We’re going to put this discussion thread up and leave it as a place for questions, ideas, comments, suggestions, and conversations about whatever topics you’d like. We hope to have some interesting discussions, while at the same time hearing from you to make sure the effort we put into this publication is tailored to what people want to read and talk about!
Gender is now better understood to be very significant from a medical standpoint, like symptom display, but I still think about this article I read not too long ago on how some astronomical percentage of lab rats used in medical experiments were exclusively male rats. I occasionally wonder what kind of discoveries might be invalid if the rat experiments included female rats. Maybe this is a very broad question, but have you found any unexpected divide in effect along gender lines?
This is an interesting question. I can't profess to know much about lab animals and how an animal's sex might influence experiments (though once upon a time I was an assistant in labs where we worked with zebrafish), but I have read about concerns of problems from using animals of only one sex in research. This is a good idea for us to look a bit deeper into!
peter Attia's podcast (and website) talk about this a lot. It has to do with getting "clean, reliable data," which is a similar argument for why human trials are often limited to a narrow group. When there's a blip in data, it becomes very difficult to discern what it causal vs. correlative. The more similar all the subjects are, the more reliable you can infer signal from noise.
Once you get out of the lab, and out of phase 1 trials, you then expand the subjects to include a wider range of people... you can then do all sorts of statistical analysis to filter out noise. If/when you get to phase 3 trials, you're now in the general population, and the real data comes streaming in.
I suspect this has been covered in this substack's article(s) on how to evaluate clinical trials. (Peter Attia has one as well--worth reviewing to get the deep dive.)
Why aren't doctors and nurses always sick? Are their families more sick than average (in which case the médical professionals are probably asymptomatic carriers) or no sicker than average ( the doctors and nurses have immune systems that entirely fight off transmissible illnesses).
i say this as a primary school teacher who, before covid, had an immune system of steel.
I'd have to dive into the data on this, but my hunch is that it has to do mostly with hand hygiene. Long before COVID came around, we've been pretty good at wearing masks around sick people and washing/sanitizing our hands--even if we don't do it as often as we should, we're still cleaning our hands dozens of times a day. Speaking from my own personal experience, I can't remember catching anything from a patient, but I have caught colds from my co-workers (working in close quarters in tiny little work rooms with shared computers/phones). And now that I have toddlers at home, the number of colds I catch has gone through the roof...
i was thinking it is à very "podcastable" question. It accords plenty of opportunities for listeners to record their own observations, it may suggest surprising conclusions (in terms of acceptability to colds, first year Kindergarten teachers seem to be always sick, whereas their senior colleagues rarely so, but i think they have similar hygiene) and, most importantly, you can make some amusing comparisons between subprofessions. Like do ear-nose-and-throat specialists get the flu more often than proctologists and gynecologists?
Why is it that when I tell various medical people in my visits that I am a QS (Quantified Self) person who tracks nearly everything I do or consume in a spreadsheet, all I get are blank looks?
I will say that “Quantified Self” is something I had to look up, so maybe they aren’t familiar with the term! There’s more data about our health and daily lives that people now collect, but one challenge is that we often times just don’t know what to do with it. We don’t have a lot of research to inform us of what to do with all the data a fitness tracker or other monitor collects--there can be some discomfort giving medical advice in this situation. It might be easiest if you’re looking at your data to tell your doctor what *you* think it means, and ask them if they have any concerns or thoughts about those conclusions. But who knows, you might find some docs who really want to dive into the data with you!
I've yet to encounter an MD interested in the gobs of data I collect, even among those doing research when I volunteer for a medical study. I have spreadsheets of data going back to 2005 detailing everything I have eaten, when I eat, when I sleep, how many hours, all my exercise, all my vitamins and supplements that I take, etc., etc.. I also have years of Strava data tracking my hikes/jogs with a chest heart strap (~530 activates since 2016!). And DNA from a couple of services also.
To me, the clear lack of knowledge and interest in QS solidifies my belief that the vast majority of MD's aren't interested in getting to the source of any problems/complaints but are instead focused on assembly line medicine - getting the patient processed, prescribed with some drug(s) or referred to someone else and on their way.
This is head scratching given the trend towards precision medicine, which QS data could surely augment. Sadly, there is little real curiosity by most MD's, which, IMO, is why there are so many stories and patient complaints of incorrect diagnosis. Hell, I'd wager that a large percentage of MD's don't even know how to use a spreadsheet or build formulas for one.
