Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

FTA: "As hard as the job is, diagnostic accuracy in the E.R. is high overall. But a recent systematic review of published research estimated that 5.7 percent of E.R. patients will have at least one diagnostic error and 2 percent have a setback as a result."

I feel like scrutinizing the industry for a 2-3% error on an obviously difficult problem is exactly why we pay so much in the United States for health care.



2-3% error is too high, this is something we _should_ be scrutinizing, and healthcare _should_ be more expensive if the reason it's expensive is that we're pouring more resources into it to get diminishing returns on reducing mistakes.

Costly healthcare due to scrutiny is not the problem with healthcare in the US. The problem is drug monopolies, medical (mal)practice without a license by insurance companies, and the lack of taxpayer funded healthcare-as-a-right.

We need to create an environment where someone like Terblanche feels comfortable advocating for himself without feeling like he's being a burden on the ER, and physicians don't feel like they're wasting time by investigating seemingly trivial cases. Such a situation exists because we are not pouring enough money into healthcare in this country.


It feels high to me because most ER cases should be obvious i.e. heart attacks, car accidents and strokes etc. So if say 10% of cases are non standard then 2% overall is 20% off that.


Not only is the case mix much broader than you imagine but even the three things you listed all have plenty of nuance at the individual case level.


We don't have an anti-heart attack pill. Medicine hasn't developed the post-car-accident protocol. Strokes vary so much in type that they can go unrecognized by competent doctors for years.


The ER functions as primary care for a large percentage of the population. They see far far more than the kinds of thing a rational person with health insurance thinks of when they think about medical problems they’d go to the ER for.


https://pmc.ncbi.nlm.nih.gov/articles/PMC10121120/

The study performed by AHRQ used incorrect methodologies and datasets to extrapolate its findings.

The ED is far more accurate with a much lower error rate than the study found.


> 2-3% error is too high, this is something we _should_ be scrutinizing, and healthcare _should_ be more expensive if the reason it's expensive is that we're pouring more resources into it to get diminishing returns on reducing mistakes

There will be rapid diminishing returns. It may cost 5x to get to 1-2%. Maybe 10x.


Part of the reason is - people are not machines, its extremely hard to diagnose quite a few situations since every body is a bit unique. Add tons of medications and issues every older person has, within their own unique bodies. Add symptoms like chest or abdominal or head pain which can mean hundreds of conditions, some benign some deadly. Add time pressure to diagnose quickly since that's how medical systems are set up. Wife is a doctor so I can see the perspective from the other side too.

I've had a thrombosis formed in my calf after having a broken leg and using cast. I also caught covid during that time, and from what I've read now I believe it increases temporarily clotting of blood for certain people. When cast was removed, leg was still stiff as wooden plank and ankle didn't bend. I wasn't told to keep the leg higher so I didn't. Some weird mild pain started in the middle of the calf after few days, wife suggested it may be thrombosis rather than stiff muscles or tendons. Went to Switzerland's biggest hospital's ER, got blood tests, they were below limit for thrombosis, so I was just sent home.

Pain didn't go away, luckily my wife considered it suspicious and asked another doctor who is an expert on this to recheck. Voila, thrombosis there.

The cause of miss - ER doctors should have done more than just a blood test (even by their own ER protocols, checked that with wife and her colleagues), echography would have shown blood clot in the veins. If it got dislodged and ended up in lungs, that's a quick death within cca 20 mins, ambulance & CPR usually are not sufficient to keep person alive without major brain damage. Or blood clot goes into brain, cutting off some part of it with similar result. One peer from back home died exactly like that (lung variant, the most deadly one).


I am happy you were able to solve it. One of my friends had exactly this problem, but he just died the next day :(

Sorry to have the need to have shared this, but at least it's been on my mind every time I hear someone take their cast off and experience something similar.


yeah this be part of it. each aircraft is designed to be effectively identical.

yeah yeah there may be some gremlins or bugs in the airframe but in theory it should fly and handle just like each other aircraft.

each human may be wildly different


An alternative way to throw money at the problem: Instead of trying to further improve accuracy, build out space for more ER beds, and implement continuous monitoring of marginal patients.

