Are Decision Support Tools Turning Doctors into Idiots?

A HealthLeaders article by Gienna Shaw notes that some physicians are reluctant to use computerized decision support (CDS) tools because they fear losing the respect of patients and colleagues. There’s some evidence to support this concern:

In one [study], even tech-savvy undergraduate and graduate computer science students preferred physicians who rely on intuition instead of computer aids.

“Patients object when they ask their doctor a question and then she or he immediately types in the question into their laptop and then reads back the answer. It gives patients the feeling that they just paid a $25 copay to have someone Google something for them,” [study author James] Wolf says.

Shaw argues that this is a transient phenomenon in any case because soon everyone will use CDS as payers demand it and the tools get built in to electronic medical records in a way that’s invisible to patients. She’s probably right, but she’s sparked some interesting thoughts.

I prefer physicians who uses sophisticated decision support tools such as SimulConsult, which allows physicians to extend their expert knowledge to make differential diagnoses of rare conditions that even excellent, experienced specialists may see rarely in the course of a career. Doctors are trained to see horses, but there are a lot of zebras out there that are being missed as a result.

Other point of care information tools, such as UpToDate are terrific for keeping current with the latest knowledge. I’m always happy for a doctor to use UpToDate to confirm what he already knows or to find out if there are new developments.

I’ve been favorably disposed toward computerized decision support ever since a summer job at a VA hospital 30 years ago where I programmed an early tool in MUMPS on a DEC PDP 11/34. Yet I can understand where Wolf’s skeptical computer science students are coming from based on my own experience with computerized tools. I’m not a doctor so I can’t relate these directly to medicine, but here are a couple of my own observations for what they’re worth:

  • I love using my GPS when visiting new places and to find shortcuts. It’s great because it lets me explore new places I would have hesitated to drive to in the past. But excessive reliance on the GPS may have dulled my map reading ability, sense of direction, and ability to learn new routes. I’ll admit that I sometimes end up taking a less optimal route just because the GPS suggests it.
  • In the olden days doing research required some thought as to the best way to frame the question, what data sources to pursue and the most promising, efficient way to find information. Now the easiest thing to do is usually to type whatever the question is in to Google and see what pops up first. I still know how to go well beyond that but it’s a skill that seems to be eroding.

These tools tend to level the playing field, bring up low performers and reduce costs. The best professionals –in health care or any field– are the ones who can build from the easily available knowledge and add something differentiated on top.

There may be some doctors who really are just typing questions in to Google and reading out the results –although even there they are likely adding value by drawing on their training and experience. The good news is that routine tasks can now be performed by less expensive people or machines. The problems that remain are tougher, which gives ample opportunity for medical experts to earn their keep by applying human judgment that takes into account all that the decision support system can tell them –and then goes beyond.

David E. Williams is co-founder of MedPharma Partners LLC, strategy consultant in technology enabled health care services, pharma,  biotech, and medical devices. Formerly with BCG and LEK. He writes regularly at Health Business Blog, where this post first appeared.

Spread the love

39 replies »

  1. Family med is the specialty with substantial breadth and limited depth.

    As a family doc at an Indian Health Service clinic with limited opportunities to refer I’m up to my ears in ‘breadth’ and I’ve got increasing demands on me to reach new ‘depths’.

    I’m expected to manage diabetes and CKD like a specialist, but there’s also all the primary care that everyone else requires.

    It’s a challenge we all face as medical providers.

    I’m not too proud to “look it up.” I don’t have the luxury of unlimited appointment slots and my patients don’t have the ability to get in to see me at the drop of a hat. A visit needs to be effective.

    If a checklist gets me there and gives value to the patient, then why not use it? If I can bolster a tough decision with evidence, it makes it a lot easier to move on to the next one. And if I can quickly move through a rote task with some tried and true boilerplate, it frees up time to deal with the long tail of one of those zebras (I guess the pun was intended).

    In short, I’d prefer to be an effective clinician rather than a amazing artist of medicine.

    And it all can get better. And we all can participate, rather than observe and pontificate. Go to soapnote.org and see.

  2. The statement that decision support tools will turn doctors into Idiots is the same as saying that using a safety net turn an acrobate into a fat clumsy dummy. If the doctor is competent and the computer will say about a 1000 times that he is right with the diagnosis he will be more self-confident and cool-headed and less nervous the 1001st time and will work fater and more effective, which is critically important in medical profession. I would certainly feel safer if my doctor’s decisions were backed by a computer system.