Just listening to your chapter on the Hawthorne effect and mortality around Joint Commission visits and wondered if you gave any thought to the mechanism being the “Dilbert Principle”---here is a time when the folks most likely to do harm to a patient are all out of clinical care and busy showing the binders to JCAHO.
It’s an interesting idea! When the Join Commission comes they poke their head in anywhere they want to ongoing care, so in theory they could collide with the best performing health care workers or the worst ones. But they don’t interrupt ongoing care. The managers that might show stuff to them are usually always managers the rest of the year, too. But it does get me thinking that maybe if you know they’re coming, if you’re a manager, you might be inclined to push your best performing health care workers pick up more shifts that week
Food and diet studies are always dodgy because of how hard it is to monitor what anyone eats over a long enough time to really validate it. One of the few places where the food is fairly controlled is prisons. In the case of long term incarceration there may be prisons with different diets that might show different outcomes. Unfortonatly I expect they are all very similar low quality food.
A better study would require stipulating a diet for the inmates but for the purpose of the study it would be unethical unless they agreed to it in which case there would be at least two groups to compare.
Nov 13, 2023·edited Nov 13, 2023Liked by Chris Worsham
For what it's worth, your articles were the inspiration for me to start my own substack, "Type 1 Diabetes: It's not that simple!", where I model my articles with the same goal: helping everyday people better understand one of the most complex diseases, and is the most time-intensive for patients to manage. I've linked to your articles already, and have a few more in the queue that cite your articles on evaluating clinical trials, how research studies can be flawed, and others.
I realize that T1 Diabetes is a very small audience (only 1.5M americans have it), but their unique metabolic conditions make them excellent study subjects that can reveal great insights on other (related) medical conditions and behavioral patterns for those with highly time-intensive management requirements.
It's a unique disease in that it's far too complex for those with the condition to fully understand it, yet in order to manage it, one must know considerably more than people they possibly can, and clinicians can't possibly teach them everything they need to know. This raises the paradox--and, as I will be writing--a moral hazard: Reducing the nature of the disease to such simple explanations (and treatments) that keep patients alive, but far from "healthy." This is essentially the problem with T1D, which I hope to address in my substack.
I don't think I could have written this without seeing your articles to form a model.
Thank you, Dan, for the kind words! I find that idea of not over-simplifying a complex disease to either doctors or patients to be compelling. Not only are the disease and its treatment complex, but the way that complexity has to weave its way into the unique considerations for each individual patient with unique needs that change over their lifetimes makes it exponentially more complex. It's a situation where I think doctors/healthcare providers often need to take an approach of being an advisor to an informed and proactive patient, who is ultimately the one who is in control of their care (and will be the second they walk out of the office). That doesn't mean handing over the prescription pad, but it does mean having the humility to recognize when patients are, in fact, experts in their own disease and to help them take control of it in a way that the evidence suggests is going to be beneficial to them.
On the assumption that upcoming milestones tend to focus our attention, are people more likely to die shortly after their birthday (versus at random moments during the rest of the year)?
You must either have ESP or a listening device at our ideas meetings! We've been discussing this very idea with colleagues and students and trying to find good ways to answer it. Working in the ICU and caring for a lot of patients at the end of their lives, it's common to see patients wanting to survive to various milestones (major birthdays, graduations, weddings, birth of grandchildren, etc.). Does pushing for milestones change the care patients receive? Does any additional treatment they receive actually lead them to live longer? Lots of fascinating questions here, challenging to answer--but the question of milestone birthdays is one where we at least have good data (unlike, say, an upcoming wedding). I anticipate we will dedicate a post to this in the future!
When my parents were dying, we would tell them of changes in world events that they most wanted to see (you can probably guess what they are). These white lies made them extremely happy, of course, but it was also noteworthy that they'd died shortly thereafter.
I suspect that having closure on big, important things--whether it's a big thing or little one (like a birthday)--gives the mind what it needs to finally "let go." A milestone doesn't have to be a personal one, but an existential one as well.
Very interesting. Thanks. (No ESP no room bugging, I promise.)
At the birthday-indifferent end of the spectrum, I wonder how many people schedule elective procedures like colonoscopies on their birthday. A positive background rate in this direction would also be interesting to document.
Fascinating new article with pretty important findings: "50+ years of randomized control trials in criminal justice and shows that almost no interventions have lasting benefit -- and the ones that do don't replicate in other settings." Is there any worry that a similar finding might be lurking in a large body of medical clinical trials? Why or why not?