Or, build devices to send home with patients which allow for cheap, continuous self-monitoring. That might be a legitimate application of AI actually, if you could use e.g. phone camera tricks to measure more health parameters. Even if imperfect, it could still pick up a few patients who should not have been sent home.


This is mentioned in the article, the fundamental problem is a capacity problem. If patients could be moved out of the ER department to hospital wards then there would be a greater ability for the ER department to monitor patients.


This is because beds are artificially expensive because hospitals deploy maximum-feature bed equipment and services, to avoid malpractice claims, and to increase billing.


There is a (somewhat artificial) middle ground in US hospitals where patients can be admitted for observation but still be considered outpatient.

https://doi.org/10.1001/amajethics.2023.901


> I feel like scrutinizing the industry for a 2-3% error on an obviously difficult problem is exactly why we pay so much in the United States for health care

All research I’ve read on this topic finds that it is the US legal system that causes the crazy prices (incentivizing more testing to cover-your-ass and avoid liability etc.)

Many comparative studies on health care cost and quality use the US military as a proxy, as it is free on the condition that you cannot litigate (very coarsely; it is more nuanced).

The costs for treating US military personel is much closer to other countries (while treatment quality remains equal).


That is a factor but a relatively small one. Several US states have instituted limits on medical malpractice liability and that has had only a small impact on total healthcare system costs. It's often the patients themselves (or family members) who insist on trying every possible diagnostic test, and even with the growth of evidence-based medicine we still don't have clear clinical practice guidelines covering many of those situations.


It´s not just the legal system. A lot of US Doctors are typically paid on a piece rate basis, and the medical records systems are extremely fragmented, so there is an incentive to order repeat tests (as you get passed around from specialist to specialist), and no incentive to put the systems in to make that unnecessary.


I believe your points, but give me a source.


> I feel like scrutinizing the industry for a 2-3% error on an obviously difficult problem is exactly why we pay so much in the United States for health care.

If only. Then the outcomes would be better.

The real reason is that it's ostensibly supposed to be a market but the pricing for everything is completely opaque and shrouded in bureaucracy and corruption.


When the consequences are lethal for that 2-3%, that scrutiny is needed.


A set back is not necessarily lethal.

We expect so much from our health care providers, and we sometimes don't appreciate that they deal with a wide array of patients.

Some will come in with a tiny brushing, asking if they are going to die. Others will walk around with a critical condition for days, saying maybe they were a little sore, but they didn't think it was too bad.


[flagged]


Ok Goodhart!

We might need other measures. It might be damn complex. Ideally 100% would die from "malpractice" (or unknown able issues) because people are so healthy and society is so safe there is barely anyone in ER.


I thought so but reading the whole article, it seems like there has actually been mistakes. Several.

Someone who goes to the hospital 3 times and still die because of an untreated disease is not just bad luck.


That 5.7% number must be wrong. Ask anybody with a chronic medical condition.

I've had uncountable number of doctor's visits including ER for 37 years before a proper diagnosis was made.


We pay too much in the US because of the self-imposed medical labor shortage and that results in mistakes by over-worked staff.


Yeah, it always seems crazy to me that in a country that is often so economically liberal and free market, medicine is still run like a medieval guild.


So, your premise is that the US is unique in its lack of checklists in the ER?


Yes. Less scrutiny, more insurance middlemen please.


It's not quite that one either. The big problem is that most people get health insurance through their employer, and then it's the employer choosing it rather than the insured. Otherwise people would choose different insurance and in particular insurance with lower premiums but higher deductibles, and then use the money they saved on premiums to pay out of pocket for things that cost less than the deductible. And then actually insist on getting a real quote and having the ability to compare prices for non-emergency medicine.

So the main problem is employer middlemen. Which happens in significant part because of tax incentives for employers to do that which you can't get if you do it yourself.


>tax incentives for employers to do that which you can't get if you do it yourself.

Is there any problem today that DOESN'T boil down to the government giving preferential treatment to some class or group?


Yes, tons. You choose to focus on the government-aided ones.


the problem is greed.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: