 Our next speaker, Stephen Navala. Now, of course, Stephen Navala, he's going to do how to think like a skeptical neurologist. We all know Steve. We all know love. We all know love. We all know love and Steve. We all love and know Steve. Here's his limerick. Stephen ends, never lived in Nantucket. But he's been to Warwick and Pawtucket. He's thoughtfully dressed with his blazer well-pressed, and if he's wearing a shirt, he will tuck it. Please welcome Stephen Navala. Thank you, George. I hope everyone's having a great conference. I know I am. I want to thank the JRAF, DJ and Randy and everybody for having us back, having me back here, and also for having a whole team dedicated to neuroscience and skepticism, somewhat in my sweet spot. I really do appreciate that. I want to talk to you about how to think like a skeptical neurologist, and I want to emphasize two things. One is how to understand what's going through your doctor's head when you're talking to him or her. Because I think that doctors don't always have time to explain their entire thought process. Patients and family members and loved ones, et cetera, go into a physician's office with a lot of preconceptions, maybe a lot of misconceptions about how the whole thing works, how the clinical decision-making works, and how the interaction is supposed to go. There isn't always time to correct all of those perceptions and misconceptions before the doctor gives you his opinion and recommendations. I have the good fortune to be an academic, which means that I have a little bit more time to spend with patients doing that, and I find that's often, sometimes that's all I do. To be honest with you, I'll get a second opinion where the previous neurologist did absolutely everything that I would do, I don't have anything to add to their workup and their impression. The only thing they didn't do was spend 15 minutes explaining everything to the patient. The patient comes to me and I explain to them everything that their previous doctor did and what they were thinking. I'm going to start with a case, actually. This is a real case, although it's also sort of a representative, sort of merged case. This is the one really interesting thing that happened with this patient, and this has happened more than once to me, is when I was seeing the patient back for the second time, I told them that the MRI scan of their brain was completely normal. No brain tumors, no strokes, no bleeds, no lesions, no inflammation. It's perfectly normal, you have a perfectly healthy brain. The patient's response to that information was to start crying. Now, that seems perplexing, you know, okay, I was thinking this was good news. The patient interpreted that as bad news, and I really wanted to understand why. Now, I already had an idea, but I kind of crystallized it for me. So let me back up a little bit and tell you, this is, again, very typical. So a patient presents to me a two-year history of fatigue, dizziness, muscle cramping, memory difficulty. I'm the fourth doctor that they're seeing. Prior workup by the previous three doctors has all been completely normal. I do a thorough neurological exam, which is also completely normal. What's the diagnosis? The diagnosis is Dr. House Syndrome. Now, what do I mean by that? The reason that the patient is seeing me as a fourth specialist for that list of complaints is because they have a misconception about what's supposed to happen. And otherwise, they would have been okay probably with the previous two or three opinions, but they're kind of stuck in what we call the endless million-dollar workup for their symptoms because they're stuck in the Dr. House Syndrome. What I really mean by that is that is the impression that patients have about how the medical, the clinical pathway, how the process is supposed to work based upon what they learned on TV. And the most dramatic example of that is the storytelling narrative of the Dr. House series. This is what happens on Dr. House. The patient comes in with an interesting illness. You never want to be an interesting patient, by the way. And what eventually house, the patient gets worse and worse and worse until House in the third act figures out, because of some plot device that's entertaining, what the very specific, what we call pathophysiological entity or diagnosis is. That means there's one specific pathological thing that's happening to the patient. You need a proper expert who's smart enough to make that very specific pathophysiological diagnosis. That then initiates treatment, treatment based on and targeted to that diagnosis and the patient gets better. That's what all patients want to have happen and that's perfectly rational. That's what I would want to have happen. You find something very specific, you fix it and I'm better. If that doesn't happen, and here's where the preconceptions really come in, the only other pathway is that the patient doesn't get a diagnosis, therefore they don't get treated, they get worse, they become debilitated, and maybe even in the worst case scenario they die from whatever their illness is. So now let's go back to the patient who cried when I told