The Case of the Missing Evidence

The most telling detail in Robin Hanson’s lecture about doctors was about a nurse assigned to measure hand-washing rates among surgeons at her hospital. After she measured the hand-washing rates, she — as ordered — correlated them with death rates. It turned out that the surgeon who washed his hands the least had the highest death rate. For reporting this — as she was ordered to — the nurse was fired. Robin learned this story from his wife, who was a friend of the ex-nurse.

I was very impressed by Robin’s lecture, which was both accessible and profound, and it was one reason that during my next encounter with a doctor I was more skeptical than most patients. As I blogged earlier:

I have a tiny hernia that I cannot detect but one day my primary-care doctor did. He referred me to Dr. [Eileen] Consorti, a general surgeon [in Berkeley]. She said I should have surgery for it. Why? I asked. Because it could get worse, she said. Eventually I asked: Why do you think it’s better to have surgery than not? Surgery is dangerous. (Not to mention expensive and time-consuming.) She said there were clinical trials that showed this. Just use google, you’ll find them, she said. I tried to find them. I looked and looked but failed to find any relevant evidence. My mom, who does medical searching for a living, was unable to find any completed clinical trials. One was in progress (which implied the answer to my question wasn’t known). I spoke to Dr. Consorti again. I can’t find any studies, I said, nor can my mom. Okay, we’ll find some and copy them for you, she said, you can come by the office and pick them up. She sounded completely sure the studies existed. I waited. Nothing from Dr. Consorti’s office. After a few weeks, I phoned her office and left a message. No reply. I waited a month, phoned again, and left another message. No reply.

Yesterday Dr. Consorti finally got back to me, by posting a comment:

Seth, While I am in the process of finding papers in the literature to satisfy your scientific curiosity on why this hernia should or should not be fixed I am additionally trying to care for around 30 new patients referred to me for their new cancer diagnosis in the last 3 months. This may or may not explain why I have not been motivated to answer your call regarding your ambivalence about fixing your hernia. Yes, it is small and runs the risk of incarceration at some time. I will call you once I clear my desk and do my own literature search. Thanks for the update. Eileen Consorti

Fair enough. She’s busy. And I am glad to have her reply and her view of the situation. On the other hand, I am pretty sure the studies she was so sure existed — that justified the surgery — don’t exist. To call my curiosity about whether the proposed surgery would do more good than harm “scientific” has a bit of truth: No doubt scientists understand better than others that you can test claims such as “you need this surgery”. But it isn’t “scientific” in the least to worry that a medical procedure will do more harm than good. Everyone, not just scientists, worries about that. Surgery is scary. Let’s set aside the death rate, which is low but non-zero. How many brain cells are killed by general anesthesia? Dr. Consorti doesn’t know, nor do I. The number is plausibly more than zero. I suspect a power-law distribution: Most instances of general anesthesia kill a small number, a small fraction kill a large number.

I pointed Robin to Dr. Consorti’s response. He replied:

I wonder if she even realizes that she in fact doesn’t know why you should get surgery.

What I know and Dr. Consorti, very reasonably, doesn’t know, is that my mom was a librarian at the UCSF medical library and has done a vast amount of medical-literature searching. If she can’t find any relevant studies, it is very likely they don’t exist. And my mom did find a study in progress, which, to repeat myself, shows that my question about cost versus benefit is a good one. Others had the same question and launched a study to answer it. Robin’s lecture helped me ask it. Thanks, Robin.

More. Robin’s version of the fired-nurse story is here. Thanks to Charles Williams.

41 Replies to “The Case of the Missing Evidence”

  1. it’s an agent principal problem, doctors can enrich themselves at the cost of wasting their patients’ money, and additionally if they dont conduct every imaginable procedure they run the risk of being sued for malpractice by a trial lawyer, another individual who earns money by wasting their clients’ instead of their own, it’s no surprise medical care grows more and more expensive when all the incentives are aligned against the patient.

  2. “it’s against the patient” in the sense that the patient could take an unnecessary risk of surgery; this risk is present regardless of the financial implications.

  3. We can all grant that this surgeon stands to make a nice buck in a short amount of time (provided your insurance company pays her) if she does this routine hernia surgery. So she has that incentive.

    But I don’t understand why, in all this Consorti business, no one has acknowledged the elephant in the room, which is the legal incentive this surgeon has to advise you to undergo surgery.

    She knows a little about abdominal anatomy, so it doesn’t take a wild flight of fancy to imagine a more substantial hernia developing.

