In which a neurologist writes a bad study claiming we are all idiots

[emperor tamarin]

More than 7 million incorrect diagnoses made in US emergency rooms every year, government report finds | CNN

A new study finds that nearly 6% of the estimated 130 million people who go to US emergency rooms every year are misdiagnosed, which translates to about 1 in 18 patients getting the wrong diagnosis.

www.cnn.com

---

Members don't see this ad.

 

[emperor tamarin]

Glancing over this, basically a Hopkins neurologist used data from 3 other studies of middling quality totalling 1700 patients, none in the US, and extrapolates to the entire ED visit population of the United States, to claim were misdiagnosing millions of ED patients and killing hundreds of thousands of them.

Insulting. Someone else on Twitter said it best. Nothing says emergency medicine like being told how to do your job better by people who have never done your job...

Making headlines on CNN and NYTimes and I'm sure elsewhere. Thanks a lot AHRQ.

 

[Dr Mantis Toboggan]

Glancing over this, basically a Hopkins neurologist used data from 3 other studies of middling quality totalling 1700 patients, none in the US, and extrapolates to the entire ED visit population of the United States, to claim were misdiagnosing millions of ED patients and killing hundreds of thousands of them.

Insulting. Someone else on Twitter said it best. Nothing says emergency medicine like being told how to do your job better by people who have never done your job...

Making headlines on CNN and NYTimes and I'm sure elsewhere. Thanks a lot AHRQ.

This job really sucks lol.

 

[GonnaBeADoc2222]

This ostrich is more than welcome to consult on every Neuro complaint in my ED.

 

[Rekt]

What an absolute POS. Both that neurologist and the study.

 

[bearstanley]

this study is pretty terrible, and CNN just did a piece on jake tapper's daughter's delay to appendicitis diagnosis. a lot of common tropes of 'the doctors weren't listening, i knew something was wrong, i asked about appendicitis, i called the administrators and demanded answers' etc etc.

what a terrible climate for emergency medicine. seeing these studies after getting absolutely destroyed by unprecedented patient volume, lack of staffing and bed availability, lack of accepting hospitals and EMS crews for transport, medicare cuts... it's really something to experience.

 

[Birdstrike]

“CNN —
A new study finds that nearly 6% of the estimated 130 million people who go to US emergency rooms every year are misdiagnosed…”

In other words, 94% are diagnosed correctly. Nothing to see here.

 

[EctopicFetus]

F this study.

 

[pkwraith]

This ostrich is more than welcome to consult on every Neuro complaint in my ED.

Yes, the standard of care obviously is any patient with a complaint that contains dizziness will now get a Neurology consult. I am excited to see how the profund increase quality of care and cost and ED throughput with the increase in CT angio and MRIs. Exciting.

 

[Brigade4Radiant]

Oh how quickly the healthcare hero position has fallen. In two months they will say ED utilizes too many resources. NYTIMES made an article about who expensive the ED was in it turned out her mother had gerd

 

[ScepticSkeksis]

I could also tell CNN and the NY Times how to do journalism......but none of them are actually journalists. Just paid propagandists. I'd wager I'm better at their job than they are at mine.

 

[WilcoWorld]

That headline is journalistic malpractice, and thanks Ryan Radeki for the summary tweet - pretty accurate.

 

[thegenius]

I listened to that piece on Jake Tappers daughter and it irritated me to no end. The teenage patient spent three days in the hospital before she got a sonogram, and then by that time it had ruptured. She needed a few drains and didn’t have her appendicitis removed for another 12 weeks.

And all along the parents were begging FOR THREE DAYS to rule out appendicitis. And on the third day they got the scan.

98% of abdominal pain admits having imaging prior, and prob 100% the case given the high profile nature of the family.

There is more to that story

 

[wholeheartedly]

someone should do a study that looks at misdiagnosis rates in the ED in understaffed private equity owned shops vs SDGs. maybe also for ****s and giggles look at it in relation to CEO pay.

 

[wholeheartedly]

I listened to that piece on Jake Tappers daughter and it irritated me to no end. The teenage patient spent three days in the hospital before she got a sonogram, and then by that time it had ruptured. She needed a few drains and didn’t have her appendicitis removed for another 12 weeks.

And all along the parents were begging FOR THREE DAYS to rule out appendicitis. And on the third day they got the scan.

98% of abdominal pain admits having imaging prior, and prob 100% the case given the high profile nature of the family.

There is more to that story

Maybe, maybe not. We all tend to get defensive when it’s our team. But we’ve all also seen our colleagues do epically bad or stupid stuff that makes you want to bang your head into a wall. such is healthcare.

 

[thegenius]

Maybe, maybe not. We all tend to get defensive when it’s our team. But we’ve all also seen our colleagues do epically bad or stupid stuff that makes you want to bang your head into a wall. such is healthcare.

C'mon...15 yo girl admitted for gastroenteritis...three days pass and then imaging shows ruptured appendicitis...and all along per parents they are begging to r/o appendicitis? The patient's daughter is a famous TV reporter on CNN?

are you serious....

