Hacker News Reader

Comment by 0xDEAFBEAD | Hacker News Reader

>In his effort to understand hospital safety and risk, he learned these oft-cited projections: more than 200,000 people will die each year from preventable medical errors. He was shocked. Conservatively, these estimates amount to at least one fatal Boeing 747 crash per week, Terblanche calculated.

This reminds me of the book The Checklist Manifesto by surgeon Atul Gawande. The book argues that aviation has achieved such a good safety record largely through the use of checklists, and Gawande describes his attempts to apply them to the field of medicine. Recommended. (Edit: I see that checklists are discussed later in the article; I would still recommend the book, as it has thoughts on how checklists can be applied more effectively.)

View on HN

If you read the complete article it specifically mentions that Sam's condition did trigger some checklists, but those checklists were willfully ignored in favor of expediency.

The problem isn't that there's not enough checklists, the problem is that there's one pilot and he's trying to fly 3 Boeings at the same time from the air traffic control tower.

> These unreliable warnings can lead to “alert fatigue” and, sometimes, a mental habit of discounting them.

This is a problem the airline industry also struggles with and even more checklists is not the answer. A lower workload and better crew resource management is.

I think it's more like Sam's condition did not clearly fit into a checklistable entity. Our heart rate and temperature go up when we have the flu, but we don't all go in to hospital for antibiotics or die at home. Probably they should have done more work-up the second time he came in but as the article points out that could also have been negative. He was probably just too young for checklists built for older people to pick up on his condition.

Also sounds like he was left alone which does not help either. I had similar to Sam, in early sepsis but none of markers were there until I went delirious. Wife picked up on it right away and that time, ER picked up on it.

I can't remember where I read or saw this, but it struck me as the obvious key difference: In aviation, procedures and practices are developed in concert with experts in aviation maintenance, aviation engineering, various parts of system design, and the people who fly the darn planes. In medicine, the lobbyists, politicians, and software companies have political and economic incentives and communication structures quite divorced from the practiced expertise of actual end users, not to mention the people being treated. So you have all these 'best practices' being imposed that have little to do with the sorts of best practices health practitioners would do or want to do or what patients need.

Being a pilot I can attest how important checklists are, and I do advocate for using them in medicine (have practicing relatives, and I have them tired with that).

But maybe you oversimplified the book? (Or the book oversimplified how safety was achieved?)

There are some other 100s of reasons why aviation is safe. Heck, some of them could also be applied NOW: people must rest! I do NOT want to be treated by a doctor doing an idiotic 24hs shift, which is the norm in every country I know of…

There is a whole list of things that can be transferred from aviation to medicine.

Another point I know of is the “handover” of patients. Just as ATC hands over planes from one controller to another, some procedure should warrant the correct transfer of information between shifts. Oh boy I have hear some funny (and some bot at all funny) stories about it.

The long shifts are the most baffling thing to me. AFAIK they're also common in law enforcement.

"Hey these people make life or death decisions. You know what's going to help? Fatigue."

From the outside, it just seems insane.

how much more are you willing to pay? more law enforcement = more taxes

more shifts for docs = more $$ paying for more medical staff.

I can't speak to the police, but there have been a bunch of studies that showed that handoffs between shifts at hospitals is where things go bad. Someone doesn't document they gave an extra 2 cc of a drug to a patient, and next shift gives them more and causes issues, etc.

Basically longer shifts = more fatigue, and the number of errors caused by fatigue were still lower than hand-off related errors.

> more shifts for docs = more $$ paying for more medical staff.

I don't follow. It can't be more expensive to pay 2 doctors for 8 hour shifts than 1 doctor for 16 hours; if anything, I'd expect it to be cheaper (no overtime).

> but there have been a bunch of studies that showed that handoffs between shifts at hospitals is where things go bad. Someone doesn't document they gave an extra 2 cc of a drug to a patient, and next shift gives them more and causes issues, etc.

Hence pushing for checklists so that doesn't happen?

The docs dont get paid per hour, they are salaried, so 2 docs is double the cost of 1 doc.

This is why they are overworked, why pay 2 docs if 1 can do the work, the burnout of the doc is irrelevant as there are more docs to hire after they burn-out.

Do doctors do extra hours per week? I thought they work less days but do this crazy long shifts

sometimes, but extra hours dont get paid extra, so very little incentive to do so. there are many different models for compensation but you can think of it as a 'fixed salary with optional bonuses'.

EDIT to add:

Most places have a base + bonus structure. You get your base salary, and you see patients, for each patient seen you generate 'RVUs' which is how your group/practice generates income ( by billing insurance companies ). Once you generate enough RVUs to cover your base salary, you start accumulating 'bonus' and that gets paid out down the line using whatever formula your employer uses. There is some variation to this but for the most part groups follow a similar scheme.

EDIT #2: This is US centric, i dont know how other countries do it.

Yes I don’t get the comments about salary vs hs. You need the same amount of people. The question is if you have 3 people doing 24hs shifts or 3 people doing 3 8hs shift a day… has nothing to do with more people/salary/money is just organization of work.

