“Incidence rates for deaths directly attributable to medical care gone awry haven’t been recognized in any standardized method for collecting national statistics,”
That's because they're hard to standardize. What qualifies as a medical error? It's easy to go back after the fact and analyze a case where someone has a poor outcome and say 'we should have done this another way'. That doesn't mean there was any indication at the time that the other way would have been better or that the poor outcome could have been foreseen.
I do believe there are a lot of medical mistakes made. I'm a little skeptical that there are legitimately 250k that result in unnecessary death each year. I assume many of those 250k were being treated for serious illnesses that may have taken their lives anyways (not all medical interventions are 100% effective). It would be nice to have a standardized definition, but I'm not sure how that's possible unless we are only talking about treatment received was different than treatment ordered.
I'm a little skeptical that there are legitimately 250k that result in unnecessary death each year
From purely anecdotal experience, that number is totally believable. Again anecdotally, the medical field seems to only have two kinds of people, the extraordinarily competent and dedicated, and people who don't give a whiff of a shit. The medical bell curve looks more like a bathtub.
My anecdotal data (which is primarily second-hand from friends who work closely with doctors in a hospital setting) agrees with your anecdotal data, for whatever that is worth. There are some spectacularly bad doctors out there which the medical system provides cover for, mainly, as far as I can tell, because they generate revenue.
It is important to understand that in the American private healthcare system, the doctor is the true “customer” of the hospital, in that, with the exception of services provided by hospital-employed physicians, the things the hospital itself can actually bill for have to be ordered by physicians in private practice who have patients admitted to the hospital.
There is a tremendous power imbalance between physicians and non-physician clinicians and a powerful incentive to cover up bad acts by bad doctors who perform a large number of profitable procedures.
I believe this is a big part of it. Anecdotal as well, but my grandmother had a new leadless pacemaker installed for dubious reasons by a doctor that specialized in this particular type of leadless pacemaker. The pacemaker migrated at one point and it resulted in her chest filling with fluid, the removal procedure was very tough on her (leading to other complications on an already frail body) and marked the beginning of the end for her.
"installed for dubious reasons by a doctor that specialized in this particular type of leadless pacemaker"
That's the problem with a fragmented health system. A lot of doctors are specialists in a certain procedure and want to repeat their pet procedure on every patient they see. I saw this with my ex who had chronic leg pain. Most of the doctors we met just wanted to repeat what they do every day (and makes them good money). There are only a few doctors who have a wide knowledge of the area and can recommend a treatment that actually fits.
It's a double-edged sword. For surgeons, specialization improves outcomes as practice makes perfect. Since this is HN, I'll liken it to using the same programming language daily, versus being asked to fix a bug in a language you last used years ago, within a fixed time window.
The only missing peace is the a doctor on the care team who chooses which type of surgery is best, becasue once one is referred to a specialist: they are going to get that specific specialized care.
That doctor on the care team is usually missing. Actually I would guess that most people don’t have a “care team”. They are referred to a surgeon or other specialist that then does their pet procedure.
I don't think that any profession or system with extensive state-imposed restrictions and regulations (education requirements, medical boards, licensing, exams, continuing education requirements, and so on) should be considered "private".
Perhaps such a system incorporates some aspects of private enterprise, but ultimately it should still be seen as a non-private system.
A truly private healthcare system would allow anybody to practice any kind of medicine at any time, without specific requirements and restrictions being imposed by the state.
It would be up to each patient to decide whether or not he or she will deal with a particular practitioner or provider of services. (Of course, this can be complicated in the case of an incapacitated patient, without prior planning.)
As is typical with non-private/socialist/communist systems of any sort, where natural free market balancing forces such as consumer choice, competition, and reputation are limited or absent, we should expect to see the inefficiencies you described, including significant quality variation, bureaucratic games, and misaligned incentives.
Healthcare in America seems very non-private in nature to me. It may not be as overtly non-private as the healthcare systems in Canada or many European nations, but it certainly does not seem to be private, either.
With private healthcare as you describe, poor people are left with subpar care, and you've got a revolution on your hands. At least I would be very prone to violent action in such a system.
I support affordable universal access to basic healthcare regardless of ability to pay, but you're overestimating the impact of the healthcare system. Up until a few generations ago there was no real "system" at all, and treatments frequently caused more harm than benefit. This wasn't a significant cause of violent revolutions.
Even today, organized healthcare beyond vaccines and antibiotics has only a marginal impact on lifespan and healthspan. It's way behind sanitation (clean drinking water, sewage systems, waste disposal) and lifestyle issues (substance abuse, obesity, lack of exercise).
Sure, but not long ago medical care was essentially useless even for rich people. If now, poor people see rich people get cured while they die or suffer major consequences from common illnesses, I don't think that's gonna go well.
My wife for many years has been complaining of bone and joint pain, strange rashes, and other odd symptoms suggestive of an auto-immune condition. I kept telling her to ask her doctor for at minimum an ANA to see if it comes back with anything. Her doctor kept brushing it off as "stress and anxiety" - something I've learned incompetent medical professionals throw around without ever gathering a mental history or referring the patient to a psychologist - and basically refused to do further tests. The GP basically wrote her off as a hypochondriac with zero evidence; no tests, no referrals to a mental health provider, no looking back at her history, nothing. Doesn't "give a whiff of a shit" is spot on.
I finally convinced my wife to see another GP. This one definitely fits the bill of "extraordinarily competent and dedicated" and on that first visit the new GP ordered an auto-immune panel. Well well well, she came back with high ANA and Rheumatoid Factor markers. She's been seeing a rheumatologist since then who has been evaluating her for rheumatoid arthritis and lupus.
EDIT: this is not my only anecdotal experience either. There's so many of them. For myself: over many years I was slowly losing sensation from the neck down along with a myriad of other neurological issues that were strange but had no specific point of origin. Got several MRIs all of which came back inconclusive.