I suspect the doctors really wanted to play "doctor" and show how smart they are by using evidenced based medicine, their clinical acumen, decision rules, repeated physical exams, blah blah. Pt was undoubtedly at an academic hospital which made things worse because a half-dozen residents were going through every day. It's just incredulous. Family asks for appendicitis scan EVERY DAY. LOL. And then on day three they get a simple, harmless ultrasound!!!

At a community hospital she would have gotten, minimally, an US in the ED and probably a CT within the first 4 hours.

 

[johnny_bananas]

More than 7 million incorrect diagnoses made in US emergency rooms every year, government report finds | CNN

A new study finds that nearly 6% of the estimated 130 million people who go to US emergency rooms every year are misdiagnosed, which translates to about 1 in 18 patients getting the wrong diagnosis.

www.cnn.com

View attachment 363450

I’m not sure those critiquing this study have bothered to read the abstract, let alone the entire 744 page report written by a group of multidisciplinary experts in research methodology.

The only critique provided of the study is that the authors utilize non-U.S studies. This is only for overall error rates, which is a small component of a huge report that primarily focuses on disease specific error rates that are derived mainly from U.S based studies.

For overall error rates, the authors themselves acknowledge the limitation and state that the estimates are only accurate “if overall rates are generalizable” to the U.S. The authors also state in their conclusion, “estimated ED error rates are low (and comparable to those found in other clinical settings)”.

Yet, those with unmistakably fragile egos take scientific work as a personal attack directed at their profession. Astonishing!

 

[turkeyjerky]

I knew it!

I was waiting for him to chime in. Lol

 

[DrDrummer]

Lots of haters in here. Misdiagnosis is super common. I've missed things, my co-workers miss things, you've all things. This is the nature of our job.

There are definitely strategies and actions that can help us miss things less (which I think should always be the goal) ranging from raising awareness of the phenomenon to clinical decision support tools to doing research like this which shows that certain pathology is typically missed on the index visit. It's inevitable that prudent, reasonable practice will still miss some diagnoses but it should be the aim of our field (which is built on trying to become the best at recognizing and stabilizing serious pathology) to aim to be better as a function of time.

Additionally, Newman-Toker does/has done a ton of great work, much of it relevant to emergency medicine. The hot takes / angry reactions in here seem a bit extreme to me. At least re: the study. The criticism of the Times' description of the results is valid and I agree with it, but it's what I've come to expect unfortunately from NYT on medicine.

 

[Apollyon]

What I would tell pts was that classical presentations of disease were highly advanced in their course, whereas, today, we see them much earlier, and a lot looks like a lot, but, if we let the process come to it's natural conclusion, we would get 1. an accurate dx and 2. a LOT more dead people.

Briefly, these misdiagnoses might, for a large part, just be nonspecific.

And, wait a minute - 250k extra deaths? With 6k EDs, that's 40/year, or about one every 9 days. That could not be hidden, any way you cut it.

 

[thegenius]

I’m not sure those critiquing this study have bothered to read the abstract, let alone the entire 744 page report written by a group of multidisciplinary experts in research methodology.

The only critique provided of the study is that the authors utilize non-U.S studies. This is only for overall error rates, which is a small component of a huge report that primarily focuses on disease specific error rates that are derived mainly from U.S based studies.

For overall error rates, the authors themselves acknowledge the limitation and state that the estimates are only accurate “if overall rates are generalizable” to the U.S. The authors also state in their conclusion, “estimated ED error rates are low (and comparable to those found in other clinical settings)”.

Yet, those with unmistakably fragile egos take scientific work as a personal attack directed at their profession. Astonishing!

johnny is back!

 

[thegenius]

Lots of haters in here. Misdiagnosis is super common. I've missed things, my co-workers miss things, you've all things. This is the nature of our job.

There are definitely strategies and actions that can help us miss things less (which I think should always be the goal) ranging from raising awareness of the phenomenon to clinical decision support tools to doing research like this which shows that certain pathology is typically missed on the index visit. It's inevitable that prudent, reasonable practice will still miss some diagnoses but it should be the aim of our field (which is built on trying to become the best at recognizing and stabilizing serious pathology) to aim to be better as a function of time.

Additionally, Newman-Toker does/has done a ton of great work, much of it relevant to emergency medicine. The hot takes / angry reactions in here seem a bit extreme to me. At least re: the study. The criticism of the Times' description of the results is valid and I agree with it, but it's what I've come to expect unfortunately from NYT on medicine.

Evaluating efficacy of ER care should not measure whether we get a diagnosis correct anyway. It's only the can't miss diagnoses, and thankfully there aren't that many (e.g. maybe a few hundred can't miss diagnoses).

 

[CliveStaples]

I’m not sure those critiquing this study have bothered to read the abstract, let alone the entire 744 page report written by a group of multidisciplinary experts in research methodology.

The only critique provided of the study is that the authors utilize non-U.S studies. This is only for overall error rates, which is a small component of a huge report that primarily focuses on disease specific error rates that are derived mainly from U.S based studies.