  3. Here at Body Tonic Tea, our goal is to improve the quality of life, and the overall well-being and enjoyment of people who live everyday lives. We do this by providing quality healing herbs to keep the body purified of toxins. We also provide rare, gourmet teas from different parts of the world.

    Drink Body Tonic Tea after each meal and the herbs will detoxify your body. You can keep your current eating habit and drink this tea to remove the harmful properties of the food. It is highly recommended that we change our eating habits into more healthy diets and this is one way to do just that. By adding this tea to your meal, you are altering your eating habit by including herbs designed to detox your body; toxins that would otherwise be preserved in your colon, lungs, intestines, etc. and cause severe health problems. It is the one tea for the full body.*

    Body Tonic Tea is a tea that is designed to complement a healthy diet but also to boost a healthy diet from an “unhealthy” diet as a supplement. Body Tonic Tea is a mix of herbs reported to have the value to perform the function of cleansing the body from head to toe. Extensive research has been done on these herbs by M.D. Herbalists and each herb can be researched on one’s own. This is the perfect mix of herbs to allow you to eat what you choose and reverse the effects by drinking this tea. We also encourage a diet change to a more fruit and vegetable diet, however, if that is a dramatic step, this tea can supplement your current diet to help you stay internally cleansed and purified. Thank you ahead of time for taking the initiative to help yourself, and to support your health.


  4. We aren’t going to make cashiers at the grocery store calculate the purchase on a sheet of paper, by using abacus of in mind, right? We let them read off bar codes by the means of computers and that works fast and effective. So why then not let doctors use decision support systems? I believe those who want to rely only on a human doctor don’t know the actual numbers of medical mistakes that take place every day, without computer support. We should be grateful to doctors who prefer to back their decision with the computer system.

  5. I think using computerized support is absolutely normal, unless the doctor uses it for every trivial case as a major decision making tool. At least. let them use it to check thier compitence AFTER they make their own conclusion. It’s better to find out that you made a mistake and correct it than prescribe wrong treatment. There is nothing “idoitic” about correcting mistakes or knowing that you were right. The “idiotic” thing is a wrong diagnosis or killing a patient. Patients need to understant that.

    • I would suggest using it like you do spell-check in writing, run it through to check for obvious errors, and then read it again yourself to make sure the machine didn’t miss anything. It is ONE tool of many available, starting with the doctor’s own training and experience. Used as simply the opportunity to share experiences with others in the field the CDS has the potential to become an excellent resource. How many people who object to this would be upset if their doctor had a copy of the Physician’s Desk Reference on their desk? Same thing a respository of information. The ability and willingness to use available resources will vary among humans, even doctors.

  6. @ Jason Smith
    I certainly hope you are not a practicing physician or that you treat patients in any way. The way you speak to people on this post (and others, I’ve read your responses on other topics) is terrible. First and foremost you should always listen to your patients…I can tell you do not, I think you like to listen to yourself ramble and, of course, you are ALWAYS right. The mother had a suspicion and the pediatrician should have listened…mother’s gut feelings or instincts for their child are much stronger than a physician’s gut feelings. Shame on you for the way you treat people.

  7. Margalit:
    I can’t agree both are right.
    Intuition and rule-based decision-making may come to the same conclusion. They may also be related as you suggest, if your intuition is based on years of following rules and understanding their strengths and weaknesses.
    But they are most definitely not the same thing – and all you have to do is teach novices to realize that. Telling them to follow their intuition is completely different than telling them to follow rules.
    That’s why Groopman’s philosophy and Gladwell’s are different. Not so much for the experienced observer, altho they are often different there too, but for the person learning the practice. You can’t teach them to do both a the same time- they have to learn one, usually rule based, and then learn the other.
    Also, I don’t happen to agree with the idea that the three most dangerous words in medicine are “in my experience”- they certainly can be- especially in someone who doesn’t have much experience or doesn’t interpret their experience well, but the school of thought that experience is to be disregarded completely (which is what the phrase implies) is just as dangerous if not moreso. The most complex thing about practicing medicine is figuring out what the value of experience actually is, which is why I don’t completely concur with Groopman even if I think he is closer to the mark than Gladwell. There is a modicum of arrogance, unwarranted in my opinion, in completely dismissing experience – in any endeavor.