    And she knows a little bit about malpractice lawsuits (hopefully not from experience), so it’s not hard for her to imagine that if she advised you not to get surgery, and then you had a major problem with this thing, you might sue her.

    That seems like it would also be a strong incentive to operate. Where does that play into all this?

  4. Uhm — odd, that neither you nor Robin mentioned Semmelweis. in broadstrokes, the nurse re-created his seminal findings in Vienna in 1847. For his insight and standing firm against the ‘scientific consensus’ of the day — he was driven back to Budapest and his reputation thoroughly besmirched.

    BTW Crichton and others (myself included) think that are poignant parallels to be drawn from this episode of scientific history and current mainstream scientific consensus’s stance on global warming skeptics.

  5. “Robin learned this story from his wife, who was a friend of the ex-nurse.”

    This sentence stuck out for me.

    There’s a subbranch of the field of folklore called “Emergency Room Lore” by anthropologists. These are apocryphal stories passed on by nurses and doctors in hospitals. It’s a very rich and developed body of folklore.

    Jan Harold Brunvand has a section about it in one of his books. The reason why it’s so vivid in my memory is that a few days after reading Brunvand’s book, a coworker told me about an incident that happened to his mother, an emergency room nurse, which exactly matched one of the examples in the book. I did what I have learned you never should do when you hear an urban legend (and I probably shouldn’t be writing this comment!): I challenged the story and related what I had just read in Brunvand’s book. The coworker was incensed and asked if I was calling his mother a lier. Before I wised up and shut up, I tried to explain that folklore works because people unconciously move up the source of information one person when they pass it on. It’s always “a friend of a friend.”

    I think if you tried to track this story down, Robin’s wife’s friend will say, “Yes, it’s absolutely true, but it wasn’t my friend, but a friend of my friend.” And this would repeat as far back as you have patience to follow it.

  6. While not impossible that the doctor could get the nurse fired, doctors are not hospital employees or supervisors. The doctor would have to convince an administrator that the woman deserved firing for cause, pretty difficult in this day of employee lawsuits.

  7. I do not know Robin Hanson personally; I know only the persona he presents on his blog, Overcoming Bias. That persona dislikes doctors on several levels. The justifiable level is that they waste healthcare resources and participate in services that are not helpful, while being paid to do so. It galls him that such a profession enjoys high respect among the general public, because, as he frequently implies, and sometimes states overtly, he believes doctors harm their patients intentionally for money. The persona on his blog is a little out of control on this issue, as the “nurse story” illustrates (he used it in a major blog post). Dr. Hanson is so precise when it comes to the economic evidence he presents (he has persuaded me completely about the lack of value of my profession overall), yet he allows his brilliant (I mean that term literally) mind to fall into the same trap as a lesser human when he passes along such gibberish (as Mark and Dennis Mangan point out correctly). This story is so untrue on so many levels, and obviously so to any informed person, that both Hanson and his wife should be embarrassed for having passed it along. By the way, vigorous scrubbing has been on its way out for several years. Current topical solutions work better without it.

    Dr. Consorti should have been more honest with you about her lack of evidence regarding hernia surgery. This “doctor arrogance” is what causes Hanson to lose his professional perspective about the doctors. What Dr. Consorti *did* know is that tiny hernias are more likely to become acutely incarcerated causing a life-threatening emergency, frequently requiring partial bowel resection. She cannot say what your individual chances of that are, for as you have noted, the studies are not available. I’m not sure how they could be, especially in this litigious age. She recommended the surgery, because she feels the outcome is better when done electively as opposed to the emergency situation I described. Now, she’s off the legal hook, and you are free to decide as you see fit.

  8. “Dr. Consorti should have been more honest with you about her lack of evidence regarding hernia surgery.” This is unfair to Dr. Consorti. She honestly believed the evidence existed, I’m sure. However, the fact that she was wrong about the state of the evidence is a serious problem whether she is honest or not.

  9. I wonder if it would be possible to develop one’s own risk-assessment? This would be possible if there were the following numbers (all per year)
    a) the number of emergency operations for incarcerated or strangulated hernias
    b) the number of people who die due to incarcerated or strangulated hernias
    c) the number of people who have hernias each year

    As for me, fifteen years ago I had an emergency hernia surgery and it scared me seriously. That’s why I am having another hernia surgery soon, even though my last one had its problems. But then, my current hernia is large and hurts when I sneeze. It would be a much more difficult decision for a small hernia. I wish you the best of luck.

  10. Dennis Mangan — it really depends on the hospital. In some hospitals, nurses are subjected to constant and incredible levels of hostility and harassment and can be readily fired if they displease any doctor who draws in revenue.