About 3-4 years later the problem suddenly progressed to me running into walls, losing significant sense of sensation, and losing my balance. I visited an orthopedic clinic and got a second set of MRIs (brain and neck), but this time I asked for a CD so I could review the material myself. The radiologist reported no issues, but I used radiopedia and medical journals to study and cross-reference normal/abnormal MRIs with mine. I found an area between C5-C7 on my neck with significant spinal cord compression. The orthopedic doc read off my radiology report and said "your results were normal". I ended up begging her to go through the MRI stack with me just to see it with her own eyes. She reluctantly agreed. We get to the C5-C7 section and she pauses, says "I'll be right back, I need to consult with one of our surgeons" and boom, I was scheduled for surgery for the next month. Unfortunately I didn't make it a month. Over the next week things deteriorated quickly and I ended up having emergency surgery.
It isn't quite as bad as you paint but there is truth in what you say. This is simply Goodhart's Law in action. We've made fee for service the model which means the only way to make money is to see more patients. Pumping patients through as fast as possible isn't conducive to care and you don't have time to give a shit. We need to fix this by moving care to the mid and low levels as much as possible so that we can still pay people and have better care.
That said, there are improvements that can be made. We have very low health literacy in the US and too many providers assume the patients are as literate as them. We need a much stronger teach back style encounter. As above, we need most care done by mid levels. We need to realign incentives among the patients, providers and payers. This is quite broken outside of derm and ortho.
I think the no wiff of shit group can be split in two. Those who never gave a shit, and those who have been beaten down by a system that has normalized profits over patients.
Yes, the treatment runs the gamut, but the overall healthcare system - sans the higher end (naturally) - is more checklist-driver assembly line than patient care. It's not about quality, the system runs on quantity.
Of course things are missed, they were never looked for in the first place. Fwiw, I saw this happen to my father first hand. I dred the day my mother or I have to depend on the healthcare system. It's not all bad. But the odds favor the house.
My parents always warned me, based on their own experience, that you need a healthy family member watching doctors and nurses like a hawk at all times, looking into the drugs and dosages they're prescribing, and even doing work on diagnoses, to make sure they don't kill you, and that it was extremely risky not to have a competent, dedicated advocate of this sort around. I was always a little skeptical of this.
Now that I've had a little more experience with doctors and hospitals, I'd say they were spot-on.
I gather you are not a healthcare professional, hence the artificial dichotomy stemming from your perception as an outsider. Medical competency is very difficult to assess objectively. You're good/bad because people say so. There isn't any form of performance feedback.
Anecdote here, but my friend who is an anesthesiologist (MD) says:(1) anesthesiologists know exactly who the good and bad surgeons are (they see the outcomes and also must manage fluid
loss), and (2) bedside manner by surgeons has no correlation to good surgical technique so patients have no idea.
Am also MD and anesthesiologist. This is partly true. The true part is because anesthesiologists are more able to appreciate medical abilities than the layman is. The false part is that people, including colleagues, tend to judge on what they see and hear. Oftentimes, you won't work every day with the same surgeon, so you actually have no idea regarding success/failure statistics.
That's why I said you're what people say you are. Heck, I don't know my own statistics and I don't know where I stand in comparison to my colleagues. Yet I enjoy a very good reputation with colleagues, but patients often are weary of me because I'm socially awkward.
Judging surgeon competency is much easier than judging physician competency, and physicians make up a much larger share of "medical doctors" than surgeons.
Physicians sometimes jokingly refer to surgeons as "technicians" because what they are doing is often more discretely defined, clearer boundary conditions, clearer indicators of performance/success/quality (especially over X number of cases when you know the national averages for outcomes), and ultimately of a mechanical nature. The work physicians do can sometimes be described similarly, but much less frequently and it can be much harder to do.
The physician/surgeon distinction is a critical one in medicine that is under-appreciated in most conversations about "doctors".
Crucially, if you can get someone to hide which surgeon is skillful and which one is not and view the video and judge for yourself, you might be surprised at your skill in differentiation. In my experience, my friends have 100% guessed which surgeon is better correctly. So it is possible that this is Sorites-style.
From everyday work experience, I strongly believe people to be totally unable to fairly assess a surgeon in an everyday clinical context. Looks, manners and personal sympathies are the main drivers for that kind of judgment.
I know a surgeon with a substantial rate of spectacular failure and a very bad reputation. But most of the time, outcomes are excellent. Is that a bad surgeon on average? No idea. And Im in the OR every day.
FWIW, in this respect, I'm just relaying from my parents: dad's an ortho trauma surgeon who now runs a trauma department, and mum's an ophthal surgeon. As much as I'd like to claim I'm casually reading the NEJM every day, I've only ever read papers they share with me, which they do at an annoying frequency considering I'm in trading.
I assume the general silent belief would be that your guy with the spectacular failure is some kind of "take the tough cases" dude?
By the way, when I was much younger (maybe 10-15 years ago), I recall my dad coming back from an AO Foundation conference where there was a talk about some particular procedure in some Scandinavian country where the range of outcome quality was so high for a particular non-emergency THR/TKR or something like that and so they shut down the procedure in the remote hospitals and centralized it in a few (3?) tertiary healthcare centres. Do you recall something like that? I have this memory in my head but I can't find the original material. The story is so good: EBM driving patient outcomes; but I can't find it so I don't share it and I would like very much to, if it is true.
> I assume the general silent belief would be that your guy with the spectacular failure is some kind of "take the tough cases" dude?
No, the silent belief is that he's the worst surgeon we have.
> they shut down the procedure in the remote hospitals and centralized it in a few (3?) tertiary healthcare centres. Do you recall something like that?
That's become an extremely common strategy for specialized surgery throughout Europe.
GP: This condition X is much worse than normal, so I'm going to refer you to Dr. A, the best I know in the field.
Aggregate enough of this and Dr. A has worse outcomes on paper for condition X than everybody else, since s/he gets the patients with the hardest to treat coming in.
It's not a legal or insurance issue, it's a data gathering and classification issue. There is no such thing as a standard patient. Especially when it comes to surgical procedures, everyone is unique. We generally don't have ways to accurately assess the impact of factors like patient age, fitness level, body composition, and co-morbid conditions. So instituting a quality measurement system like you describe would have the perverse effect of causing physicians to keep their numbers up by refusing to take the riskiest patients on at all.