them that their MRI scan was normal. Now if in your mind there are two pathways, diagnosis better, no diagnosis dead, what path do you want to be on? And what path are you on if your MRI scan is normal? And now with that context it makes perfect sense that the patient was disappointed and upset that I told them that yet another diagnostic test came back normal. Now patients also interpret that in other way. I heard somebody shout out the word depression when I said what's the diagnosis and that certainly is on the list of things that we think about and that was deliberately a list of nonspecific symptoms that don't necessarily require that there is any neurological pathology. In other words people could have those symptoms without there being a tumor or a stroke somewhere in their nervous system. But sometimes we characterize those as functional meaning that the nervous system is not functioning optimally but it's not necessarily damaged in any pathological way. Like for example if you're sleep deprived your brain isn't working properly. That doesn't mean you have a disease, it just means your brain isn't working and you can have a lot of symptoms and in fact a lot of my patients who have that list of symptoms are just chronically sleep deprived and if you fix their sleep disorder their symptoms get better, sometimes even go away. Now when I'm in this situation where I have a patient with a negative work up with symptoms that are disrupting their life and the work up is negative and they are stuck in Dr. House syndrome the only way to keep them from going to a fifth neurologist is to teach them that this is a good news, that this is good news, that telling them that you would again it's like I have to explain to patients that not having a horrible brain tumor is a good thing. But it's true and this is not an isolated case, this is all the time, this is every week for me and what you have to tell patients is again I have to explain to them my thought process and my experience and the experience here is a great guide because I see what happens. It's like I've been here a thousand times and I know what happens when the work up is negative and I know what happens when specific abnormalities are found and trust me you want to be in the pool of patients who has a normal work up. You want the MRI scan to be normal. Patients also will tell me that just the uncertainty of not knowing is the worst thing and they would rather know they have something bad than not know what's going on but again I think that just refers back to this binary pathway they have in their head and again I have to explain to them no you don't want that. You don't want to have any of the things that I ruled out. Okay so when we can't, we always like to make the specific pathophysiological diagnosis, not saying that never happens, it does and that's our plan A is to find the specific diagnosis and fix it but that's often not the case in medicine because biology is complicated and we don't know everything and there are lots of functional things that could be wrong that don't have abnormal tests. So again that's the pool you want to be in trust me. So what do we do in those cases? So I have to explain to them here's another way this can go meaning here's another clinical pathway that we can follow. First I'll make sure that we ruled out any specific pathology that's always tricky because there's the unknown unknowns right? You could have some diagnosis I don't know how to rule out but that's always keep my mind open always have that question mark but as far as we can tell right now you don't have any pathology. We can identify contributing factors. How are you sleeping? How are you eating? Do you exercise? Do you have sleep apnea? I'll check your vitamin B levels and everything that is modifiable and then you treat what you find. So if I find that oh yeah your B12 level is really low is that causing their symptoms? There's no way for me to know. There's no way for me to know that again this correlation is not causation right? So I could find a bunch of things correlating with their illness but that doesn't tell me that that's what's causing their illness. But if your B12 level is low we might as well get it up to normal. It's not gonna hurt you to do that and in fact it may help. It may mitigate your fatigue. It may help your nerves function a little bit better. It may help that slight anemia you got going on. Is that gonna fix you? No but it's not gonna hurt and it's probably gonna help and we should do it anyway so that's always a win-win. We'll make lifestyle changes. That's always the hardest thing to do because usually patients get to where they are partly because they have some bad lifestyle habits and I can't write on a prescription pad. Start exercising three days a week, lose 50 pounds. You know I have good sleep hygiene. I could tell them to do those things but I can't make that happen but at least I can set them on the path. It's also really helpful to give patients something to do because patients want to, they feel out of control and they want to feel in control and so giving them things to focus on and things to do helps