For overall error rates, the authors themselves acknowledge the limitation and state that the estimates are only accurate “if overall rates are generalizable” to the U.S. The authors also state in their conclusion, “estimated ED error rates are low (and comparable to those found in other clinical settings)”.

Yet, those with unmistakably fragile egos take scientific work as a personal attack directed at their profession. Astonishing!

Just to be clear, are you asserting that it was "good" scientific work to extrapolate a 0.2% fatality rate from a single study of a sample of patients from a single academic center in Canada where ONE patient died to the entire population of patients in EDs in the United States in order to suggest 250,000 preventable deaths occur annually in the ER?

If so, I really don't think we can have a respectful discussion about this, and the issue doesn't have to do with my "fragile ego," argumentum ad hominem notwithstanding.

 

[wholeheartedly]

Wow. It really is a 744 page reported by AHRQ.

 

[cyanide12345678]

Maybe we should write a paper that 100 percent of neurology patients die a slow painful death as neurologists are pretty much unable to cure anything 😂

 

[Rekt]

Maybe we should write a paper that 100 percent of neurology patients die a slow painful death as neurologists unable to cure anything 😂

Maybe another study showing how "amazing" thrombolytics are for CVA.

 

[johnny_bananas]

Just to be clear, are you asserting that it was "good" scientific work to extrapolate a 0.2% fatality rate from a single study of a sample of patients from a single academic center in Canada where ONE patient died to the entire population of patients in EDs in the United States in order to suggest 250,000 preventable deaths occur annually in the ER?

If so, I really don't think we can have a respectful discussion about this, and the issue doesn't have to do with my "fragile ego," argumentum ad hominem notwithstanding.

Like others, you seem to have a myopic view of the study and are hyper-focused on the estimate they provide on overall misdiagnosis-related death rate, neglecting 98% of the rest of the study. Putting it another way, even if that part is “bad”, does that imply the entire scientific study is “bad”? It really is an impressive synopsis of the best available evidence on the topic.

As to your incorrect claim, the estimate of overall misdiagnosis-related death rate was not just based on a single study, but corroborated by another prospective cohort study and data on preventable deaths measured among ED discharges. Regardless, you confuse limitations in existing data that the authors themselves acknowledge as “bad” science. The authors temper their claims acknowledging limitations and that the estimate is based on best available evidence not strong evidence. Importantly, the value in their work is summarizing existing knowledge gaps. For instance, the authors state that due to the imprecision of the estimate on overall misdiagnosis-related death rate, “large U.S.-based studies using rigorous, prospective ascertainment are needed to validate that estimated error rates reflect current U.S. ED diagnostic performance".

 

[johnny_bananas]

Maybe we should write a paper that 100 percent of neurology patients die a slow painful death as neurologists are pretty much unable to cure anything 😂

Jokes are typically funnier the closer they reflect reality, and I am scratching my head wondering why you think “100 percent of neurology patients die”. Has this been your experience managing patients with neurological issues?

 

[RustedFox]

Jokes are typically funnier the closer they reflect reality, and I am scratching my head wondering why you think “100 percent of neurology patients die”. Has this been your experience managing patients with neurological issues?

THIS guy again?

Bro. If you don't understand that statement, we're not going to explain it to you.

Guys. I seriously had teleneurology order lytics on a stroke alert last week because the patient was drunk and demanding thrombolytics (she said she worked in healthcare and "knows her body", lol) with the neurologist remarking "I ordered it to stop her from demanding it."

Apparently, you can be hammered, throw a tantrum, demand a dangerous drug, and Neuro will give it to you because they're too lazy to discern "drunk with NIHSS of zero" vs CVA.

I'm seriously going to the patient safety board over THIS one.

 

[CliveStaples]

Like others, you seem to have a myopic view of the study and are hyper-focused on the estimate they provide on overall misdiagnosis-related death rate, neglecting 98% of the rest of the study. Putting it another way, even if that part is “bad”, does that imply the entire scientific study is “bad”? It really is an impressive synopsis of the best available evidence on the topic.

As to your incorrect claim, the estimate of overall misdiagnosis-related death rate was not just based on a single study, but corroborated by another prospective cohort study and data on preventable deaths measured among ED discharges. Regardless, you confuse limitations in existing data that the authors themselves acknowledge as “bad” science. The authors temper their claims acknowledging limitations and that the estimate is based on best available evidence not strong evidence. Importantly, the value in their work is summarizing existing knowledge gaps. For instance, the authors state that due to the imprecision of the estimate on overall misdiagnosis-related death rate, “large U.S.-based studies using rigorous, prospective ascertainment are needed to validate that estimated error rates reflect current U.S. ED diagnostic performance".

Your rhetoric is deliberately inflammatory (not that some other posters on this forum aren't equally inflammatory, don't take this to mean I think you're the biggest offender) and it gets in the way of being able to have an honest, respectful discussion.

What's the point/conclusion of the study besides that more research should be done? That 7 million misdiagnoses are made per year, 2.5 million harms that are a result of these misdiagnoses, and 350,000 deaths or permanent disability result from these same misdiagnoses.

I do not have any reason to doubt the misdiagnosis rate. It's a pretty reasonable number, consistent with physicians in all practice settings, and I would say. that most ED physicians would say feel consistent with their practice. The role of the ED is to be sensitive first, specific second. It is backed by the most robust evidence in the entire study. The most helpful parts of the study relate to diagnosis-specific errors that are made, and I would say that it again is consistent with reality (I.e. SEA having such a high misdiagnosis rate, any ER doctor will tell you this disease is oftentimes impossible to detect on initial evaluation).

The evidence on the morbidity and mortality is frankly ridiculous. Look at the range of mortalities they found in their studies. They have studies with hundred-fold differences in mortality. Great, they had one prospective study that supported the 1/500 death in the one Canadian study. Then there's one with 3/400,000, one with 1/14,000, one with 0/10,000, etc. I'm going to be honest, there are two reasons we are all perseverating on this number

1) It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis? I direct an ED in the United States that sees up to 350 patients per day. This is laughably unrealistic.

2) It's dangerous to suggest this number could be true. We owe it to the general public to do better than to publish this idea as a "best guess" that needs to be studied more. We can just say it needs to be studied more because we frankly do not know. Eroding the public's trust with emergency care will only serve to cause more preventable harm; it will not prevent a new 3rd leading cause of death in the United States.

 

[thegenius]

Your rhetoric is deliberately inflammatory (not that some other posters on this forum aren't equally inflammatory, don't take this to mean I think you're the biggest offender) and it gets in the way of being able to have an honest, respectful discussion.

What's the point/conclusion of the study besides that more research should be done? That 7 million misdiagnoses are made per year, 2.5 million harms that are a result of these misdiagnoses, and 350,000 deaths or permanent disability result from these same misdiagnoses.

I do not have any reason to doubt the misdiagnosis rate. It's a pretty reasonable number, consistent with physicians in all practice settings, and I would say. that most ED physicians would say feel consistent with their practice. The role of the ED is to be sensitive first, specific second. It is backed by the most robust evidence in the entire study. The most helpful parts of the study relate to diagnosis-specific errors that are made, and I would say that it again is consistent with reality (I.e. SEA having such a high misdiagnosis rate, any ER doctor will tell you this disease is oftentimes impossible to detect on initial evaluation).

The evidence on the morbidity and mortality is frankly ridiculous. Look at the range of mortalities they found in their studies. They have studies with hundred-fold differences in mortality. Great, they had one prospective study that supported the 1/500 death in the one Canadian study. Then there's one with 3/400,000, one with 1/14,000, one with 0/10,000, etc. I'm going to be honest, there are two reasons we are all perseverating on this number

1) It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis? I direct an ED in the United States that sees up to 350 patients per day. This is laughably unrealistic.

2) It's dangerous to suggest this number could be true. We owe it to the general public to do better than to publish this idea as a "best guess" that needs to be studied more. We can just say it needs to be studied more because we frankly do not know. Eroding the public's trust with emergency care will only serve to cause more preventable harm; it will not prevent a new 3rd leading cause of death in the United States.

Agree with everything you wrote. Some studies shouldn't be published because the conclusions are laughably not believable.

What is SEA?

 

[emperor tamarin]

Your rhetoric is deliberately inflammatory (not that some other posters on this forum aren't equally inflammatory, don't take this to mean I think you're the biggest offender) and it gets in the way of being able to have an honest, respectful discussion.

What's the point/conclusion of the study besides that more research should be done? That 7 million misdiagnoses are made per year, 2.5 million harms that are a result of these misdiagnoses, and 350,000 deaths or permanent disability result from these same misdiagnoses.

I do not have any reason to doubt the misdiagnosis rate. It's a pretty reasonable number, consistent with physicians in all practice settings, and I would say. that most ED physicians would say feel consistent with their practice. The role of the ED is to be sensitive first, specific second. It is backed by the most robust evidence in the entire study. The most helpful parts of the study relate to diagnosis-specific errors that are made, and I would say that it again is consistent with reality (I.e. SEA having such a high misdiagnosis rate, any ER doctor will tell you this disease is oftentimes impossible to detect on initial evaluation).

The evidence on the morbidity and mortality is frankly ridiculous. Look at the range of mortalities they found in their studies. They have studies with hundred-fold differences in mortality. Great, they had one prospective study that supported the 1/500 death in the one Canadian study. Then there's one with 3/400,000, one with 1/14,000, one with 0/10,000, etc. I'm going to be honest, there are two reasons we are all perseverating on this number

1) It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis? I direct an ED in the United States that sees up to 350 patients per day. This is laughably unrealistic.

2) It's dangerous to suggest this number could be true. We owe it to the general public to do better than to publish this idea as a "best guess" that needs to be studied more. We can just say it needs to be studied more because we frankly do not know. Eroding the public's trust with emergency care will only serve to cause more preventable harm; it will not prevent a new 3rd leading cause of death in the United States.