Insurance, precisely. Insurance companies are pushing for performance assessment to put pressure on prices. All well and good until you understand that you can't force hospitals/practitioners to all get a random sample of patients. So measurements are biased, and therefore people game the system. In practice, it means all healthcare providers will concentrate on easy cases to lower risk and maximize revenue. So as you see, free market thought doesn't really work in healthcare.
There are too many variables. You could have two people with the same diagnosis but one has a complication. Even if you do everything 'right' the success isn't 100% because there are slight variations in people and diseases.
We don't even know what all the factors are, and we don't have a practical or affordable way to gather and record all that data in a high-quality, consistent manner.
It's always disappointing to see comments like this on HN by arrogant, ignorant technologists who haven't done their homework and fail to understand the complexities of healthcare delivery. Biology is a lot harder than software or engineering. There are many researchers working in this area but outside of a few limited areas they haven't succeeded in creating models that are useful for assessing the quality of individual practitioners. And that generally wouldn't even be useful since healthcare is a team activity where systems matter more than individuals.
We can do somewhat better at measuring quality for larger provider organizations across patient populations. The NCQA HEDIS measures are evidence based and reflect current consensus on best practices.
To clarify: We're talking about an industry that only (relatively) recently realized ubiquitous hand santizers was a good idea. And then took additional time to motivated / train staff to use them. Let's keep this in mind.
Suggesting what is available be tracked isn't ignorance, it's innovation and improvement 101. It's a call to prevent waste and unlock societal value. To suggest otherwise is lazy and insanity.
Or perhaps the industry is afraid of something? Fear makes for great excuses.
We do try. But what we've got is not nearly good enough. You underestimate the messiness and overestimate the information retrieval capabilities of healthcare systems.
Oh. I understand the healthcare system's systems are a shitshow. I've seen it first hand. FFS the staff isn't even aware of the importance of sleep (i.e., lights left on, TV left blasting).
But.
If they can identify the importance of hand sanitation and implement programs to ensure that, then there is hope. But as it is, it feels like the land of "that's how we've always done things."
Note: this isn't only hospitals but extended care facilities. They're printing money on the backs of the suffering and losses of others.
Do no harm?? Perhaps it once meant something. No much anymore.
That doesn't mean that they make mistakes, some extremely competent workers just do what they're told. But I don't know of any evidence that hospitals have gone through the same engineered hardening process as heavy industry (eg, airlines) so I expect huge gaps between actual practice and what is possible. I've never seen a hospital process that looks to have been given an engineered finish & I don't think they have enough mathematicians floating around to squeeze for efficiencies.
That sounds like too broad a generalization. In their bid to maximize profits and "increase efficiencies" Medical institutions themselves have a huge impact on outcomes. I wish the medical industry was closer to airlines/FAA, which is more effective and investigating accidents and suggesting systemic fixes, instead of the current culture that suggests "That's a bad-apple/negligent doctor. No need to dig deeper into their load or processes. Case closed"
> Again anecdotally, the medical field seems to only have two kinds of people, the extraordinarily competent and dedicated, and people who don't give a whiff of a shit.
The former will however cover for the latter. Same thing with police unions. That's why it's so hard to prove malpractice.
Also the later group usually gets paid much more because you can do a lot more billable actions or get better performance statistics, if you only pretend to do medicine.
If we just classified things that are obvious. Wrong medication applied (ie paper says to deliver 10mg of Advil but you got a vitamin C tablet), wrong leg amputated, wrong patient data looked at leading to a bad diagnosis. We would likely find a large number of incidents, however no way on earth will you get the American medical cartel to report on them honestly.
Calling doctors, collectively, the "American medical cartel" isn't helpful language. It's just more polarizing politics that breaks down constructive dialogue.
Don't behave like a cartel if you don't want to be called one.
From limiting residency slots, making it nearly impossible to hire foreign doctors and lobbying for safe medications that are sold over the counter in the rest of the world to be sold as prescription only in the US. Just to name a few, the AMA has served it's members well to the detriment of anyone buying healthcare in the United States. That to me is a cartel, they just use a "certificate of need" to enforce it's will rather than a gun.
Anytime someone mentions the AMA I just have to ignore them completely. It’s not that the AMA hasn’t done some shady shit throughout the years (especially the 60s), it’s just that in 2022 they are so close to irrelevant that painting them as a boogieman merely shows that your name information is about 50 years out of date.
The “medical cartel” as you call it is more and more getting shaped by a number or powerful interest groups none of which are the AMA that is pushing somewhat successfully for fewer doctors in most specialities.
Amen. The American Hospital Association wields way more power now. Mergers between hospital systems have led to the large hospital systems being among the largest employers in many states, which means that senators and representatives are going to listen to what they want. Certainly more than they do to the AMA, which might only represent like 25% of doctors anymore.
Please stop spreading misinformation. The AMA doesn't limit residency slots. They have been actively lobbying Congress to increase residency slots, and have even put their own money into it.
They were actively lobbying to reduce residency slots as late as 2012. They only started the program you linked to when became common knowledge that they were attempting to limit the number of doctors.
The AMA (and other medical trade groups) lobbied to limit residency slots back in the 90s. We're still feeling the effects of that in the market today.
The research fixates on the notion of "error/mistake" (because it's more easily quantifiable), but the reality is that most people do not get "optimal" care. In fact, in any given service, in any given industry, no one gets "optimal" care - that's just human nature. Doctors are humans.
It's also completely inefficient. Humans make judgement calls based on available resources. If everyone over 45 received "optimal" care, they would all effectively live in hospitals.
there's a 4X difference between good surgeons and bad surgeons in terms of bad outcomes or complications. And the same study shows that surgeons are actually accurate in terms of assessing who is good and bad but there's really no way for patients to find out before surgery
There is a big difference between a perfectly healthy 18 year old accidentally receiving a lethal dose of fentanyl instead of 400mg of Tylenol AND a 95 year old not having a mole removal rescheduled after the Doctors office was closed for a day unexpectedly. Both things are errors. And both could lead to death.
The article stresses that all medical errors are not necessarily due to bad doctors. It's also bad coordination between different services, miscommunication, etc. Dr. Makary points out various process problems that are not necessarily the fault of any individual doctor.
