A discussion of illness scripts and how they are used in clinical reasoning, along with discussion of base rate neglect, and the famous horses vs. zebras comparison.
Thank you Doctor! I nearly skip this one but glad that I watched. This is very important video how clinician conceptualize or contemplate about diagnosis. The script will navigate the history taking. Doctors at Stanford are beyond textbooks :) I thank you once more.
Very helpful,especially for us who are studying in faculty where half of professors don't know anything & the other half don't want to learn you anything.Thanks
Thanks for this! How, if at all, do illness scripts play into the way patient education is delivered? Specifically, what brought me here was the actual repeated language that physicians use when talking about any particular condition.
Dear Prof. Strong, thank you so much for uploading this series and all your great content on youtube!!! I really appreciate it a lot!! I was just having a question about the textbook's presentation/description of an illness: Why is it at all that there is an "uncommon" presentation, if the textbook is a very well established one like Harrison's and others? Is it that "atypical" means that the key feature, which comes along with an atypical presentation is really predictive but not very frequent? Isn't that a contradiction because key features of an illness should be by definition highly prevalent in a certain disease? Is it that typical key features are highly sensitive and highly specific for a special disease while an atypical presentation is "just" highly specific? Considering an atypical presentation of myocardial infarction which consists of abdominal pain, nausea and shortness of breath: Isn't it the case that certain conditions (in this cause diabetes or even female gender) interfere with the discriminatory capacities of the key features? To put it in final words: Isn't it the case that not all so called "atypical key features" are no key features at all? Lovely Greetings from Germany!! :-)
I think this rule that states "an atypical presentation of a common disease is more likely than a typical presentation of an uncommon disease" has to do with epidemology and specifically predictive values. While sensitivity and specificity are static (ie, not related to pre-test probability or prevalence), positive predictive value (PPV) and negative predictive value (NPV) are dynamic and change regarding prevalence. When prevalence is increased, PPV will increase and NPV will decrease (take for example now in Italy, if a person comes with clinical and radiological manifestations of atypical pneumonia, statistically it's more likely due to COVID19 because of the high pre-test probability : if RT-PCR is +ve : it's more likely a true positive, and if it is -ve we can assume it's probably a false negative.). Regarding the case of acute nausea, vomiting and abdominal pain : if it were a healthy 20 years-old college student, you would assume it's more likely food poisoning. However, if it were a 65 year old man with diabetes you would assume it's possibly food poisoning but you should not forget the possibility of fatal ACS etc.
The first thing is that referring to something as an uncommon presentation isn't exactly the same as referring to it as an "atypical presentation". Unfortunately, "atypical" in the context of clinical reasoning is often used more non-specifically to mean a presentation that is different than what's described in textbooks - textbooks which were written many years ago, biased by both demographics and an era of medicine in which diseases looked different. So for example, 30-40 years ago, we didn't have a test for troponin to use to diagnose acute MIs, and ECGs are insufficiently sensitive on their own - so most patients who had MIs presenting with nausea or abdominal pain simply weren't diagnosed. This resulted in doctors believing that such presentations were uncommon, which influenced how they think about MIs, and influenced what was written in books and taught in schools. And even though we can now study how these presentations are more common than originally thought, the commonplace belief is that MIs present predominantly as chest pain, which then influences how we approach diagnosing them (even when we know about this problem). This will be discussed more in an uncommon video on cognitive bias in clinical reasoning. To answer your other question about typical key features, I haven't considered it quite that way before, but at first thought I would agree that a "typical" key feature for a disease would be more specific than it would be sensitive.
I use UpToDate for clinical care several times a day; most residents I work with use it even more frequently than that. Although I'm old enough to remember it, it's nevertheless hard to imagine practicing medicine during a time before the ubiquitous use of sites like UpToDate and Dynamed...when the only way to learn about new treatments was to subscribe to journals, use PubMed (which was far more clunky than it is currently), or hear about them from pharm reps.
Do Illnes Scripts replace Frameworsk from your earlier series on clinical reasoning? i.e the process is now: Aquire Data > Key Features > Summary Statement > Illness Script > Apply Key Features to Illness Scripts
No, definitely not a replacement. Just a different way of organizing knowledge. A framework is knowledge organized around a symptom or sign. An illness script is knowledge organized around a specific disease or syndrome. They support each other.
If there is, I haven't seen it yet. Learning Clinical Reasoning by Kassirer et al is a good, modestly sized text on clinical reasoning, but it's not truly at a beginner level. It mostly teaches through cases which I think only works once some of the fundamentals are understood (i.e. works best for clerkship students on interns hoping to up their clinical reasoning game, not so much for preclerkship students).