I was going to reply to this johnny guy but you did it more eloquently and politely than I could have. Your two points are great. I'd add a third, which is not directed at the authors of this paper per se but rather at the consequences of authoring a paper that makes poor quality estimates (even while later describing its own estimates as very low quality evidence and limited) - the lay media is simply publishing it almost as fact. The NYTimes article is actually one of the better ones (headline aside). CNN and Yahoos article both simply describe the study as "a government study finds that 7 million people are misdiagnosed in the ER every year - and a quarter million die from it"
They don't get into these nuanced limitations and so forth. The study author is quoted in the CNN article so they definitely talked to him, maybe he caveated and stressed the limitations and they ignored him but I doubt it.

So the net result of this study, which by the description is not really very helpful anyways even if viewed in the best of lights considering the limitations, is a massive lay media coverage that draws overly certain conclusions from these low available evidence studies. It reminds me of that study a few years ago that claimed medical error was the 3rd leading cause of death in the US, which was also laughably badly done analysis....

 

[emperor tamarin]

Breaking: Government report on ER misdiagnoses has "fatal flaw," internal analysis found.

A report which has made shock waves in the US medical world had flimsy methods on key outcomes, which peer reviewers and technical experts worried about prior to publication.

insidemedicine.substack.com

Here is a good in depth analysis of some additional errors - and not only did they read it, they somehow have their hands on the peer review comments which are attached. So maybe alot of the naysayers did in fact read all 700 pages and then some.

Some great points -
250,000 deaths due to ED misdiagnosis is laughable on just face validity. That's like 9 percent of the annual deaths in the United States.

Even if these numbers were accurate a total failure to asses the NET benefit of how EM is practiced. Ie, the approach to diagnostics that allows for a bit of insensitivity (missed diagnosis) is likely preserving some specificity and preventing unnecessary and dangerous over testing.

“The methodology overall is excellent—however, given the substantial heterogeneity of the studies, the variable definitions for diagnostic error, varying including/exclusion criteria, and varying outcome measures make the robust analytic methods completely moot.”
These guys may be great statisticians but they fed their model a bunch of pop tarts and given how low quality the findings are by their own admission, the results are worthless and shouldn't have been published.

 

[CliveStaples]

Agree with everything you wrote. Some studies shouldn't be published because the conclusions are laughably not believable.

What is SEA?

Spinal Epidural Abscess!

 

[Arcan57]

Check out the abstract of the report, it has a lot to unpack. A lot of the disease specific rates comes from malpractice data, and case reports per the abstract, which seems like it would over-represent medical error. The abstract also states that 90% of errors were deficits in clinical judgement which makes sense if you're reading plaintiff's malpractice filings but is going to be instantly out of step to anyone that's familiar with why things actually screw up in the ED.

 

[Janders]

Its a multi-hundred page “study”. Of course it has some good points within it. I’ve sped-read about 20% of it. Absolutely makes some great points, and has good discussion of the difficulty with “diagnosis”. Truly interesting on a philosophical level.

But it leads with the absolutely fantastical allegation that the 4th leading cause of death in the USA is ER Misdiagnosis. Come the hell on. Blatantly impossible by an order of magnitude or two. Top-order result that they place in their abstract. Deep in the text do they place some caveats? Sure. But absolutely irresponsible to lead with such ****ty head-line grabbing hot garbage.

Now if you go diving within the text you find all sorts of gems, written by non-ED clinicians, full of biases and assertions made without evidence.

Its great reading:

(1) You guys above belly-aching about this study are just poor diagnosticians (and old), and your hyper-independence is threatened by the study. The answer is to lay on this couch, talk about your mother, and consult more specialists who know more than you.

(2) All of us who looked at the basic premise of the paper and noted that decreasing “misses” requires additional diagnostic resources and ergo expense but also the potential harm of extra testing, incidental findings, etc… we’ll we are wrong. You see, if we just had “basic diagnostic methods” i.e “proper history taking and neurological examination”, we would increase sensitivity and specificity at any given threshold. Less false negatives. Less false positives. Fitter! Happier! More productive! Comfortable! Not drinking too much!

(3) Turns out we are actually excellent at heart attacks and don’t need help with that. Strokes are the issue. Especially dizzy strokes. What we need is a pathway. Just follow the pathway boys and girls. Its so simple a child could do it. You need to learn HINTS you idiots. We’re going to make an app for that. Just hold your iPhone up to the patient with your dizzy app turned on, and follow the pathway. No more missed stroke deaths. Then we just expand on our success and create pathways for meningitis and mesenteric ischemia. Simple.

 

[Kryptocoin]

A 94% success rate is pretty darn great. Any other professional body with that success rate? I think only the DOJ lawyers are higher, but that's because they screen their cases and only take the strong ones to trial. We don't get to screen patients in the ER. We see all comers.