An important factor to consider any time you see really worrisome things like medical errors or suicide as leading causes of death: counter-intuitively, that can be good news. Every time you solve your #1 highest priority problem, your #2 becomes your new #1. At some point, if you have good-enough solution for a broad-enough range of problems, the only option left is that your biggest problem is now failing to apply those solutions correctly!
By way of example, let's take a look at WHO data[0]. Across Europe, the leading cause of death for children under 4 is congenital anomalies. That sounds bad in isolation, but is actually pretty good in context. First, because the actual rate at which children die from congenital anomalies in Europe is a fair bit lower than the world-wide average, but most importantly because those anomalies float to the top simply by virtue of every other cause of death being exceedingly rare.
This really is similar to the statistic of head protection increasing the number of head injuries. With the explanation being that otherwise, many of those injuries are deaths instead.
Definitely. The important thing is to not hide the facts. In a counter example sometimes a botched procedure can lead to suicide (if the person is in too much pain for example). That should probably count as a fatal medical error but I doubt it would be counted as such.
Thank you Tesla for replacing hardware buttons with menu driven touch screens and autopilot that will drive you home safely. Additionally thank you for causing such a hype that other car manufacturers feel like they missed the boat and are copying these brilliant features. /s
This is insane. These device UIs should be standardized and any manufacturer making a significant UI change to their product should not be permitted to sell it under the same name and number. The FDA should be looking at this before approving such devices for use.
Don't aircraft have more or less standardized controls and displays?
As far as I am aware, for every significant UI change we do at the medical device company I am working for, we bring a panel of final users (doctors and techs) to our facility and we have to do some validation tests, and finally show these validation results to the FDA before releasing the product.
"As far as I am aware, for every significant UI change we do at the medical device company I am working for, we bring a panel of final users (doctors and techs) to our facility and we have to do some validation tests, and finally show these validation results to the FDA before releasing the product."
I also work for a device company and a lot of these panel members aren't really competent at giving good feedback because they lack the technical knowledge. And the FDA mostly just checks for completeness of paperwork. They look at the actual product only once there is a problem in the field.
I am not really involved with these tests: I believe they are not the same people every time, because doctors tend to be quite busy, and don't have much free time available.
They are coming at least from different continents, but obviously to answer the sibling's comment they are selected by the company, from a pool of advanced users or early adopters, so not necessarily a good reflection of the whole population of final users.
And this is an issue in itself because the product is influenced by expert users, but at least, you cannot release a totally horrible or misleading UI that no user has ever seen before.
The other side effect, intended or not, is that you cannot change the UI at every iteration of the product because these tests tends to be costly, as a result, the evolution of the design is quite conservative, which might also explain in part why medical devices UI tends to look "outdated".
"Don't aircraft have more or less standardized controls and displays?"
Commercial pilots (and military) are required to train and qualify on a particular aircraft before being allowed to fly it. Controls can differ, but it's generally the difference in flight characteristics/handling that is the difference. Although we can see that the different computer related controls and lack of training were a factor in the Max crashes.
Ah yes, the training on what to do if the computer suddenly decides to crash your airplane out of the blue.
I feel this framing mis-allocates blame away from Boeing and towards airlines. The MAX was specifically designed to not require a new type rating (they have now introduced one). And no amount of training is a reasonable substitute for the autopilot with a habit of nosediving into the ground because you were too cheap to build fault-checking into your AOA sensor subsystem (an oversight that should have made the MAX fail certification).
> Ah yes, the training on what to do if the computer suddenly decides to crash your airplane out of the blue.
My expectations might be too high, but yes I expect their training to include something along those lines. Perhaps not "computer is trying to kill you" specifically, but "computer is off or malfunctioning". And I would expect the airplane to have a secondary/backup computer at minimum, these systems are supposed to be heavily redundant after all).
"And no amount of training is a reasonable substitute for the autopilot with a habit of nosediving"
If they would know that the system even existed then maybe they would have known how to disengage it.
Yes, the design was flawed. Even then, you train for equipment failures because everything can break. It's hard to do that if your not even told that the MCAS exists.
The training aspect still squarely places blame on Boeing and the FAA for crearing/approving a design that didn't mandate additional training.
My memory is fuzzy here, but wasn't it the case that even with training, you only had a window of a few seconds to disengage the MCAS before the aircraft became unrecoverable? If you waited too long, it cranked the jackscrew to its extreme position, and there was no way to un-crank it with the MCAS disabled (i.e. MCAS had exclusive control over it). That seems like an unacceptably fragile failure mode to me.
The main cause was completely insane specifications. With a reasonable design the plane would not have become uncontrollable under any possible sequence of events, training or no, familiarity or no.
True, but we live in an age where we we trust our lives with computers so that we can do things we couldn't do otherwise. Flight control computers are necessary for most new commercial and military aircraft because they're fly by wire or can be uncontrollable by nature (Max, V22, etc). We're at the point that the FAA and people in general don't care about an inherent instability if the computer can fly it.
The computer was not flying it. The computer was actively fighting the pilot, who was desperately trying to keep it flying, while the computer was forcing it to crash instead.
To be clear: with the computer entirely powered off, the plane would have flown just fine. Not exactly like other 737s, but more like them than anything else.
"The computer was not flying it. ...
the computer was forcing it to crash instead."
Sounds to me like the computer was in fact in control of the plane if it causes it to crash. The "computer" is the MCAS system that sits between the pilot and the actual control surfaces. This is a fly by wire aircraft.
Yes they could have turned the MCAS off and had it fly ok... if they even knew it existed. Although they weren't trained on the different flight characteristics they would experience with it off either.
"with the computer entirely powered off, the plane would have flown just fine."
Just to be clear, if you had a total computer failure or power down, a fly by wire aircraft could not function. I assume you are only talking about the MCAS.
No, 737MAX is not a fly-by-wire aircraft. The controls are physically connected to control surfaces via actual physical cables.