them feel like they're in control of their own healthcare because they are, because their problem is a lifestyle problem in this scenario. Manage symptoms, that is something that gets disparaged so that says you're just putting a band-aid on it. Well band-aids are good for boo-boos. I mean there are, nothing wrong with a band-aid when that's contextually appropriate. If I could make your symptoms go away, that's you know, we're pretty much there. I can't make somebody who has migraines into somebody who doesn't have migraines because nobody can do that but I could make you have fewer migraines and I can make them not so bad when you get them and then we'll monitor you clinically. I'm not gonna send you out into the world or cut you off, I'll be your physician. You can come back to me whenever you need to, maybe I'll see you once every so often to monitor your exam to make sure we're on the right path. That's all reassuring, that's the clinical pathway you want to be in because those patients do fine. Those are the patients that do well. The patients who have the deletion I can't diagnose, that's a different pool of patients. The patients who have a horrible brain tumor who have ALS believe me, you don't want to be in that group. Okay, not that we can't treat those patients, I don't want to make it sound worse than it is, but you know, you definitely want to be in the group of patients who have the normal exam. Okay, part of understanding clinical decision-making, what your doctor is thinking is knowing how to ask the right questions. Sometimes patients are asking different questions than I'm asking and that's okay. They're asking the questions that make sense. They're asking the questions that medical students ask because they haven't been practicing medicine for 20 years. They don't know that that question is not always the right question. So just to give you one example, again, patients think the only real question is, do I have diagnosis X? And what's interesting is that as physicians mature clinically, they mostly stop asking that question except for cases when we can definitively answer it. For most of the, at least I'm biased, this is how to think like a skeptical neurologist because I'm telling you, my experience as a neurologist, I'm not a surgeon, you'll get a very different talk from a surgeon, but the kind of problems that I see most of the time, we can't make a specific diagnosis and diagnoses themselves are a range of things. Sometimes all I could do is tell you what category you're in. That's kind of a diagnosis, but it's not what patients think of a very specific label. But the question that mature clinicians ask are not do you have X, but should I treat you for X? That's a different question. It may seem like the same question, but it's not. It's a different question. And the good news is that that's a question I can answer often much better. Multiple sclerosis, for example, is a very squirrely disease. It's fuzzy around the edges. Now, yes, there are some people who definitely have it. There's a huge, blurry, fuzzy line of people who may have it. And there's actually no definitive, objective way to draw a sharp line of demarcation. That would be fake anyway. All we could say is that you may have multiple sclerosis. You may not, we can't really know for certain. But I can say definitively whether or not I should treat you for MS because we have data that predicts who should be treated based upon specific operational criteria. That's what we look for. And that's the research that we do. Not many people are studying who should I attach the label of MS to. They wanna know who should I treat? If I give this patient a drug that is meant to treat, prevent long-term progression of MS, do they do better? Where's the cutoff of people who do better? So it's a subtle difference, but it makes all the difference in the world. And so patients ask me, do I have multiple sclerosis? And I have to say, I don't know, but that's okay. Because I know I don't have to treat you for it. And that may not be as comforting as they want, but I can explain to them why that's a good, good enough answer. And then we're gonna monitor you. And if at any point you get to the point where we should treat you, we will. So even though I don't have all the information, I know what we need to do. Often we summarize that when I'm giving this talk to students as we don't always have to have the answer, but you do need to have a plan. So that's what, and the information, the research, the clinical research that we do is to help us decide what is the plan? What do we do? Not what's the ultimate answer to what's going on? So physicians, again, amateurs talk about signs and symptoms and clinicians talk about predictive value. They seem like the same thing, but there's an important difference. So what I wanna know is not is fever typical of this syndrome, but how predictive is having a fever or not having a fever of somebody responding to or not responding to treatment? Very different than is a fever a symptom of this disease? Do they have the disease? So I need to know, are symptoms specific or are they non-specific? Fever