Gender is now better understood to be very significant from a medical standpoint, like symptom display, but I still think about this article I read not too long ago on how some astronomical percentage of lab rats used in medical experiments were exclusively male rats. I occasionally wonder what kind of discoveries might be invalid if the rat experiments included female rats. Maybe this is a very broad question, but have you found any unexpected divide in effect along gender lines?
This is an interesting question. I can't profess to know much about lab animals and how an animal's sex might influence experiments (though once upon a time I was an assistant in labs where we worked with zebrafish), but I have read about concerns of problems from using animals of only one sex in research. This is a good idea for us to look a bit deeper into!
peter Attia's podcast (and website) talk about this a lot. It has to do with getting "clean, reliable data," which is a similar argument for why human trials are often limited to a narrow group. When there's a blip in data, it becomes very difficult to discern what it causal vs. correlative. The more similar all the subjects are, the more reliable you can infer signal from noise.
Once you get out of the lab, and out of phase 1 trials, you then expand the subjects to include a wider range of people... you can then do all sorts of statistical analysis to filter out noise. If/when you get to phase 3 trials, you're now in the general population, and the real data comes streaming in.
I suspect this has been covered in this substack's article(s) on how to evaluate clinical trials. (Peter Attia has one as well--worth reviewing to get the deep dive.)
Why aren't doctors and nurses always sick? Are their families more sick than average (in which case the médical professionals are probably asymptomatic carriers) or no sicker than average ( the doctors and nurses have immune systems that entirely fight off transmissible illnesses).
i say this as a primary school teacher who, before covid, had an immune system of steel.
I'd have to dive into the data on this, but my hunch is that it has to do mostly with hand hygiene. Long before COVID came around, we've been pretty good at wearing masks around sick people and washing/sanitizing our hands--even if we don't do it as often as we should, we're still cleaning our hands dozens of times a day. Speaking from my own personal experience, I can't remember catching anything from a patient, but I have caught colds from my co-workers (working in close quarters in tiny little work rooms with shared computers/phones). And now that I have toddlers at home, the number of colds I catch has gone through the roof...
i was thinking it is à very "podcastable" question. It accords plenty of opportunities for listeners to record their own observations, it may suggest surprising conclusions (in terms of acceptability to colds, first year Kindergarten teachers seem to be always sick, whereas their senior colleagues rarely so, but i think they have similar hygiene) and, most importantly, you can make some amusing comparisons between subprofessions. Like do ear-nose-and-throat specialists get the flu more often than proctologists and gynecologists?
Why is it that when I tell various medical people in my visits that I am a QS (Quantified Self) person who tracks nearly everything I do or consume in a spreadsheet, all I get are blank looks?
I will say that “Quantified Self” is something I had to look up, so maybe they aren’t familiar with the term! There’s more data about our health and daily lives that people now collect, but one challenge is that we often times just don’t know what to do with it. We don’t have a lot of research to inform us of what to do with all the data a fitness tracker or other monitor collects--there can be some discomfort giving medical advice in this situation. It might be easiest if you’re looking at your data to tell your doctor what *you* think it means, and ask them if they have any concerns or thoughts about those conclusions. But who knows, you might find some docs who really want to dive into the data with you!
I've yet to encounter an MD interested in the gobs of data I collect, even among those doing research when I volunteer for a medical study. I have spreadsheets of data going back to 2005 detailing everything I have eaten, when I eat, when I sleep, how many hours, all my exercise, all my vitamins and supplements that I take, etc., etc.. I also have years of Strava data tracking my hikes/jogs with a chest heart strap (~530 activates since 2016!). And DNA from a couple of services also.
To me, the clear lack of knowledge and interest in QS solidifies my belief that the vast majority of MD's aren't interested in getting to the source of any problems/complaints but are instead focused on assembly line medicine - getting the patient processed, prescribed with some drug(s) or referred to someone else and on their way.
This is head scratching given the trend towards precision medicine, which QS data could surely augment. Sadly, there is little real curiosity by most MD's, which, IMO, is why there are so many stories and patient complaints of incorrect diagnosis. Hell, I'd wager that a large percentage of MD's don't even know how to use a spreadsheet or build formulas for one.
Just listening to your chapter on the Hawthorne effect and mortality around Joint Commission visits and wondered if you gave any thought to the mechanism being the “Dilbert Principle”---here is a time when the folks most likely to do harm to a patient are all out of clinical care and busy showing the binders to JCAHO.