But, the computer can yank on those cables, and it can also run the elevator trim up and down, and does. So, no, I am not just talking about "the MCAS". There is no such thing. That is a software feature that is part of the autopilot system, not any sort of hardware. They could have pulled the circuit breaker on the whole flight computer, and flown without it.
> Don't aircraft have more or less standardized controls and displays?
No, not really. There are similarities in avionics, bare minimum instrumentation and historical conventions but there are type ratings for commercial pilots - they train on each one specifically. Some aircraft are similar enough that they have “common type ratings”. But an Airbus and a Boeing look very different.
They aren't very different. For normal flying, the biggest difference is that Boeings have yokes and Airbus uses sidesticks. The important stuff is all roughly the same.
The yoke/sidestick is really a superficial difference. Though I will also point out that anybody that is good at the 10 differences puzzles will call bullshit on your claim they aren’t very different - in appearance at least - just look at the primary flight display. More relevant, look at the autopilot stack - they not only appear different they have some deeper fundamental differences in operation.
Oh come on. The PFDs are damn near identical. The biggest difference is that the Airbus attitude indicator has rounded corners while Boeing's is squared. That's an entirely superficial difference that changes nothing about how it works or is interpreted.
Yes, as stated, just like with the side stick/yoke it’s a superficial difference - but since you were making your case with pictures - whether they look similar or not really has to do with how detail oriented you are.
Though you ignore the relevant point made about the autopilots. The entire workflow of the flight computers differs. And ask any pilot type certified in both, they typically have preferences.
I’m not arguing these need to be made more standard but there’s a continuum where at one extreme the standardization is so rigid that muscle memory transfers from type to type. That’s not the case here at all.
In other words, there’s no free lunch. Regulations can be incredibly expensive to support and almost always stifle innovation and increase time to bring a product to market.
Sometimes it’s still worth it. The aviation industry has had to deal with this forever because its hard to hide a plane crash. But in the end we have a very reliable mode of transportation despite its inherent risks.
Accidentally overdosing an elderly or critical patient? Hell the nurse that did it might not even know.
I’m sure there are ways to classify products to help with this. Anything where operator error creates a life safety issue, like IV pumps, could have some pretty rigorous controls over UX modifications.
Note that the more rigorous the regs the more likely that a particular treatment or tool will not be affordable and the more likely someone will die because there just isn't enough X to go around.
There isn't a single definite answer, it depends on how expensive the regs are and how much they help avoid damage.
A quick Google search will tell you whether the FDA in the US has similar requirements or instead perhaps medical devices in the US are the unregulated Wild West.
I will remind myself of this when I feel an oncoming panic attack before releasing to production the code that will kill absolutely zero people, even in theory.
Interesting, according to the medium article: 1) the numbers used in the study are wrong 2) they are not representative and 3) the numbers do not represent medical error
From the article:
The problem is that this is very subjective and mostly assessable in hindsight — it’s easy to say on reviewing a person’s chart that they might’ve done better on different medication, but whether that really does constitute a medical error is complex and not nearly as cut-and-dried as the BMJ paper suggests.
Yeah, there are medical errors like "patient got the opposite treatment of what they should have been given" or "gross misdiagnosis" but there's also "didn't get the ideal drug for their case - but got something ok"
Not just flawed argument, but wildly inaccurate data that was then grossly misapplied.
Unfortunately a lot of people on HN will just read the headline and think this is true, when it's very far from being the case. I wish we could link this whole thread to the refutation instead of further propagating this misinformation.
Based off the fact that mistakes that lead to death are so high... If we had statistics on errors that didn't lead to death, the results would be an astronomically larger illustration of incompetence then Just mistakes that lead to death.
Mistakes don't sort neatly by causes into fatal and non-fatal. Some of almost any kind are fatal. Anything that prevented fatal mistakes would eliminate most others.
I’m not in the medical industry, and I’m a Brit and don’t know what the estimated stats are here. But the way I think about this is that medicine must be at least as complex and side effects are at a minimum as hard to anticipate as software development.
Bugs are a fact of life in development, nobody thinks software having any bugs at all is normal or that bugs are an irrefutable indicator of incompetence. That’s because anyone with any knowledge of Tyne subject has developed software and spent a huge part of that time fixing their own bugs.
The way you deal with that is through testing, defensive coding, making software maintainable, etc. You minimise the impact bugs have and try to ensure mistakes are recoverable. You also avoid counterproductive measures like punishing bug reporting, because that will just lead to cover ups and sweeping the problems under the rug.
The last thing we want to do is punish the reporting of medical errors. So yes I absolutely agree, and the way to do that is to build an honest and healthy culture around reporting of medical mistakes.
We already know the most effective way to reduce, and often eliminate, medical mistakes: rigorously applied checklists. They are still used only here and there, years after this was well demonstrated. It is well past time to treat not rigorously following checklists as negligent malpractice.
Medicine depends overwhelmingly more on manually applied procedures that would be automated, in a computer, as patients and staff manifestly are not. So, the same mistakes are repeated again and again. A software fix would correspond to changing a checklist.
Still, it would be no bad thing to use checklists for manual processes in software development, but I have never heard of any.
Exactly. Every time I come to a medical thread on HN I realize how bad even intelligent people are at understanding medical statistics and recognizing when to raise an eyebrow of skepticism. It really does take a lot of study and/or professional training to develop those skills.
Interesting. I would be one of those people who can't fully understand the accuracy of these numbers, so I have to take other peoples' word for it.
Are you part of the medical field yourself? If so then I wouldn't trust anything you say given that if the statistics were true, what you're saying has a high chance of just covering yourself up.
I'm not, but I've spent a lot of time over many years cultivating an understanding of epidemiology and med statistics, especially as they relate to study/trial design. The article linked above points out many of the problems with the parent's study. Medical errors are a problem, but we don't know much about it because of these problems, so I worry, as always, that people will take the wrong message away from headlines like this.
I wonder what the average/aggregate expected loss in years of life is due to this. It's very different to make a fatal error on somebody with a life expectancy of 1 year vs 50 years. This data doesn't seem to be discussed in the article.
The US incentivises blaming healthcare providers in order to recoup costs. I'm sure there are issues with substandard care, but I'm not sure the data can be fully trusted given that putting the blame on a healthcare provider in any way possible is often in the best interests of the patient.