The 94% success rate was probably pre-pandemic numbers. My ER volume has doubled, and the nursing staff has been slashed in half. So I expect more increase in M&M. That is why when a recruiter calls me to cover any shift, I tell them $400/hr or GTFOH. If I'm going to risk my license practicing American EM, I'd rather be rich than broke.

 

[WilcoWorld]

Its a multi-hundred page “study”. Of course it has some good points within it. I’ve sped-read about 20% of it. Absolutely makes some great points, and has good discussion of the difficulty with “diagnosis”. Truly interesting on a philosophical level.

But it leads with the absolutely fantastical allegation that the 4th leading cause of death in the USA is ER Misdiagnosis. Come the hell on. Blatantly impossible by an order of magnitude or two. Top-order result that they place in their abstract. Deep in the text do they place some caveats? Sure. But absolutely irresponsible to lead with such ****ty head-line grabbing hot garbage.

Now if you go diving within the text you find all sorts of gems, written by non-ED clinicians, full of biases and assertions made without evidence.

Its great reading:

(1) You guys above belly-aching about this study are just poor diagnosticians (and old), and your hyper-independence is threatened by the study. The answer is to lay on this couch, talk about your mother, and consult more specialists who know more than you.

View attachment 363521

(2) All of us who looked at the basic premise of the paper and noted that decreasing “misses” requires additional diagnostic resources and ergo expense but also the potential harm of extra testing, incidental findings, etc… we’ll we are wrong. You see, if we just had “basic diagnostic methods” i.e “proper history taking and neurological examination”, we would increase sensitivity and specificity at any given threshold. Less false negatives. Less false positives. Fitter! Happier! More productive! Comfortable! Not drinking too much!

View attachment 363522
(3) Turns out we are actually excellent at heart attacks and don’t need help with that. Strokes are the issue. Especially dizzy strokes. What we need is a pathway. Just follow the pathway boys and girls. Its so simple a child could do it. You need to learn HINTS you idiots. We’re going to make an app for that. Just hold your iPhone up to the patient with your dizzy app turned on, and follow the pathway. No more missed stroke deaths. Then we just expand on our success and create pathways for meningitis and mesenteric ischemia. Simple.

View attachment 363523

Appreciate the Radiohead reference you snuck in there.

 

[Dred Pirate]

Jokes are typically funnier the closer they reflect reality, and I am scratching my head wondering why you think “100 percent of neurology patients die”. Has this been your experience managing patients with neurological issues?

in all honestly 100% of all patients die

I mean 100% of people who drink water die.

 

[johnny_bananas]

Your rhetoric is deliberately inflammatory (not that some other posters on this forum aren't equally inflammatory, don't take this to mean I think you're the biggest offender) and it gets in the way of being able to have an honest, respectful discussion.

What's the point/conclusion of the study besides that more research should be done? That 7 million misdiagnoses are made per year, 2.5 million harms that are a result of these misdiagnoses, and 350,000 deaths or permanent disability result from these same misdiagnoses.

I do not have any reason to doubt the misdiagnosis rate. It's a pretty reasonable number, consistent with physicians in all practice settings, and I would say. that most ED physicians would say feel consistent with their practice. The role of the ED is to be sensitive first, specific second. It is backed by the most robust evidence in the entire study. The most helpful parts of the study relate to diagnosis-specific errors that are made, and I would say that it again is consistent with reality (I.e. SEA having such a high misdiagnosis rate, any ER doctor will tell you this disease is oftentimes impossible to detect on initial evaluation).

The evidence on the morbidity and mortality is frankly ridiculous. Look at the range of mortalities they found in their studies. They have studies with hundred-fold differences in mortality. Great, they had one prospective study that supported the 1/500 death in the one Canadian study. Then there's one with 3/400,000, one with 1/14,000, one with 0/10,000, etc. I'm going to be honest, there are two reasons we are all perseverating on this number

1) It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis? I direct an ED in the United States that sees up to 350 patients per day. This is laughably unrealistic.

2) It's dangerous to suggest this number could be true. We owe it to the general public to do better than to publish this idea as a "best guess" that needs to be studied more. We can just say it needs to be studied more because we frankly do not know. Eroding the public's trust with emergency care will only serve to cause more preventable harm; it will not prevent a new 3rd leading cause of death in the United States.

“Your rhetoric is deliberately inflammatory”
In your first reply you write that if I don’t agree with your point then we can’t have a respectful discussion. Yet, you accuse me of being “deliberately inflammatory” because I wrote: “you seem to have a myopic view of the study” “incorrect claim” “you confuse limitations”?

Tone can be hard to disentangle in writing, I do appreciate your reply and agree with some of the points you are making. I am always interested in having a respectful discussion regardless if someone agrees with what I write.

“What's the point/conclusion of the study besides that more research should be done?”
How do you know in the first place more research needs to be done if you don’t formally do a systematic review? This becomes more important when you are trying to obtain funding for an area you are claiming needs more research, you need to convince people of this claim with evidence.

