime is not just about preemptive restriction, but also enforcement of punishment after the damage has been done. Our strongest regulations came about through egregious disasters to avoid, and pointing to a person to shame. When governments increase the difficulty of accountability, it necessarily is going to have an impact on what sort of preemptive restrictions they have on themselves.

    What happens when you receive bad advice from a un-registered expert at a government facility? Who do you blame? How do you punish them? What sort of forces do you go up against?

    Predictions from this hypothesis:
    Governments that are more accountable will have tougher internal regulations and be closer to the public norm. (True)
    In small nations, large transnational corporations may have fewer restrictions, and governments might have more restrictions on themselves. (True?)
    It should be easier to sue a corporation or an individual than sue the government, and the rewards for the later should be smaller. (False?)

  • Pingback: Trust Govt More? « Daniel Smith()

  • dave

    I used to go to minute clinic all the time for colds and flues, it was a a lot easier. Then one time I went for what I thought was a flu but it was really mono combined with another really nasty hard to diagnose condition. The medicine minute clinic gave me for flu caused an allergic reaction. I ended up in the hospital for three weeks.

    Now, 99.9% of the population isn’t going to turn out to have a rare auto immune deficiency that only gets turned on by mono and they are finding out for the first time. But nonetheless getting the wrong treatment in that case was really the wrong move. About six months afterward I was diagnosed with type 1 diabetes, following in my fathers footsteps. The stress on my body and the things that happened during those three weeks may have contributed. So now I’ve got diabetes the rest of my life, in some part because I decided to go to the cheaper, quicker nurse rather then a doctor.

  • Pingback: I'm not asking, I'm driving. » School Nurses and Econ 101, by Bryan Caplan()

  • medical expert systems are rarely used. Doctors say it takes just as long to enter patient info into a computer as it does to diagnose patients themselves.

    What, and the people in charge actually *buy* this excuse? I’m ashamed to be the same species!

    Why do we allow the exceptions of school nurses and military doctors, but no other exceptions?

    Children and patching up soldiers are important. So a less cynical answer is that the truth actually *matters* here, and the people making the decisions pay for being wrong. So just like with betting, this creates an incentive structure where the deciders are a lot more willing identify wasteful medical training or requirements, and thus they make exceptions to the normal cartel rules.

  • EH

    Out of curiosity, what exactly do school nurses do? What are their powers and privileges? I was under the impression that they did basic triage (i.e. are you sick enough to go home) and dispensed medicines previously prescribed or over the counter (such as ibuprofen). Or do school nurses manage kids who’ve been sent home as well (to the best of my knowledge, they don’t). The only times I went to the nurse I (a) had my temperature measured, was diagnosed with a fever, and sent home (to my doctor’s office) and (b) told that the cut above my eye looked pretty severe and that I should go to the ER. If the school nurse had not been present, presumably one of the teachers would have had to take my temperature (and disrupting class, etc.) I wonder if school nurses might more appropriately be considered teacher extenders (freeing teachers to teach rather than making sure the sick goes home).

  • Patrick McCann

    A simple google search of ‘medical expert system open source’, without the quotes, yields many promising results. Particularly, the source code for EGADSS at and also something called Project N, at,

  • Patrick McCann

    How is different from an expert medical system?

  • Pingback: Relaxing Quality of Health Regulations()

  • Pingback: Nurses Can Practice Medicine — Only if They Work for Government | John Goodman's Health Policy Blog |

  • Phil Goetz

    “Congress has also exempted itself from rules against workplace discrimination and stock insider trading.”

    Wait – what? Details, please?

  • Pingback: Overcoming Bias : Religion Gets Bad Rap()