The incentive for retaliation is simply customer unhappiness. Patients don't just sue doctors because doctors are sitting on a pile of money, people don't behave this way. Most People sue because they are pissed.
However if you caused someones death and don't want to go to jail... well that can be swept under the rug, people actually do behave this way.
I believe it. In my 40+ years of dealing with the medical profession, Dr. Google, Dr. Wikipedia, & Dr. Pubmed have wiped the floor with them, time after time. If you get sick, do the research. I can almost guarantee that it will pay off.
And these days, medical research is excellent--if only by volume. I know people here like their canned responses--"small n!", "in mice!", "correlation != causation"--but when you're trying to make a diagnosis, every little clue helps!
I can't honestly attribute excellence to the typical medical practitioner.
People that have lived with a chronic disease have seen this first hand. I had specialist Doctors five up on me in frustration because they just dont know what's wrong with me. I've been to several doctors in 3 different countries during 20 years. You really comprehend that doctors are just humans, not too different from say, an IT professional trying to troubleshoot a failing PC.
Forgetting the measurement issue with no standardized definition of what even constitutes a medical error, this is still a very poor way of framing a change in rank. Is the absolute rate at which people are killed by medical errors higher than it used to be? Is it higher than in comparable countries? Or did the rank change because the old 3rd leading cause of death now kills fewer people? Those are very different pieces of news and this press release doesn't indicate which is the case.
From personal experience and friends and family experience, I actually am surprised it is only 250k annually in the U.S.
> The researchers caution that most of medical errors aren’t due to inherently bad doctors, and that reporting these errors shouldn’t be addressed by punishment or legal action. Rather, they say, most errors represent systemic problems, including poorly coordinated care, fragmented insurance networks, the absence or underuse of safety nets, and other protocols, in addition to unwarranted variation in physician practice patterns that lack accountability.
I respectfully disagree. All these are systemic problems as indicated and need to be solved by a revamp of the whole system. They are not magically going to fix themselves, there is no financial incentive to do so.
It may actually be good news. Depending on how the stats are made.
That a death qualifies as a medical mistake means that proper treatment existed. For example let's say that 50% of people die from "the plague". The plague is lethal and untreatable, no medical mistakes can be made because nothing can be done. One day, a cure is found, cheap and 100% effective, unfortunately, the procedure is tricky and as a result, 1 out of 5 times, it is done improperly and the patient dies.
In the end, 4/5 cases survive, which is great news, is used to be zero, but if we look at the share of deaths because of medical mistakes, it has increased a lot.
Of course, it needs to be fixed, but usually, it is something we know is fixable, better than a disease without treatment.
If a person is depending on medical intervention to stay alive, and the situation is permanent, it seems inevitable that their death will be a "medical error".
He’s apparently so important and busy that he has to make these videos while driving in his super car from point A to point B. 15 years ago, this “content” would have been a blog post, which we could have skimmed in 30 seconds to get the gist of the counter argument. Now it’s a 10 minute video with no notes, buried behind 2 ads. Next time, give us a 2 sentence summarization of the argument you’re linking.
Afaik he does mostly videos and dont do blogs. Anyways one argument presented in the video (but dont know if its relevant to this post) is that the datasets used sometimes dont differentiate between medical error and medical harm. Medical harm is where you need to do a particular intervention, but the possible side effects of the intervention leads to the patients death. Furthermore the line between medical harm and medical error is fuzzy especially when you dont have all the information at hand when doing the procedure. The datasets for the papers mentioned in the video were retrospective
He's a consultant (the most senior and advanced type of doctor you get here in the UK) cardiologist, meaning that he likely works crazy hours (read 60-80 hours per week and above). He's not a professional content creator.
- the claim is absurd on its face: that's more deaths than from road traffic accidents, half as much as from all cancer, and accounts for a significant proportion of the number of people who die in the hospital for all causes (and would be the number one reason
- the figures come from studies with methodological issues (non-representative populations, overly-broad criteria for what's counted as a death from medical error)
The latest Peter Attia Drive podcast episode has an interview with Dr. Makary where they examine the issue of preventable medical errors in more detail.
It "surpassed... respiratory disease, which kills close to 150,000 people per year."
For proportion, according to the CDC[1] there were an estimated 100,300 drug overdose deaths in the United States during 12-month period ending in April 2021
Although this report is from 2016, which was before COVID and mandatory masking. WebMD says that in the interim, deaths from respiratory diseases have also soared, but I couldn't find any numbers.
Lots of these are drug interactions. Have a symptom? Here is a drug for that. Doctors rarely check and often are not aware of the potential for harm. Seniors taking 2-3 dozen drugs regularly are not uncommon and often feel much better when medications are removed from their regimen. Data science should be able to help at both ends connecting problems with drug interactions that trigger them and then warning of possible problems when considering or prescribing medicines.
Yes there was. Questioning of corporations, their motives, or their claims, was verboten. Maybe it still is, someone just accused me of being "far-right" for the thought-crime of questioning Pfizer's profiteering, lmao.
It seems like the bullies and censors and corporatist bootlickers are still far louder. It's great if you're hearing them being called out by those speaking their truth to power though.
Bullying can be very silencing and disempowering actually. I see them as closely related actually -- censorship is bullying and bullying can be silencing, although the simple kind of bullying is usually driven by their more basic fear and embarrassment of their own inadequacies.
Ah yes, the world paid Pfizer about $70 billion for vaccines in the last couple years, and in return those vaccines saved about 1 million lives. $70k/life sounds like we got a pretty good deal to me. Much cheaper per life saved than nearly any other new medical intervention nowadays.
Mark Cuban's project of selling off-brand medication for cheap is great - insurance companies don't want to work with them, but if said insurance companies still require their customers to pay for medication and that project is even cheaper than that, they'll get competed out of the market.
Also that free market thing is a lie, it's in the medical industry's interest to keep prices high and to not have competition / undercutting happen.
>> Mark Cuban's project of selling off-brand medication for cheap is great
It's great and going to help a lot of people. But it's still not cheap compared to a lot of other countries which either provide free medication or medication for a fixed fee per item (usually around $10-20, varies by country).