is pretty non-specific, meaning there's a thousand things that can cause a fever. Fatigue is just about the most non-specific symptom there is, everything makes you fatigued. But a bullseye-targeted rash is really specific for acute Lyme disease. So that is much more useful. It's also much more useful when it's present than when it's absent, because it's only present about 60% of the time. So if you have it, you have Lyme disease. But if you don't have it, you still may have Lyme disease. I can't tell you that you don't have Lyme disease. When a sign or a symptom has a 100% correlation with a specific diagnosis, we call that pavagnimonic. That's the technical term for that. Those are gold. We love those because it's a one-to-one. You have it, you have the disease, that's great. But there are precious few of those in medicine. There's not a lot of those signs where you don't have to think you just can plug it in and a computer could tell you what the answer is. All right, risk versus benefit. We talk about this a lot at Science Based Medicine when we're talking about vaccines, for example. Naively, humans think our risk of verse and we focus on risk and we focus on the wrong risk. And we are also much more worried about the risk of doing something than the risk of not doing something. But that none of that's rational. It's not rational in that it doesn't lead statistically to the best outcome. And this is where the physician's job is to be detached, which doesn't mean we don't care. It doesn't mean we're unemotional. It just means that we keep enough professional distance that we could be the ones to make the rational calculating decisions. And a big part of that is risk versus benefit. So we always need to ask ourselves, what's the risk to the patient and the potential benefit to the patient if I do this test versus not do the test, give this treatment versus not give the treatment. Not doing something has risk too. So every decision I make, which includes not doing anything, has a risk versus benefit. And that's what we're always striving for, understanding as much as we can, informing the risk versus benefit and then making the best decision with the patient. That's where informed consent comes in. And this is why, because patients will often ask me, well, why not just get the MRI scan? So give me some peace of mind. You know, I'm saying, well, the MRI is not indicated in this case. I don't need to do an MRI of your brain because you don't meet the criteria. And then they will tell me, well, why just can't we just do it because I wanna just be sure. And that's again, superficially seems rational, but the data shows that the risk of doing the study and is greater than the risk of not doing it. There's no benefit there. And so the chances are greater that I'm going to see something that's not really a problem. And then I'm gonna be obligated to do a follow-up test and subject you to more radiation or whatever, or maybe give you a treatment you don't need or worst case scenario, like an invasive follow-up test versus what's the probability that I'm gonna find something that's going to enable me to treat you in a way that there's a benefit to you for that. So it's all the numbers game, which sounds cold, but believe me, that's what you want. You want the numbers game, right, because that's what's gonna lead you to the best outcome. This is just a little interlude. This is a CAT scan of the thorax that in a study was shown to a number of radiologists who are experts in reading CAT scans. And 70% of them missed the pathology. You guys see where it is? Oops, come on, where's my pointer? All right, that's not getting my pointer. Okay, look in the, to you, with the upper right-hand quadrant, you see the iconic gorilla. The radiologists weren't looking for gorillas. They were looking for things that they know and they didn't see anything, so they thought it was normal. All right, a little interlude. Okay, another case. 45-year-old patient presents with a two-year history of progressive frontal temporal dementia, which that means the frontal lobes and the temporal lobes were not working, causing cognitive memory and thinking problems. Now, there's more possibilities here, but the differential diagnosis came down to two probable things. Alzheimer's disease or PICS disease. The question is, which one of those diagnoses is more likely? So what if I told you that in PICS disease, it's classic for PICS disease to have a frontal temporal distribution of dementia, whereas Alzheimer's disease is more typically global. It's everywhere, equally. It can be frontal temporal, but it's usually global. Alzheimer's disease usually presents in the 70s. You know, it's more risky as you get older. PICS disease usually presents younger. Younger than 60, even younger than 40. Younger than 50. So a 45-year-old with frontal temporal dementia is typical of PICS disease, but very atypical for Alzheimer's disease. So what's the more likely diagnosis? Who thinks it's PICS disease? Clap. You're wrong. It's not PICS disease. You made the medical student mistake, which is what I wanted you to do, so thank you. But there's a specific cognitive error that we tend to make, and it's a heuristic. It's called the representativeness heuristic. When medical students make this mistake, we call that diagnosis a zebra. You're diagnosing zebras, which comes from Sherlock Holmes. When you hear hooves clopping in the streets behind you in London, think horse, not zebra. Not because the clopping is more typical of a horse, but because there's a lot more horses in London than there are zebras. So what the representativeness heuristic is, is to be over-impressed with typicalness, representativeness, and to fail to ask the question, well, what's the base rate? How common is PICS, and how common is Alzheimer's disease? It turns out that Alzheimer's disease is really common. 