    In other hospitals, swearing once at a nurse will get your credentials revoked.

    Abuse from doctors is a huge issue in the nursing profession and the subject of a great deal of literature. You would think that they could file hostile work environment claims, but it doesn’t seem to be the case.

  11. You would think that they could file hostile work environment claims, but it doesn’t seem to be the case.

    Disclaimer: Anecdote

    Twenty years ago, the technically-best and highest revenue-producing cardiovascular surgeon in town was brought before the executive committee at my community hospital. A nurse had brought a summons indicating a lawsuit against the hospital and the doctor for habitual “sexual harassment” (sexual only in the sense that his verbal abuse was aimed at someone of another gender). She agreed from the outset to drop the suit if the doctor were removed from the staff. He was, and from the other local hospital staffs as well. As you say, perhaps it varies.

    I think there’s more to the “nurse story” than she revealed to Mrs. Hanson.

  12. 1. Seth, your post refers to a link to the full fired-nurse story, but I can’t find it. Could you have left it out?

    2. About urban legends: yes, Robin reports that his wife spoke to the actual nurse. However, Mark is saying that this is exactly the classic form of the transmission of urban legends. Whenever you hear that someone heard something from only a single link away (a friend of a friend, or FOAF), it’s a red flag. When that single link is followed, too often it leads to another link, and another, and another. Did Robin talk to the nurse himself? Does he know (even if he’s keeping it private for whatever reason) the name of the hospital, what year this happened, and so forth?

    3. Patrik, I think the reason Semmelweis wasn’t mentioned was simply that the connection was so plain and obvious, not because no one knows about Semmelweis and the birth of the germ theory of disease. Besides, the moral I got from the reported correlation between death and less hand washing was this: less careful doctors provide worse care overall. I didn’t think the story was supposed to point to hygiene, specifically.

  13. Vesna, thanks for pointing out the missing link. I fixed it. As for your Point #2, I’m sure “my wife knows the person involved” is not the classic form of urban legends.

  14. Thanks for fixing the link!

    I notice that Robin doesn’t say “my wife knows the person involved,” but “a colleague of my wife.” Which could mean someone that she doesn’t actually know. I posted a comment on Robin’s blog asking if he could clarify. Now I’m curious!

  15. @Seth: my question about cost versus benefit is a good one.

    It seems that when all the evidence was weighed, it was the anecdotal, experiential bias of the surgery profession versus a study in progress. With nothing else to go on, what did you decide to do, and what does your decision tell us about dealing with matters for which there is not perfect rational evidence?

  16. I decided to wait for the surgeon to tell me about the studies she told me she would find. If she never find those studies (or at least never calls me back), I’ll wait for the study in progress to finish. A few days ago my mom found a study of a different (larger) type of asymptomatic hernia which found that doing nothing (“watchful waiting”) was a good option. If no more evidence surfaces, I’ll do nothing.

  17. This sort of encounter is becoming more common among doctors and their patients of equal or superior intelligence. The medical guild has published all their information on the Internet.

    Well, no. They haven’t published all their information. They’ve published the information they bother to publish.

    Some years ago my girlfriend had a deep venous thrombosis. It started out extremely serious and then it stabilized. She’d need to use a pressure bandage for the rest of her life and maybe have some circulation problems etc. I did a lit search and the results were pretty horrendous. I was upset. She got a referral to a prestigious surgeon who reassured us.

    “Why do you think there’s any problem here?”
    “I did a lit search and the published literature was pretty scary.”
    “Why do you think you could understand about that?”
    “I’m a biomathematician. I can follow the literature but I don’t actually have any experience.”
    “Well I do. I see DVTs *all the time*. There’s nothing to worry about.”

    My girlfriend started a regular exercise program. Pretty soon she was feeling mostly recovered.

    The trouble was, they didn’t publish about it unless it was something serious. There’s a great big publication bias.

    If you go to a plumber he won’t have engineering studies that show the way he does his work is the best way. There are probably engineering studies taht show everything he does is sub-optimal, that there are newer methods that work better. But he has a big body of experience, most of it handed down to him, that says what he does works.

    Doctors are a lot more like plumbers than they are like doctors or engineers. When you consult a doctor you are depending on his experience and his training. Hardly any of that is published or publishable. It is deeply unscientific. But on average the doctor has some idea how well it works, although you can’t depend on him to truthfully tell you how well it is likely to work.

  18. Google “deep vein thrombosis” and 560,000 hits appear. Each of those have multiple references to other (or cross) references. Contrary to what you say, almost every thought that goes through doctors’ minds has been published somewhere. Much of it is locked behind subscription-only journals, but even these are available in medical and hospital libraries.