“It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis?”
It seems that you disagree with the ‘best available evidence’ because it is contrary to your personal experience. As you know, anecdote is not evidence, though I agree that in can be valuable it providing a sense that the data is off. It seems that you agree, that we need more evidence since the current evidence is not stellar and we don't know what the 'true' estimate is.

“We owe it to the general public to do better than to publish this idea as a "best guess" that needs to be studied more.”
This point seems to be major issue for most people. Should we not publish the ‘best available evidence’ because it runs the risk of being sensationalized and misconstrued by the media? I do agree with your point, and it can be argued that the claims could have been more tempered regarding the supporting evidence, though the authors did acknowledge many of the limitations and it seems the bigger issue is the fact that the NYT wrote an article. Is that the authors fault?

 

[Rekt]

A 94% success rate is pretty darn great. Any other professional body with that success rate? I think only the DOJ lawyers are higher, but that's because they screen their cases and only take the strong ones to trial. We don't get to screen patients in the ER. We see all comers.


The 94% success rate was probably pre-pandemic numbers. My ER volume has doubled, and the nursing staff has been slashed in half. So I expect more increase in M&M. That is why when a recruiter calls me to cover any shift, I tell them $400/hr or GTFOH. If I'm going to risk my license practicing American EM, I'd rather be rich than broke.

Agree. How about they also run the headlines again with how the ED provides 50% of all medical care in the US as well.

 

[southerndoc]

Agree. How about they also run the headlines again with how the ED provides 50% of all medical care in the US as well.

That's not as catching as "ED Physicians Are the 4th Leading Cause of Deaths in America"

 

[Kryptocoin]

Agree. How about they also run the headlines again with how the ED provides 50% of all medical care in the US as well.

Right. I mean, don't get me wrong, there's absolutely room for improvement in EM. We saw how the airline industry came together in a concerted effort, in addition to an act of Congress, to make it the safest mode of transportation. It wasn't always that way.

But, unfortunately, this article is more blame shifting and less solutions. So, in the end, nothing changes. In fact, I expect things to only worsen as the nursing exodus continues.

 

[RuralEDDoc]

All of this is mind boggling. Most of my patients are seen and treated in a crowded noisy hallway. Patients are sent from PCPs for emergent studies that aren’t emergent and so I spend a good deal of my limited time on damage control between expectations and emergency room capability. Drug users come to the ER for preventative treatment because their friends are dropping dead and the county doesn’t have an appointment for two months.
I’m watching my clinical career burn down to the ground here. Nationwide our ERs are a damn mess. What the hell is this report even talking about? Improve stroke pathways? Seriously what kind of advice is that for hallway Pete and his stinky feet?

 

[namethatsmell]

“It's nonsensical. One in 350 ED patients is dead from an ED misdiagnosis?”
It seems that you disagree with the ‘best available evidence’ because it is contrary to your personal experience. As you know, anecdote is not evidence, though I agree that in can be valuable it providing a sense that the data is off. It seems that you agree, that we need more evidence since the current evidence is not stellar and we don't know what the 'true' estimate is.

"Best available evidence"-- two errors with three words, nice. Not only is nothing about this paper is "best available," nor is it "evidence."

And it's always delightful when somebody seeks intellectual refuge with a bogus proclamation of "until there's "data" I can't say X is incorrect" in cases where the likelihood of the claim in question being true is essentially zero. Tends to come from folks who are honestly naive, can't get past their own preconceived notions/biases, or just have an ax to grind.

The entire point of the face-validity test is to help us separate the wheat from the chaff wrt to the perpetual heap of tests, data, studies, claims, etc being thrown at us which can potentially change our decision making. Pragmatically, it can also help give us an idea if we really do need more data or if what we have is good enough to go on. And most people who actually do stats, EBM, see enough patients in their own practice, etc, actually understand this.

As others have said, the claim is absurd.

Mr/Ms. Bananas, if 1/350 out of the 100-130million folks who go to an ER each year die from a misdiagosis, ask yourself:
-Do you think anybody would still go to an ER? At that rate, many Americans would know somebody (or at least knows somebody who knows somebody) who died "because" of the ER. But instead of visits going down due to a populace fearful that their local ER will kill them, "best available evidence" shows us that the rate of use has consistently gone up over time.
-How could EM physicians get medmal insurance needed to practice? Nobody knows the claims numbers better than the insurance companies, and there's no way their actuaries would recommend issuing policies to a field with such a costly/high-risk profile.
-What hospital would want to have an ER? They'd be too much liability to afford.
-How would any lawyer possibly be unemployed and not a millionaire? They'd have so much work, it'd probably allow for conditions to arise for legal midlevels to be born.

Even as click-bait, this piece is laughable.

 

[Arcan57]

I think the author’s whole point was to create this “dizziness is dangerous” paradigm and support that dizziness needs pathways (and reimbursement) similar to STEMIs. Never mind that there’s all sorts of data on benefit of PCI in STEMI. I have yet to see high quality evidence that posterior circulation CVA pts are any less f&@$ed because a neurologist told us to get an MRI STAT.