Perhaps if we didn't expect superhuman schedules from doctors, doctors wouldn't command as much of a cost as they do now.

From the doctors I know, it seems like most don't get into it for the money, but they put up with it long-term because of the money. If we treated them better and increased supply, they would almost certainly cost less.

Doctors aren't paid by the hour.

One doctor is one salary and one package of benefits. Two doctors is 2x that.

> Doctors aren't paid by the hour.

Not explicitly, but do you think the salary wouldn't change in the medium to long term if the hours changed significantly? Of course, in the short term you can burn out your doctors by making them work longer.

If the supply of doctors wasn’t artificially suppressed as mentioned by comments above, it’s likely that wages would go down. Whether that would make things overall more or less costly isn’t easy to answer.

Doctors are paid in loads of different ways including profit sharing, RVU-based models, and hourly.

Two people take twice the vacation of one person, have twice the healthcare costs, etc. It is almost always cheaper to have 1 person work 16 hours than it is to have 2 people work 8 hours.

Not if you count the cost of errors.

Also, just bring in more affordable doctors from overseas. Have them take a test to qualify.

US doctor comp is much higher than any of our peer states due to industry protectionism. Other industries don't put a cap on training and licensing and haven't been so distorted.

Personally, I haven’t had a US born doctor in ages. In most areas it’s typically Chinese, Indian, or Vietnamese. Nurses it’s almost as common too.

I personally have no issues paying more taxes for better staffing.

> handoffs between shifts at hospitals is where things go bad.

It'd cost more money, but the solution here is overlapping shifts.

The reason shift handoffs go bad is it's usually a singular information dump right as the next round is getting into work mode.

Overlap by an hour, long enough to pair on a round or two, and that information is much more likely to get remembered.

I've been in hospitals a few times for shift changes and there have been a few times I've been the one to inform what the last shift was doing simply because it wasn't communicated.

We do need more shifts and almost as important we need shift overlap.

The doctor is going to have to go home some time - some patients are going to have care needs that get handed off regardless. If they are going home after 8 hours instead of 24 hours, maybe they'll be better at remembering to hand off everything properly to the next person.

Today’s doctor shortage can be traced directly to government policy in the early 80s that lasted for 25 years. They assumed we’d have too many doctors and pressured medical schools to reduce enrollment.

The lobbying by the AMA may go back to the 80s, but the 1997 "Balanced Budget Act" set the limit on residency slots.

Nice background: https://www.washingtonexaminer.com/opinion/1692395/thanks-to...

No, the “limit” is the expectation of free reimbusement by the hospitals. This is lobbying for them to get more socialized gov handouts.

this is vastly complex than aviation. it is like 10 pilots and co-pilots trying to fly 100 planes and simultaneously switching between them. and with everyone overworked due to no mandated breaks.

no amount of checklist would prevent mistakes. we need legislation to limit medical workload, which is unlikely due to the shortages.

> no amount of checklist would prevent mistakes.

So, don't even try?

Why not both?

Autopilot makes such a scenario safe.

>But maybe you oversimplified the book? (Or the book oversimplified how safety was achieved?)

Quite possibly!

>There is a whole list of things that can be transferred from aviation to medicine.

Please recommend more books!

Unfortunately, healthcare is cursed.

The demand is just short of infinite, it requires an extremely specialized and highly capable labor force, and it has piss poor labor productivity forever.

Which is why the staff is stretched thin, as a rule and not as an exception.

Not that there's a shortage of other issues that compound that. But even if those issues weren't a thing? The curse is far too strong.

> The demand is just short of infinite, it requires an extremely specialized and highly capable labor force, and it has piss poor labor productivity forever.

Just because demand (typically) outstrips supply doesn't mean demand is just short of infinite. It just means it's hard to measure the demand. This is just like highway traffic --- you can't know what the demand is when it's all full, you just know there's more demand than capacity/supply.

If you built a crap ton more hospitals, and forced everyone into mandatory service in healthcare for 20 years, I'm sure you'd have more supply than demand. That's a terrible plan, but it would solve the supply problem. You could modulate the mandatory service period to adjust to the needs, and it would still be a terrible plan. :)

Something better would be some steps to address the bottlenecks. How can we attract / train a larger labor force; how can we retain the labor force; how can we increase productivity; something about facilities. Who can make the changes and how can they be incentivized to do it.

I'm outside of healthcare, but here are some armchair ideas. There's a lot of "administrative busy work" that makes everything harder to do; if you ever need to call around to multiple pharmacies to get your meds, there's two problems there: the first problem is that shouldn't need to happen, the second one is that it's amazingly difficult for pharmacies to communicate; it's not uncommon for a physician to order a test and the wrong test is performed, etc ... it's not easy to streamline communications, but it would improve productivity if done correctly. There's also a lot of things that reduce quality of life of healthcare professionals which reduces desire to go into the field and reduces time spent in the field. And of course, there's limitations on the number of residency spots.