As someone who lives in a country where medication is "free", I always like to clarify that "free" in these kinds of contexts means "free at the point of delivery" - we're all perfectly aware that they are paid using money from taxpayers.
It's not just negotiating power, but that they're actually allowed to negotiate.
E.g. Medicare has a substantially higher demand for drugs than most single payer universal healthcare systems due to a combination of number of people covered plus covering mostly older, more expensive, patients. It's larger than the UK's NHS for example. But Medicare is subject to various legal restrictions on their ability to negotiate the best prices possible, while the NHS can use whatever leverage it pleases.
The excuse is that they're concerned about it distorting the market. So instead they're distorting it by ensuring one of the bigger customers of pharma companies won't try to get the best prices possible...
> insurance companies don't want to work with them
I'm not familiar with the peculiarities of the US healthcare system, but why wouldn't insurance companies work with off-brand medications? If it costs less it's good for them, no?
Medical insurer's profits are limited to a percentage of their gross income, so, more expensive medications and procedures are preferred to less expensive ones in order to preserve overall income.
I don't understand. If you pay $5000 a year for your medical insurance, but it paid $10k for your medication, how have they made money? If they could purchase the same medication for $2000 instead, surely it would be more profitable for them?
Part of the U.S Affordable Care Act is the so-called 80/20 rule, which limits profit margins to 20%. Therefore, the greater the overall healthcare spend by insurance companies, the greater the allowable profit.
While the Affordable Care Act (Obamacare) did introduce a minimum medical loss ratio, insurers still have a strong incentive to hold down costs by encouraging use of cheaper generic medication so that they can offer lower prices to plan purchasers. Most medical insurance is purchased by employers and other group buyers, and they look closely at costs. In most cases, physicians have to obtain specific approval in order for more expensive medications to be covered.
Who is "they"? Medical insurance companies don't really set deductibles. They offer a menu of options to customers (mainly large employers) and let them pick. Insurers are happy to offer plans with zero deductible, but the premiums will be higher. This is all based on actuarial calculations; insurers generally don't make any more profit on high-deductible plans. If you think your deductible is too high then complain to your employer.
There is a tax incentive at work here as well. Individuals enrolled in high-deductible health plans (HDHP) can often receive a federal income tax break, so you have to factor that in when calculating your net cost.
The configuration of your health insurance is negotiated between insurance and employer. It seems pretty obvious that they are setting things up in a way that’s good for them.
You don't seem to understand how that negotiation works. Higher deductibles aren't good for medical insurers. In terms of insurer profit margin, higher deductibles are neutral to slightly negative. It's the employer (customer) who has complete control over deductibles. That is the reality of how the market works.
Deductibles are generally going higher because healthcare providers and pharmaceutical companies raise prices every year, and because we have an increasingly unhealthy population that demands more services. Insurers try to hold down those increases but with limited success, so they pass those costs on to their customers (employers), and those customers in turn pass on the higher costs to their employees.
No, they're required to spend at a minimum 85% of the money they receive from premiums on healthcare or healthcare quality improvement. As the price of medical procedures goes up, so does the 15% amount that they're allowed to keep and so too do the premiums rise to increase their income.
Exactly - in the Netherlands we have the reverse problem - if a generic drug exists, some insurance companies will only reimburse that drug. In some cases that leads to problems. One example was a generic drug where some patients were allergic to the excipients in the generic drug.
Our healthcare system is an absolute travesty for the poor, but it's pretty lousy even for people who can afford care. I know plenty of people with 7- and 8-figure net worth, people who have connections to the medical system, who've still received substandard care. It's at the point where no one's winning.
Sugar causes excess uric acid in the blood, which causes type 2 diabetes and high blood pressure, among a dozen other ills. And, does not reduce hunger like real food. On a choice between fat and sugar, always choose fat.
To be precise: sugar without enough fiber causes illness. With enough fiber, as is provided by all fruit except grapes, sugar absorption is delayed long enough that intestinal bacteria get to eat most of it. Keep your intestinal bacteria fed; if not, they eat you.
I don't think "age" is a valid cause of death. Old people are more likely to die than young people, but they still die from particular causes that should be distinguished and recorded.
For people of any age it's not obvious how to record a combination of circumstances or a cascade of conditions. For example, a very common fate for old people is to fall, break a hip ("neck of femur"), be admitted into hospital, and then die from pneumonia a week or so later. I think the bone breaks easily in old people often because of vitamin or calcium deficiency. In one case the fall was also indirectly caused by a preceding stroke, so the cause of death was ...?
Why did that fall happen? There is a hidden epidemic of sarcopenia among the elderly. If we trace the causal chain back we often find the root cause was muscle atrophy caused by lack of resistance training and insufficient protein intake (protein digestion becomes less effective as we age). But sarcopenia usually isn't recorded as a contributing cause on the death certificate.
It’s almost as if the medical community should come up with a general term for a death caused by a combination of complications due to circumstances highly correlated with being old.
Uhm, considering our technological advancement in the medical sector I'd argue "age" should be the only valid cause of death. Sadly it isn't for obvious reasons (firearms, car crashes, everything else related to safety etc., hunger, lack of medical equipment and supplies etc. etc. etc.) ... and ironically the article is one of those causes that could and should be easily avoided.
> Uhm, considering our technological advancement in the medical sector I'd argue "age" should be the only valid cause of death.
For this to be true we would have to be able to predict and treat every possible disease and condition, wherever it occurred.
When I had my Ventricular Fibrillation event I was lucky that it occurred just outside the door of a place that had a defibrillator. If it had occurred twenty minutes later I would have been on my own in the countryside. No amount of credible technological advancement would have saved me then.
Oh, and there are quite a few cancers and other diseases for which there is no known cure.
> "No amount of credible technological advancement would have saved me then."
Assuming there was a cause of your fibrillation; maybe advancement in imaging and detection could have spotted worsening rhythm or small damage developing in the heart, or growing chemical imbalance in time before it happened, and given you an Implantable defibrillator[1] in advance, or a hypothetical wearable one, or treatment to repair the damage and avoid it happening.