 

[namethatsmell]

All of this is mind boggling. Most of my patients are seen and treated in a crowded noisy hallway. Patients are sent from PCPs for emergent studies that aren’t emergent and so I spend a good deal of my limited time on damage control between expectations and emergency room capability. Drug users come to the ER for preventative treatment because their friends are dropping dead and the county doesn’t have an appointment for two months.
I’m watching my clinical career burn down to the ground here. Nationwide our ERs are a damn mess. What the hell is this report even talking about? Improve stroke pathways? Seriously what kind of advice is that for hallway Pete and his stinky feet?

Don't worry, after awhile you'll become a little numb to hot-garbage stuff like this.

Whenever one of these hit pieces come out, I remind myself of a small shop where I briefly worked. Place literally had blood stains on the curtains and always short on RNs (way before the recent severe shortage) which predictably caused long wait times. Basically every patient would complain about one or both of these 2 things. Then one day a new hospital VP of whatever came down and lectured us that all our problems (really, their problem of low pt sat) could be solved with AIDET. Lol. When we asked about getting more nurses to actually make things safer/better for patients we got crickets. And when we asked about the curtains they said they could only replace ~50% of them due to the "budget."

Basically, you come to appreciate that almost nobody has any idea of what the hell is going on (and actually matters) in HC. Especially the muppets who've don't treat patients or only see them in a well-resourced ivory tower.

 

[namethatsmell]

I think the author’s whole point was to create this “dizziness is dangerous” paradigm and support that dizziness needs pathways (and reimbursement) similar to STEMIs. Never mind that there’s all sorts of data on benefit of PCI in STEMI. I have yet to see high quality evidence that posterior circulation CVA pts are any less f&@$ed because a neurologist told us to get an MRI STAT.

There's nothing wrong with that intention. But he seems to go quite far afield to so and drowns out that message.

 

[johnny_bananas]

"Best available evidence"-- two errors with three words, nice. Not only is nothing about this paper is "best available," nor is it "evidence."

And it's always delightful when somebody seeks intellectual refuge with a bogus proclamation of "until there's "data" I can't say X is incorrect" in cases where the likelihood of the claim in question being true is essentially zero. Tends to come from folks who are honestly naive, can't get past their own preconceived notions/biases, or just have an ax to grind.

The entire point of the face-validity test is to help us separate the wheat from the chaff wrt to the perpetual heap of tests, data, studies, claims, etc being thrown at us which can potentially change our decision making. Pragmatically, it can also help give us an idea if we really do need more data or if what we have is good enough to go on. And most people who actually do stats, EBM, see enough patients in their own practice, etc, actually understand this.

As others have said, the claim is absurd.

Mr/Ms. Bananas, if 1/350 out of the 100-130million folks who go to an ER each year die from a misdiagosis, ask yourself:
-Do you think anybody would still go to an ER? At that rate, many Americans would know somebody (or at least knows somebody who knows somebody) who died "because" of the ER. But instead of visits going down due to a populace fearful that their local ER will kill them, "best available evidence" shows us that the rate of use has consistently gone up over time.
-How could EM physicians get medmal insurance needed to practice? Nobody knows the claims numbers better than the insurance companies, and there's no way their actuaries would recommend issuing policies to a field with such a costly/high-risk profile.
-What hospital would want to have an ER? They'd be too much liability to afford.
-How would any lawyer possibly be unemployed and not a millionaire? They'd have so much work, it'd probably allow for conditions to arise for legal midlevels to be born.

Even as click-bait, this piece is laughable.

“Nor is it evidence”

I disagree, the authors estimate is based on 2 prospective cohort studies, and data on preventable deaths measured among ED discharges that all provide a similar estimate. Now you could argue that there are limitations and the estimate may not be accurate, I agree with that, but suggesting it is not evidence is incorrect.

What is you estimate based on? Your personal experience? Anecdote is not evidence. Until there is better evidence the ‘best available evidence’ would still be the 2 prospective cohort studies and not your personal experience.

As to your suggestion that the estimate is absurd because with such high rates EM physicians would not be able to obtain medical insurance or it would prevent patients from going to the ER. The former point assumes all of the misdiagnosis would lead to litigation and the rates are incredibly higher compared to any other speciality and the cost to insurance companies would be too high. Do you have any data on this? For your latter point, I don’t follow your logic, why would patients stop going to the ER with a 0.2% case fatality from diagnostic error, the benefit of going to the ER would greatly outweigh the harms from diagnostic error.

 

[johnny_bananas]

I think the author’s whole point was to create this “dizziness is dangerous” paradigm and support that dizziness needs pathways (and reimbursement) similar to STEMIs. Never mind that there’s all sorts of data on benefit of PCI in STEMI. I have yet to see high quality evidence that posterior circulation CVA pts are any less f&@$ed because a neurologist told us to get an MRI STAT.

Here are some studies on the benefit of endovascular thrombectomy in basilar occlusions.

https://www.nejm.org/doi/full/10.1056/NEJMoa2207576

https://www.nejm.org/doi/full/10.1056/NEJMoa2206317