    I prefer the two most common fallacies in medicine -“old is good” and “new is better”

    • Oh, no, Cory. I did not mean it that way. Of course you teach rules for decision making. And after applying those rules and constantly adapting them by continued learning, over and over, the rules become part of who you are, and to the untrained eye, they appear as intuition. Novices, by definition cannot have this type of “intuition”, which is really not intuition at all. It’s just expertise.
      And expertise, in my opinion, is the ability to interpret and apply your experience to novel situations. Some folks can have lots of experience and very little expertise, but you cannot have expertise without vast experience. Am I getting too convoluted again?

  8. Speaking of Dr. Groopman- has anyone noticed the basic premise of his book on how doctors think (systematic, following rules) is almost diametrically opposite from that in Malcolm Gladwell’s writings on how doctors think (experts using intuition).
    They both can’t be right.

    (of course the right answer is somewhere in-between as these posts suggest).
    But it’s fascinating to see the two popular writers on how doctors think write such contradictory things – and I have seen no reviews that allude to this.

    • I would take Groopman over Gladwell any day (and I have the books by both) when it comes to cred on doctors’ inductive and deductive cognitive processes.

      Groopman delves into some of the heuristic liabilities that doctors are prone to (really just a special case of some of the cognitive hazards that “experts” in general are prone to).

      One of my favorite quotes, btw, from the good folks over at sciencebasedmedicine.org is “the three most dangerous words in all of medicine are ‘in my experience’.”

    • That’s a pretty interesting observation, Cory. The more I think about it, the more I think that it is possible that both are right.
      What is intuition? The dictionary says “The ability to understand something immediately, without the need for conscious reasoning.” But then what is immediately and do we really know what level of reasoning went into it?
      I can only speak for myself, but I can sit in a meeting and estimate cost and duration for complex programs, running over several years, “off the top of my head”. After weeks of research and accounting, the official numbers and resources are almost always within a few percentage points of my initial “intuition” based numbers. I know I am not running algorithms in my head, and if I am asked to explain the number, I usually can go through reasonable explanations, but I also know that I didn’t consciously go through those steps in my head. So what happened? I think there is a level of understanding that becomes automated and largely unconscious, as a result of experience, or Gladwell’s 10,000 hours. At that point it looks like intuition, but I don’t think it is.
      I think Groopman’s rules and systematic analysis become embedded in one’s brain and get executed in a background process – sorry for the computer analogy (probably another item embedded in my brain).

  9. This is kind of a foolish premise. Decision support tools (especially diagnosis ones) aren’t turning doctors into ‘idiots.’ They are just a supplementary tool. There is no way I would want a physician to simply rely upon the diagnosis generated from a decision support tool. Physicians just can’t keep up with the amounts of information that are constantly being amended and added. That’s all. What is more worrisome to me is how many physicians think they know about the drugs they are prescribing, their side affects, and their potential interactions.

  10. Unfortunately they all were probably using an approach similar to a clinical decision tool. That is in fact what we are trained to do; make clinical decisions.

    To assume some tool would give different results is not valid.

    My cable company customer support uses decision tools, too. Each time I would call with the same unsolved problem I would get the same shtick, until I could get to someone who could just listen and not use the tool.

    I am always seeing patients who are not getting better after a visit to their doctor. I am always seeing military patients after they have been through military medicine, which already uses tools and protocols

    If I can’t think differently than the previous doctors, then I get the same results. When I think about the other possibilities, I sometimes find the real truth. Take away my freedom to be independent and require me to justify my pursuits “out of the box” and you get the same efforts repeated mindlessly.

    • If you’re just arguing against a silly view that using more decision support tools will solve all our problems, fine. And if you’re just pointing out that a superficial application of protocols can lead to new problems, great. Of course!

      But you so consistently chime in to resist any promotion of evidence-based medicine that I can’t help but think you don’t want doctors to change they way they practice. In particular, you don’t think it will help to make changes to be more consistent about considering all proven possibilities in diagnosis and treatment, or adding more checks to ensure they are following today’s (and not 20 years ago’s) clinical guidelines, etc.

      What you won’t accept is that the guidelines are just that. They don’t stop anyone from thinking “out of the box.” If the protocol says look at A, B and C first, and you do but still can’t find a diagnosis or effectively treat the problem, the protocol does not say try A, B and C again. It doesn’t say give up. It doesn’t say there must be no problem. There are almost always less common D, E, F, G mentioned that have been identified in the past. And if you rule out D, E, F and G as well, you can go with your gut or some untested theory. But in using the protocols, just like in learning medicine itself, you are drawing on the collective experience of the profession, how is that harmful?