The pharmacies issue is a constant problem: patient lives out of town so prescription is sent to his home pharmacy at his request, on the day of discharge he realized his pharmacy is closed and wants them sent to a local pharmacy, of course this always happens at 5pm when you are driving in traffic, the patient is angry because they want to leave but there is not much you can do. This happens very frequently, doesnt matter if you ask ahead of time for the patient to confirm the pharmacy, something inevitably happens.

The other issue is peer to peers and prior authorizations, these take up a significant amount of time and are essentially ways the insurance companies put barriers to care and reduce their costs.

I think some of your ideas could work but good luck getting anything past the politicians, some of these things would be expensive and others would be unpopular to those that donate to the politicians.

For your first example, wouldn't the friction be reduced just by telling patients the business hours of the pharmacies nearby? I hate how this question is always posed, as if I'm supposed to come up with a name and address out of a hat. If it's the middle of the night or Christmas Eve and I'm trying to get medicine for the baby, the provider probably has a better intuition than myself as to which pharmacy will actually serve me. If I ask explicitly, the provider is usually happy to suggest some options. Even a simple web interface listing hours of operation would be better than the current method, where the patient is expected to pick a pharmacy from memory before they even know what medicine they need or how long it will be before they are discharged.

This setup is crazy, as someone from another country.

Why don't you have a unified system for the pharmacies and doctors to tap into?

In my country, if I get a prescription it goes into my card. Then any pharmacy can read the card, see what prescriptions are yet not used, and provide the product (which marks the prescription as covered). Recurring products, like allergy medication or chronic illnesses, become automatically available again after a certain time, like a cooldown. You only need doctor intervention during the original diagnosis and prescription, or after rare issues (like needing an extra prescription because you lost the meds).

I'd have thought this system or a very similar one is universal.

We used to! Doctor would write prescription on a pad, and you could take the script to any pharmacy.

Of course, doctor penmanship is terrible, and we're going paperless, so we've got to digitize. And every doctor's office and every pharmacy has their own system, and sometimes they can talk (but I think there's a lot of faxing behind the scenes)

Of course, you can't know what drugs will be covered, so the doctor has to guess, and if they guess wrong, the pharamacist will want to check with the doctor to see if something else is OK to save you money, but nobody can be reached, ever.

In the United States, the government deliberately creates a shortage of medical residencies through a longstanding cap on federal funding for graduate medical education, primarily administered via Medicare. Residencies represent the essential postgraduate training phase that new medical school graduates must complete to become licensed physicians, yet the vast majority of these positions are financed by Medicare payments to teaching hospitals. This funding mechanism traces back to the Balanced Budget Act of 1997, which Congress passed amid concerns over a perceived surplus of physicians at the time. The act froze the number of Medicare-supported residency slots at their 1996 levels, effectively limiting hospitals to reimbursements for a fixed quota of resident positions without adjustments for population growth or expanded medical school enrollment. As a result, while the number of U.S. medical school graduates has surged by over 30% since the late 1990s to meet rising healthcare demands, the pool of federally funded residency spots has remained largely stagnant, creating a persistent bottleneck that prevents thousands of qualified applicants from advancing into practice each year.

This cap not only constrains overall physician supply but exacerbates shortages in critical areas like primary care and rural medicine, as hospitals hesitate to expand programs without guaranteed reimbursement. Recent legislative efforts, such as the bipartisan Resident Physician Shortage Reduction Act, seek to add thousands of new slots over several years, but until such reforms pass, the 1997 policy continues to throttle the pipeline of trained doctors, leaving patients with longer waits and uneven access to care.

Nothing is stopping the market from paying for this themselves. They’re just cheap.

The market has to get money to pay more. Health insurance is already expensive - raising it to ultimately hire more care givers doesn’t work for most people - do you have the most expensive insurance option or the cheapest?

The market has plenty of money to pay as it is, the highest of any first world country in fact . They just divert it to shareholders / admin instead of patient care.

I'm perfectly willing to believe that US has many, many issues that compound the curse - with some low hanging fruits among them.

But there are numerous countries that aren't US, and don't share US laws.

Do they have medical staff that's not overworked, or a healthcare system that doesn't suffer from a constant labor shortage, long wait times, poor treatment quality, or all of the above?

The root of the issue is deeper than just "US is uniquely dumb".

It looks that way in hospitals but the demand is not infinite. The problem typically is that primary healthcare is unaffordable and or unavailable because they are fully booked and everyone ends up at the ER.

There has been some improvement in terms of long hospital shifts, but there is value in maintaining continuity of care. Research has shown that preventable medical errors are correlated with the the frequency of handover. Proper documentation in the patient chart can help to an extent but there's tacit knowledge that comes from directly observing a live patient which can't be documented in any codified way. So a balance has to be struck in terms of errors due to fatigue versus errors due to care discontinuities.

I've been in the hospital more than once for a week at a time. At no point did I ever see the same doctor more than once in a 24 hour period - from that perspective, it seems irrelevant to continuity of care how long their shift was.

14 handovers is a lot less than 21 when it comes to opportunity for error.

The Checklist Manifesto: How to Get Things Right by Atul Gawande https://www.amazon.com/-/en/Checklist-Manifesto-How-Things-R...

Reply on news.ycombinator.com