It’s an interesting idea! When the Join Commission comes they poke their head in anywhere they want to ongoing care, so in theory they could collide with the best performing health care workers or the worst ones. But they don’t interrupt ongoing care. The managers that might show stuff to them are usually always managers the rest of the year, too. But it does get me thinking that maybe if you know they’re coming, if you’re a manager, you might be inclined to push your best performing health care workers pick up more shifts that week
Food and diet studies are always dodgy because of how hard it is to monitor what anyone eats over a long enough time to really validate it. One of the few places where the food is fairly controlled is prisons. In the case of long term incarceration there may be prisons with different diets that might show different outcomes. Unfortonatly I expect they are all very similar low quality food.
A better study would require stipulating a diet for the inmates but for the purpose of the study it would be unethical unless they agreed to it in which case there would be at least two groups to compare.
For what it's worth, your articles were the inspiration for me to start my own substack, "Type 1 Diabetes: It's not that simple!", where I model my articles with the same goal: helping everyday people better understand one of the most complex diseases, and is the most time-intensive for patients to manage. I've linked to your articles already, and have a few more in the queue that cite your articles on evaluating clinical trials, how research studies can be flawed, and others.
I realize that T1 Diabetes is a very small audience (only 1.5M americans have it), but their unique metabolic conditions make them excellent study subjects that can reveal great insights on other (related) medical conditions and behavioral patterns for those with highly time-intensive management requirements.
It's a unique disease in that it's far too complex for those with the condition to fully understand it, yet in order to manage it, one must know considerably more than people they possibly can, and clinicians can't possibly teach them everything they need to know. This raises the paradox--and, as I will be writing--a moral hazard: Reducing the nature of the disease to such simple explanations (and treatments) that keep patients alive, but far from "healthy." This is essentially the problem with T1D, which I hope to address in my substack.
I don't think I could have written this without seeing your articles to form a model.
Thank you, Dan, for the kind words! I find that idea of not over-simplifying a complex disease to either doctors or patients to be compelling. Not only are the disease and its treatment complex, but the way that complexity has to weave its way into the unique considerations for each individual patient with unique needs that change over their lifetimes makes it exponentially more complex. It's a situation where I think doctors/healthcare providers often need to take an approach of being an advisor to an informed and proactive patient, who is ultimately the one who is in control of their care (and will be the second they walk out of the office). That doesn't mean handing over the prescription pad, but it does mean having the humility to recognize when patients are, in fact, experts in their own disease and to help them take control of it in a way that the evidence suggests is going to be beneficial to them.
On the assumption that upcoming milestones tend to focus our attention, are people more likely to die shortly after their birthday (versus at random moments during the rest of the year)?
You must either have ESP or a listening device at our ideas meetings! We've been discussing this very idea with colleagues and students and trying to find good ways to answer it. Working in the ICU and caring for a lot of patients at the end of their lives, it's common to see patients wanting to survive to various milestones (major birthdays, graduations, weddings, birth of grandchildren, etc.). Does pushing for milestones change the care patients receive? Does any additional treatment they receive actually lead them to live longer? Lots of fascinating questions here, challenging to answer--but the question of milestone birthdays is one where we at least have good data (unlike, say, an upcoming wedding). I anticipate we will dedicate a post to this in the future!
When my parents were dying, we would tell them of changes in world events that they most wanted to see (you can probably guess what they are). These white lies made them extremely happy, of course, but it was also noteworthy that they'd died shortly thereafter.
I suspect that having closure on big, important things--whether it's a big thing or little one (like a birthday)--gives the mind what it needs to finally "let go." A milestone doesn't have to be a personal one, but an existential one as well.
Very interesting. Thanks. (No ESP no room bugging, I promise.)
At the birthday-indifferent end of the spectrum, I wonder how many people schedule elective procedures like colonoscopies on their birthday. A positive background rate in this direction would also be interesting to document.
Perhaps because many people set a birthday as a some sot of goal to achieve and once they reach that, they feel free to check out?
Fascinating new article with pretty important findings: "50+ years of randomized control trials in criminal justice and shows that almost no interventions have lasting benefit -- and the ones that do don't replicate in other settings." Is there any worry that a similar finding might be lurking in a large body of medical clinical trials? Why or why not?
https://twitter.com/MeganTStevenson/status/1742263139331088480