50% of people with dementia have Alzheimer's disease. PICS disease is really rare. It's so rare that even if you see a patient who is textbook for PICS disease, it's still more likely to be atypical Alzheimer's disease than typical PICS disease, which is counterintuitive. But that's the representativeness heuristic. And we make this mistake obviously in our lives every day. This is just applying it to medicine, as I said it was gonna do. But it's the failure to ask the question, well, what's the rate, how often does this occur? All right, that's the brain. The availability, in case you were wondering, the availability heuristic, this is a mistake we all make every day as well. It's a bias towards examples that are available to us. And the assumption is that if I can readily think of an example, it must be common, or it must be the answer. So when physicians make this mistake, so yeah, I saw a patient once who had this. Therefore, this is the diagnosis. So that's a pitfall, a clinical pitfall to fall into. But when patients do it, it's more, I heard on Dr. Oz, or I know somebody who this is like every day. Oh, I had my neighbor, they had headaches, and it turns out that they had an aneurysm. So I must have an aneurysm. It's like, no, that's just the example at hand. Or it's the popular diagnosis, right? So everyone who has fatigue and muscle aches, depending on what part of the country you live and you think you have blind disease, why? Because that's the disease that's in the news. Not because you're actually more likely to have it. Or gluten sensitivity or fibromyalgia, these are available diagnoses. And so that's what people fixate on. But it doesn't mean they're actually more likely. And it's hard to get people to unfixate on a diagnosis once they've decided that this is what they have. But, and I've actually had conversations this weekend with colleagues, like what do you do in that situation? Usually I tell patients, well, I don't want to miss what you really have. So I try to engage that thinking. And that usually helps them unfixate enough that they'll at least entertain the possibility that maybe this isn't Lyme disease because all the tests for Lyme disease were negative. All right, here's now, I'm the case study here. So I had one point in my career, this from anecdotal observation, I came up with the hypothesis that migraines are associated with sleep disorders. So I asked my patients with migraines if they had, if they, I started asking as a routine screening question of all my migraine patients, how's your sleep? And then it turns out that most of them have sleep disorders. There you go. Sleep disorders are associated with migraines. I'm done, right? That's the congruence bias. What I didn't do, I did eventually. I said, okay, well, wait a minute. What you need to do is ask patients who don't have migraines if they have a sleep disorder, and they all do too. So just that, at least patients who come to my neurology clinic frequently have sleep disorders. So you could also just ask the general population, how's your sleep? Most people over 40 or 50 will tell you that their sleep is not good. So it's just, that's the age of patients who tend to come to my clinic, at least mostly. So I'm just seeing a common problem in the age group that I'm seeing, but if you just test your own hypothesis and don't test other hypotheses or specifically test whether your hypothesis is wrong, you can falsely be led to the conclusion that your hypothesis has been verified. This is a form of confirmation bias, but it's specifically this congruence bias of only testing your hypothesis and not specifically trying to disprove your hypothesis. That's why good scientists, they set out not to prove their hypothesis right, but to prove it wrong. The test of a good theory is that it has survived trying really hard to prove it wrong. The fact that you were able to find support for your idea when you looked for it doesn't predict anything. It's not predictive that your idea is probably correct. I wanna make sure I end on time. That's another interlude. It's like my cute little drawer friend. A little facial pareidolia there. What emotion would you ascribe to that guy? Unwee, I like it. It's amazing that our brains will assign a motion to a little piece of metal because that's how our brains work. That's a different talk. Okay, I'm gonna finish with this. The most challenging thing