  19. RU, I googled ‘deep venous thrombosis’ and got 1.5 million hits. I googled “post phlebitic syndrome” and got only 28,000 hits. Years ago when I did that search it was only 9000. The ones I looked at emphasized the worst cases.

    Maybe my search was biased but I tried to include lots of randomness to pick up things I didn’t know I was looking for. I believe the bias was in the literature and not much in my search of the literature.

    The prestigious surgeon we talked to had not publisned anything about DVT that I could find. But he had sufficient experience to say that her PPS would probably subside with no problems. And he turned out to be right. DVT was not a research interest for him, it was just a complication he saw a lot.

  20. J Thomas: “only 28,000 hits” on post-phlebitic syndrome”. Yep, sounds like a conspiracy to hide info to me. The “medical guild”, for a surgical specialist, requires a minimum of 13 years training, at little or no pay. Chances are your knowledge and experience (and that of your biomathematician girlfriend) may still fall short of that of a surgical specialist, even after hours of “lit search”. I would also suggest that the two of you not do any complicated plumbing based on a lit search. Nor any complicated engineering. God knows what they may be hiding from you.

  21. Chances are your knowledge and experience (and that of your biomathematician girlfriend) may still fall short of that of a surgical specialist, even after hours of “lit search”.

    RU, that was my point. My understanding of statistics etc from my biomath training might give me a much better ability to understand randomized trials than most physicians — a large portion of them wind up not at all showing what the authors think they do.

    But that’s no substitute for experience. I couldn’t look at her leg and tell whether her PPS was really serious. All I had to go on were a few pictures and a lot of wordy descriptions. The surgeon could tell at a glance that it wasn’t that bad because he had experience. He had DVT and PPS as complications for his surgery patients all the time.

    I wouldn’t be surprised if I knew much more of the literature about those topics than he did. But that was no substitute for experience, and I knew it.

    The medical guild has published all their information on the Internet. Good for the patients, bad for the guild.

    It isn’t that simple.

  22. J thomas, we are in complete agreement about the unscientific conclusions drawn from randomized studies by the medical profession. It is the central idea behind the original post to which you are responding , as well as a facet of several of my other posts at “It’s Not Hard”. I’m having a little trouble understanding the source of your displeasure. it sounds as if you were upset that your lit search led you to believe your girlfriend’s situation was dire, while you are relieved that the doctor visit changed your impression. It sounds as if you would have have eliminated the anxiety by seeing the surgeon to begin with, rather than researching it on your own. That’s the way the way the system has always worked, pre-Internet.

  23. RU, I suspect we’re pretty much in agreement all round.

    I did my own lit search because that was what I knew how to do. I had access to a medical library, so I could get any paper that wasn’t too obscure. I knew how to do lit searches quickly and efficiently and I was quick to pick up the material.

    But that was no substitute for direct experience. I had no idea that my girlfriend’s symptoms would subside in 6 months. It wasn’t true for the people in the studies. But those people were chosen because their symptoms were serious enough to deserve careful attention.

    I could tell after the fact that she was a plausible candidate for DVT. She was on estrogen, and was somewhat overweight. She got a bad bruise on a mountaineering trip and then drove long hours with no rest break. But I’d never seen it before.

    The surgeon, however, gave DVT to over a hundred people a year. He’d seen a lot of cases and he knew what to expect.

    I’m not complaining about that. I’m just pointing out that doctors are more like plumbers than they’re like scientists. When you call a plumber you don’t expect the latest scientific results in plumbing. You expect he has the judgement and experience to do the job, and if he doesn’t you hope he’ll refer you to somebody who does. Same with MDs. We currently have no way to transfer that judgement and experience over the internet or through medical libraries, either one.

    I can imagine a way to begin to do that. We’d put all patient records into a database, with some attempt to allow anonymity. So then say you have flu symptoms. You put your symptoms into the database and it tells you about what happened to other people with symptoms like that. 97% of them had flue and got over it. A few had histoplamosis, a few went on to have coronary disease, a few had this or that other problem. So then you can reduce the search to people who’re around your age, and/or gender, and/or your geographical area, and/or whatever else is available, and maybe you get a different result — fuzzier but different.

    Every few days the database could email you to remind you to tell it how you’re doing, and if you don’t respond it checks the death records etc to see if you can’t respond….

    It wouldn’t substitute for the personal touch but it would provide a whole lot of data that anybody could mine. We have something like that now for insurers, but the data is mostly private and it’s organized around diagnoses more than symptoms and test results.