I sometimes wonder what it would take in terms of FutureTech to "save everyone"; how about if you were alone but concsious, defibrillator tech builtin to every vehicle (electric vehicles at least have some high voltage power already; sensing and cables wouldn't be a big addition to them) would have got a treatment close enough to be useful. If you were unconscious and wearing a watch with ECG features, it could detect you falling over and alert somewhere, a defibrillator drone takes off, today's fast drones can do 50+mph it could have spikes to get through your clothing and be alerting a telemedicine doctor to remote view it...
I suppose we have to rule out teleportation, but humanoid robots are fairly credible, suffusing the planetary surface with nanotech is uncomfortably credible and that possibly changes a lot. e.g. blood additives which carry much more oxygen and can keep your brain alive after your heart stops pumping until help can arrive, or etc.
> maybe advancement in imaging and detection could have spotted
It's possible that I have hidden Long QT syndrome [0]. Mine can be detected following drug provocation using adrenaline but under normal conditions it doesn't show on my ECG. It is undetectable on MRI or angiograms because it is electrochemical rather than mechanical in nature. There are emerging genetic tests that can pick up some of the possible congenital causes.
> If you were unconscious and wearing a watch with ECG features
Current watches with ECG features don't detect my Right Bundle Branch Block [1], let alone the hidden Long QT.
[Edit]
> and given you an Implantable defibrillator
They are quite expensive to implant prophylactically, create a small risk of infection in the heart itself and require explanting every few years for a battery change. I had one implanted after the event.
One of the main purposes of recording cause of death is so that we know what to concentrate on for improving public health. For example, perhaps we could make old people's bones less brittle by adding vitamins to bread, or do something to persuade more of them to eat something other than just bread and jam, or perhaps we could stop them getting infected with pneumonia in hospitals, or somehow make pneumonia more survivable, or ... Putting "age" as the "cause of death" would be about as useful as putting "birth" or "cessation of the vital functions".
I agree and I might add that we probably don't want to normalise the idea that the elderly just randomly die of "old age". That could easily result in a decline of standards in hospitals and elderly residential homes, or it could even give the Harold Shipmans of the world free reign to do as they please: https://en.wikipedia.org/wiki/Harold_Shipman
Everybody dies and we can't deny that, there likely won't be some breakthrough within my lifetime where we conquer death. But we should still be striving to understand the causes, to know whether someone died because of heart failing due to being really old[0], or whether it was down to something preventable or that otherwise indicates a failure in our system
[0] - not a doctor, I don't know what sort of things they normally use for this. Substitute an actual cause of death if this doesn't sound right :)
Fixing old people is incredibly ambitious. How about we concentrate on low hanging fruit first? Young person with broken hand, knife stabbing, or dog bite comes into hospital, gets treated, leaves 2 hours latter. Nobody gets bankrupt, has to sell kidney, nobody waits 20 hours....
This is a little bit naive, I don't think you've really thought a great deal about how healthcare works. You're talking about a basic triaging system in ER (or A&E or whatever it's called in your locale) - this is something that every hospital does already.
It's kind of obvious you're young (20-something or at the very most early 30s), but just remember you're going to be "infected" by old age one day. You'll likely feel a little bit different about the whole old people are just ill and die attitude then. And that's not even to mention the elderly who are alive today who we are responsible for and should feel compassion for. A bit of empathy goes a long way :)
Tempted to use that "always has been" meme template here.
As an ex-doctor, I can tell you that I very quickly realised that at the top of my "surgical sieve" (a mnemonic that helps docs consider categories of differential diagnoses) one should always start with the category "Iatrogenic" before considering other, more textbook-medicine differentials.
And also that for some reason, most doctors seem to treat this category as last on the list, if they think of it at all.
Makes sense, everyone I met in college who said they were pre-med would immediately suffer narcissistic injury and refuse to be professional in any future context when I'd remind them Pitt has no pre-med program then go back to doing whatever I want on the wifi.
(I also met a few folks who took what they needed to do well on the MCAT then also did some practical courses on information science or focused on art so they could have something to do that would let them turn off their brain after work other than binge reality tv, and most of them are doing OK.)
It means that someone would say I am "pre med" as in, they are sure they will one day be a medical doctor. However, not only is that not guaranteed -- you need to do a standardized test called the MCAT, as well as a number of other things.
Further, while at many universities there is a defined "pre-med" major, there is no such major at the University of Pittsburgh, where I obtained my BS.
So if someone who is a Pitt student says "I'm pre med", they're engaging in a best, a bit of stretching of the truth in order to impress people.
When myself, more of a science oriented person, would say "There is no premed at Pitt. What do you study? Biology? Chemisty? Something else?" they would get angry and refuse to interact.
This is known as "narcicistic injury":
>Narcissistic injury and narcissistic scar are terms used by Sigmund Freud in the 1920s. Narcissistic wound and narcissistic blow are other, almost interchangeable, terms.[76] When wounded in the ego, either by a real or a perceived criticism, a narcissistic person's displays of anger can be disproportionate to the nature of the criticism suffered;[12] but typically, the actions and responses of the NPD person are deliberate and calculated.[2] Despite occasional flare-ups of personal insecurity, the inflated self-concept of the NPD person is primarily stable.[2]
I think it's a huge stretch to assume that anyone who simplifies "I am taking all the required courses to position myself to do well on the MCAT and go to med school" with "I'm premed" has a personality disorder. Also, maybe they stopped interacting because what you said is just kind of rude? People really hate it when you "get them" on technicalities.
That's because they're hard to standardize. What qualifies as a medical error? It's easy to go back after the fact and analyze a case where someone has a poor outcome and say 'we should have done this another way'. That doesn't mean there was any indication at the time that the other way would have been better or that the poor outcome could have been foreseen.
I do believe there are a lot of medical mistakes made. I'm a little skeptical that there are legitimately 250k that result in unnecessary death each year. I assume many of those 250k were being treated for serious illnesses that may have taken their lives anyways (not all medical interventions are 100% effective). It would be nice to have a standardized definition, but I'm not sure how that's possible unless we are only talking about treatment received was different than treatment ordered.