      As for your examples, how many of your former patients are being seen by other doctors because they didn’t get the results they wanted from you? Patients move around, sometimes for reasons that have little to do with quality of care delivered.

      Have you ever practiced at a place like InterMountain where protocols are in place? Is InterMountain a horrible place to practice medicine? Are the docs there idiots? You are allowed to deviate if you think the case at hand is an exception to the rule, and all you have to do is explain why. Is that too much to ask? No. It is exactly the right thing to ask.

      • Decision tools will be used by less qualified people to judge care of the patient. Ever do worker’s comp? Ever have to get authorization for tests? Ever have to call an insurance plan to get coverage for a prescription they prefer not to buy?

        I have seen “evidence” make many 180’s in my career. I am very aware that no one can make you sicker quicker than a doctor. I also must make the choices as I see them, not as the cheap seats see them. If I must answer for a situation, it will be my choices and not some generic algorithm that choose. I also work in real time. I also must maintain a productiviy level. I am not caring for one person at a time.

  11. Yes, there were a number of other diagnoses, but several of those had already been ruled out. Most importantly, I quickly hit descriptions, written by parents, of what their toddlers looked like at diagnosis – the distended belly, the weight loss, the leg pain and refusal to walk, the odd intermittent fevers – and realized that my son was exactly the same. I also noticed that in all of the official medical sites that popped up with common diagnosis, the same line “malignant neoplasms must be ruled out” or “tumors and other malignancies are a differential diagnosis”.

    It wasn’t just the pediatrician who had the blinders either. My son spent a week at a pediatric hospital, supposedly one of the better ones in the country, and was seen by the chief of pediatrics, who noticed my son’s distended belly and matchstick legs and arms, but then stated that he couldn’t possibly be very sick because he hadn’t lost weight (yep, in fact he gained weight due to ascites).

    One of the common symptoms of neuroblastoma, which my son did not exhibit, is “racoon eyes” – extensive bruising around the eyes. I have talked to several families that were suspected of child abuse because of this, delaying the proper diagnosis. Misdiagnosis of childhood cancer is very common, even though childhood cancer isn’t all that rare. It is the leading cause of non-accidental death in children. Neuroblastoma is actually the most common childhood solid tumor. It has pretty distinctive symptoms that could easily be flagged by a good decision support system. Unfortunately, pediatricians tend to have blinders about childhood cancer, so this is a place where use of guidelines and decision support systems could help.

    • “One of the common symptoms of neuroblastoma, which my son did not exhibit, is “racoon eyes”

      Once again you display your ignorance. Neuroblatoma is not even on the top 10 list of the most common causes of “racoon eyes”

      • But it needs to be ruled out, especially in a child who is also losing weight, or has mysterious fevers over several weeks or months, or leg pain. That is exactly the point – a decision support system can remind a doctor of other possible causes for something like raccoon eyes – causes that need to be considered. A doctor who has never seen neuroblastoma may forget that it is a possibility. You just made my point.

  12. I *wish* my son’s pediatrician had used a decision support system, or even just Google, back when he was being diagnosed with what turned out to be stage 4 cancer. My son, who was 2, was a walking textbook example of abdominal cancer. I had already Googled his symptoms and was certain he had neuroblastoma. Well, the pediatrician just couldn’t believe that, so we went round and round for 4 miserable weeks, including a one week stay in the hospital during which NO ONE EVER ran a scan, while my son got sicker and sicker. Well, finally, the pediatrician gave in to the “hysterical mom”, had the scan ordered, and guess what? It was indeed stage 4 neuroblastoma. My son was so sick by that point that when we got him over to Sloan-Kettering, they sent him to the ICU. They told us he was endstage and might not survive chemo. Fortunately, he did, and is still with us today – but we could have been spared a lot of horror, and my son’s treatment would have been far smoother, if the idiot pediatrician had known how to type his symptoms into Google.

    • The idiot pediatrician would have had to believe what he was reading. Google would not change the fact he was an idiot.

      • Don’t get hung up on Google. The physician’s problem is that he was going by some half-remembered facts from med school plus his gut feeling, maybe a conversation with a colleague. As I read the story, he refused to actually look at the state of the art diagnosis guidelines. That is cowboy medicine, the doc as lone ranger who shoots from the hip because it’s faster and shows off your skill better.