about being a physician in a way is taking a good medical history. And what I mean by that is just this is the most intellectually difficult part. This is why computers can't practice medicine, at least not yet. Even if we have super advanced computers with the Watson expert systems who can go through so many, many different diagnoses and has all of the data at its fingertips, it's a great tool for physicians. But the reason why the computers can't be practicing medicine themselves is because there's no way they could take a medical history because you have to take a medical history from a person and people are extremely flawed conveyors of information. This doesn't mean I don't trust my patients. It doesn't mean that I disbelieve them or I think ill of them. This is people. This is how people are. My job taking a history is to be skeptical. It's to be a skeptical detective. It's to completely deconstruct my patient's narrative because that's what patients come to you with. They don't come to you with information. They come to you with their own narrative about their illness. And that narrative has been filtered through a lot of things, their hopes and desires, their other physician visits, what their friends have told them, what they've seen on TV, watching Dr. House. And then they give you their narrative and it's totally biased and manipulated. And what you have to do is triangulate and ask questions multiple different ways until you can deconstruct the most reliable bits of information and then you reconstruct a new medical narrative that is meaningful from a medical point of view that's likely to help me then know what diagnostic tests to do and how to treat the patient. So here are just some general principles. People are poor historians. We know that, right? We're skeptics. Telescoping, that means that people contract their memory in the past. If they think, oh yeah, I've been having headaches for a year, double it, triple it. They've been having them for two or three years. That's what the research shows and that's what my personal experience is. Whenever I have an opportunity to compare what the patient tells me with the written chart, they always grossly by two to three factors underestimate how long in the past something has happened unless they're committing the next error, which is anchoring. Anchoring means I say, how long have you had back pain ever since I was in the car accident? That's their answer. They don't tell me two years. They anchor it to an event ever since I had my car accident. The problem is, now if that anchoring is correct, that's great. That's really useful. But it's a coin flip. Sometimes the anchoring isn't correct and sometimes I read the charts like, well, you were complaining of back pain to your primary care doctor a year before you had the car accident. Oh. Well, then I guess I've had it for longer. That's what they say. Okay, it's written down. Or maybe it got worse when I had my car accident or whatever. So that's how our memory works. We make assumptions about cause and effect and then our memory literally brings those timelines in sync. And so the patient remembers their pain starting after their car accident because that's how their memory changed the details to fit the narrative. And their narrative is, I got into a car accident and that caused my back pain. But you can't believe that. You have to deconstruct that. Okay. Memory shifts to fit a clean narrative. I said that. Often does not match the medical record. And there are also the medical record. You've got to be skeptical of that too. Because just because somebody wrote it down doesn't mean it's true. There is what we call, or at least what I call chart lore, where a patient tells a physician, this other physician told me I have X. I have fibromyalgia. So the physician dutifully writes in the past medical history, the patient has a history of fibromyalgia. And then the next physician reads that and believes it and passes it on. But that was just hearsay. Maybe the physician told the patient, I'm going to do a test for fibromyalgia. The patient hears that or remembers that as I have fibromyalgia, they tell that to their next physician and now it's chart lore. So that's why I always tell my medical students, how was the diagnosis made? Go back to the original documents. The first references. Don't believe the chart lore. Who made the diagnosis? When? How? With what information? With what tests? That's your job. To reconstruct the story from objective information and not believe the narrative that you're reading in the chart or that the patient's telling you. This is the last slide. This is a picture of a little demon sitting on the chest of a woman who is paralyzed. This is an artist's, so Fuseli's depiction of hypnagogia. This is sleep paralysis. And this is what people thought was happening. The demon was paralyzing you and you couldn't breathe because they were sitting on your chest, but really it's just a parasomnia. It's just something that happens in your brain when you're in the transition between waking and sleeping. All right guys, thank you so much. I had a good time. Dr. Steve Novella.