  24. J Thomas:

    As you can see from all these replies, “retired” means I have a lot of spare time. You sound a lot different now than in your original post. Plumbers=doctors is not only right, it is exactly what I said (“guild’). Your attitude seems a bit naive in that you wish for doctors to promote a system that would somehow transfer the benefit of their experience to the patients without a paid encounter. That’s where the guild concept comes in. Doctoring is a job, which requires a incredibly long apprenticeship. The doctors plan for you to pay for the service, as much as the traffic will allow, for as long as you are willing to do so. Just like plumbers.

  25. RU, the cost for medical care plus insurance overhead etc is something like 16% of GDP and rising. It’s already more than we can afford.

    I can’t expect MDs to do anything that might reduce their guild; privileges, but it’s plausible that insurance companies and government might want things to reduce expense. That could include public health measures and improved self-screening approaches.

    My father is a retired dentist. Back when fluoridation was new he campaigned heavily for it. Some other dentists said that he shouldn’t because he’d be cutting back on business. “Cutting his own throat.” But he argued that there would be plenty of dental business even after fluoridation, and he was right.

    Currently, more things to make people ill won’t increase medical payments much. The money just isn’t there.

  26. “Mark, Robin’s wife did not hear a story about someone else being fired — she heard from the person who was fired.”

    Yes, FOAF would seem to imply that these stories are two levels away, but many people will move the level up one. It’s just an unconscious need to make the story more immediate and vivid and place oneself at the center of the action.

    In the case of my workmate, his mother knew the person supposedly involved. Now I didn’t contact her and ask her about it, but the story (involving a person who fainted from cut-off circulation due to some sort of device in his pants to make it appear that he had larger genitals than he really did) was a classic piece of emergency room lore described by Brunvand.

    Mind you, I’m not saying that anyone at all is lying. It’s just that people’s minds play tricks. This is the entire basis of the study of folklore, and not only anthopologists, but many journalists have run across this phenomenon. And I’m not saying that such a study, if actually done, would not turn up the results claimed. It’s just that the particular details and person involved are unconscously fabricated.

    Why don’t you try to actually contact the nurse involved. If it’s really just one person away, it shouldn’t take much time. And if she was fired, she has nothing to lose and shouldn’t mind speaking to you. Make sure to get back to us with your results. 😉

    The way I see this is that at some point some medical professional hypothesized to another that if such a study were done such a result would ensue. Then a couple of people down the line, the story became that the study actually was done. Then at some point in passing this fable along, someone made the crack that, yeah, and the nurse would probably get fired for her trouble. And a few levels away from that it wasn’t a crack but a fact. And the urban legend went on from there, each time with a little more certainty added to it. (By the way, an excess of details that anticipate objections are one characteristic of urban legends.)

    The comments to this post show how much interest there is in the medical community about the issues raised by the story. Urban legends flourish the most when they “confirm” the opinions of large numbers of people in the community in which they are passed on.

  27. Mark,
    Thanks for re-opening this discussion. If it did not involve a world-class thinker who states in his widely-read blog that his significant interest is “the rationality of paternalism and other kinds of disagreement”, and who makes a living teaching medical economics, perhaps it would have no importance, and bear no further attention. However, a clear thinker like Seth Roberts has said that the incident was “the most telling detail in Robin Hanson’s lecture about doctors”, the result being that he “was very impressed by Robin’s lecture”. Consequently, I continue to feel that it should be fleshed out for what it is. I tried here and here to get Hanson to reconsider the matter or give evidence, but he would not (see his comments, and judge for yourself).

  28. “give evidence”? I guess you mean give more evidence. Doctors’ widespread disinterest in evidence is the subject of my latest post. I would be interested in your thoughts on the subject.

  29. “There’s a subbranch of the field of folklore” … reminds me of the stories of mice in soft drink bottles … and reading about how that was all folklore, except, of course, I had seen the citations and the actual case law involving those cases on appeal.

    Or, in a case closer to home, the actual maggots in the package of crackers I went out and bought my wife when she was pregnant.

    On a different example, which I will skip the details, you can imagine my surprise at having worked on a case and then having read about it being urban folklore. Since the law suit involved someone suing over the facts being disclosed … I rather believed the facts were acknowledged, not folklore. Pleadings at the county courthouse.

    Just a comment.

  30. FWIW when I had my visible hernia that didn’t hurt my doctor here in rural Wisconsin said he’d never seen one he couldn’t push back in so I might as well wait for it to incarcerate before I got the surgery. But I eventually got the surgery anyway, and now a nerve in my hip burns when I get tired.

Comments are closed.