        Or, let’s say the doc did look and as you say refused to believe what he read. It may be idiotic, but what licenses that idiocy and fosters its persistence is an attitude of disregard for “evidence-based” medicine and a preference for intuition.

    • I call BS on your post. Please post your son’s symptoms and I will quickly google them and easily find 10 more common diagnoses than neuroblastoma.

      You are a fool for saying that neuroblastoma was teh most likely cause of the symptoms. Neuroblastoma is NEVER the most likely cause of ANY set of symptoms.

      • Exactly. Moreover, this is probably a good example of how evidence-based and cost-effective medicine looks like. Both concepts are based on statistical calculations and are aimed at providing the cheapest treatment for the most likely diagnosis first. In aggregate this should work well for most patients, and lower costs for the population.

        The problem here is not that the idiot pediatrician didn’t Google the symptoms. The problem is that the idiot pediatrician never developed a keen eye for the unusual, and a “gut feeling”, a.k.a. lots of thoughtful experience, for something that is not quite right, not quite by-the-book.

        Suggested reading: Jerome Groopman’s book on how doctors think (and they do need to think, even with lots of evidence-based info available)

    • Exactly. Professors and scientists use Google too when they don’t have something at their fingertips. not to say google is the best approach when a specialized (re)search engine is available.

  13. “In one [study], even tech-savvy undergraduate and graduate computer science students preferred physicians who rely on intuition instead of computer aids.”

    Change the word “physicians” with “airline pilots.” We might toss in the word “checklists” as well.

    Just sayin’.

    • An airline pilot does not:

      discuss his decisions with the plane, let the plane participate in choosing among various courses of action, consider the psychological condition of the plane that may be contributing to the problem, discuss the financial aspects of his decision with the plane, or communicate with the plane’s relatives.

      Not the way I want to be treated as a patient.

        • “THAT” was in response to the post using the cliched analogy of physicians and airplane pilots. Checklists are useful in only a few very limited situations in medicine, but they don’t make stupid doctors smart.

          Ask any doc who’s been in practice 10 years or more about her gut feelings. She’ll reply, 9 times out of 10, that every time she’s ignored them, she’s made a mistake. Tone of voice, body posture, eye movements: all this adds up to the “intuition” that BobbyG wants to eliminate from medicine.

          • Idiotic response.

            “the “intuition” that BobbyG wants to eliminate from medicine.”

            How juvenile can you get? Why don’t we just eliminate ALL documentation and instrumentation aids so that physicians can just intuit away free of any “constraints?

            Lot of people post serious comments here. You are not one of them.

  14. There is not enough money in the world to pay for care if doctors have to “show their work”.

    This used to be called “cookbook medicine”. There is nothing new here, except the illusion of “evidence”.

    Well, most decisions have no evidence either way. Also some clinical situations require a negative test. How can there be any evidence to support a negative. Famous guidelines as decision tools do not keep you either lawyer proof or judgement proof.

    This will be fun to watch.

  15. Idiots? No ! Lemurs going off a cliff, yes. One cannot retain all medical knowledge, internet references, yes, drug lists, side effects, yes. All things have to be in balance….relying on one source…..dangerous, indeed.

  16. Speaking of turning doctors into idiots, how am I supposed to find yesterday’s posts (or the day before, etc.) on this blog? Now they disappear into never never land unless, presumably, I can remember their titles and figure out what category you decided to put them under?

    This has singlehandedly decreased my attention to this blog by 90%. Some might say that’s a good thing, but…..

    • Ah, the airline pilot analogy. So, eminent physicians: surely you would refuse to fly in a plane where the computers on board advised the pilot, but he could not overrule them.

      But would you feel safer flying in a complex modern aircraft where the pilot told the passengers he’d been flying safely for 30 years and had no need for anything more than his “gut” feeling? Probably not.

      Any physician who huffs and puffs about decision support as cookbook medicine should be required to drop all the decision support he takes for granted in his car, his cell phone, his kitchen, his office and everywhere else. No one in this field believes that decision support should be utterly prescriptive; that’s silly. No one believes the gut instinct and human contact aren’t important; they are. But anyone who believes computerized decision support can’t improve care shouldn’t be reading this blog because it appears on an actual website — and you’re stuck in the 19th century or somewhere without access to the medical literature on this topic.

Leave a Reply

Your email address will not be published.