> working those extra 3 to 4 hours per day might mean you're making enough money to retire in your early 40s.

The only way you could retire in your 40s as a doctor these days is if you were already independently wealthy before becoming a doctor.

If you go into medical school today, you're looking at finishing off your student loan payments in your 40s (depending of course on which school you go to, where you do your residency, whether you do a fellowship, and what specialty you choose).

The pre-tax, pre-insurance, pre-expenses, gross pay looks high on paper, but the lifetime ROI on going into medicine is nowhere near as good as what you describe, no matter how many hours you work.

That is a gross generalization. It is true in the US if you are a primary care physician or one of the lower-paid/lifestyle specialists. It is far from true if your are a higher-paid specialist or sub-specialist.

Also consider that unlike tech, where you get paid more for working in higher-cost-of-living cities -- in medicine you get paid more for working in lower-cost-of-living cities.

Also note that unlike tech, the peaks and valleys are much less pronounced.

> Also note that unlike tech, the peaks and valleys are much less pronounced.

The point is that, unlike tech, you have to take on massive amounts of debt before you can even begin to work[0]. And even after that, there's a really high latency before you're earning enough to make more than the minimum payments. Six-figure debt compounded over a decade or more really adds up.

> That is a gross generalization. It is true in the US if you are a primary care physician or one of the lower-paid/lifestyle specialists. It is far from true if your are a higher-paid specialist or sub-specialist.

It's actually more true if you're a specialist or subspecialist. People who don't practice medicine tend to overestimate what specialists get paid, and to underestimate what PCPs get paid. It also doesn't help that, in the last ten years, the relative reimbursement rates have changed dramatically, and to top it all off, private practices have been rapidly going the way of the dodo - so it's understandable that the public perception would be out-of-date.

Overall, reimbursement rates for specialists have been slashed far faster than they have for primary care (with some specialties hit harder than others). Combine that with the fact that they have to spend even more time in training (during which time their debt compounds

> Also note that unlike tech, the peaks and valleys are much less pronounced.

Quite the opposite - tech salaries are relatively flat and stable compared to most other industries, and we get to take home pretty much 100% of our salary, less taxes. Yes, the industry as a whole has booms and busts, but so does literally every private industry. In medicine, unlike tech, even if you're salaried, you're oftentimes paying for a large chunk of your business expenses out of your "salary". This is one of the reasons why looking at listings of what doctors make is incredibly misleading. The insurance rates vary dramatically between fields of medicine (and practice locations), as does the necessary equipment and mandatory continuing medical education expenses (again, even salaried doctors will oftentimes have to cover these out-of-pocket).

In tech, we don't have to worry about any of that. We have it pretty easy.

> in medicine you get paid more for working in lower-cost-of-living cities.

This used to be the case. It still is, for some fields like emergency medicine. It's a lot less the case in other fields, partly because of the consolidation of practice groups and the downward pressure on reimbursement rates across the board.

[0] Unless, as I said in my previous comment, you're independently wealthy beforehand, in which case this whole discussion of early retirement is moot.

With respect - you are comparing apples vs oranges. You are comparing independent/priv-practice doctors vs salaried technologists. If you had an independent contractor technologist, they also have to eat all the overhead (Liability insurance, E&O insurance, benefits, office, supplies.) Similarly, if you have a doctor working at the hospital, just like a salaried technologist they don't have to pay the overhead.

To be fair, doctors' overhead is higher for certain practices, but remember that facility fees can be billed back to the insurance or directly to the patient and even sent for collection. Those are extra fees on top of the doctor's pay -- essentially a form of rent forwarded to the patient for the doctor's facility costs.

Finally, we're not comparing tech salaries to other industries, we're comparing them to doctors. I've rarely seen doctors face a 2001 market crash or 2008 market crash and mass unemployment or had their entire training wiped out when JE22 went south. Tech is very volatile. Also consider that doctors generally make more as they age, whereas technologists often face increasing age discrimination. Yes, MLK in LA and St Vincent's in NYC did shut down, but have you ever seen the equivalent of a Lucent or Lehman or IBM or Intel in medicine? Rare...

You are right that I don't practice medicine, but I work 75hrs a week in hospitals since my startup sells to hospitals. I've seen the relative function and dysfunction of both American and single-payer overseas systems. And BTW, i pay gigantic liability insurance bills for my diagnostic technology product , and it effectively comes out of my salary.

I do think doctors have a hard and stressful job but they are mostly well compensated. A while back I was wondering who were owning the most expensive homes in our city. Turned out 7 out of 10 were doctors and the others made money being a CEO/executive of a tech company. I was expecting more technology/engineering people given that is is in the middle of the bay area.

Are they sleeping 3 hours a day every day? Or is that just their average when they get a shitty on-call rotation?

I think both cases are ridiculous, but it's worth separating the two cases.

But would it be better for the patients if they didn't work that extra? Waiting months for surgery is not good either.

It is usually not because of doctor-shortages that doctors work long shifts -- it is because hand-offs create errors themselves. Think of it as a game of telephone. 8 hour shifts means 3 hand-offs. 12 hour shifts mean 2 hand-offs. One is 50% greater risk of mis-communication as the patient shifts from one doctor's shift to the other.

During training, the long hours serve a double-purpose -- compressing more years of training into fewer. If residencies were 1.5x longer, more doctors would choose not to be doctors. Residents working 80 hours vs 40 helps compress the training time and let doctors get on with their lives and careers faster. Unlike in tech, medical residents get paid little for their training period. Shifting 3-6yr residencies into 6-12 year residencies would make residencies viable only for already-wealthy individuals who can withstand 4yrs undergrad + 4yrs med school + 6-12 yrs of low salaries in a captive residency.

(btw, w/r/t/ shortages, it is probably due to the doctors themselves -- specialty boards such as the Board of Radiology essentially throttle the number of available training positions (residencies) per year creating artificial scarcity and keeping up their own pay.)

> It is usually not because of doctor-shortages that doctors work long shifts -- it is because hand-offs create errors themselves. Think of it as a game of telephone. 8 hour shifts means 3 hand-offs. 12 hour shifts mean 2 hand-offs. One is 50% greater risk of mis-communication as the patient shifts from one doctor's shift to the other.

So long shifts are due to errors ? I hardly see the logic in that... We work long shift because we are expected to work long shift, because nothing is done to prevent medical exhaustion, and because there is a strange societal archetype of the never-sleeping doctor that goes against any recent (or old) research on learning, stress management and work performance (among other).

> During training, the long hours serve a double-purpose -- compressing more years of training into fewer.

That is also incorrect. In every country except maybe Switzerland, residency programs last 3-5 years, whatever the number of hours worked per week, and medical errors are not the first, the second nor the 3rd cause of death in those countries... Cortisol is a brain cell killer, stress reduces our ability to learn and adapt, sleep deprivation reduces both encoding and storage of long term memory, and most research on attention shows that you cannot work more than 5 full hours per day with a NORMAL lifestyle and sleep cycle. BTW 80 hours is on small weeks in surgery, at least it was for me.

> If residencies were 1.5x longer, more doctors would choose not to be doctors.

I believe this also to be incorrect, there is no shortage of applicant to medical schools. Actually it would be a problem to finance residency programs..

> So long shifts are due to errors ? I hardly see the logic in that...

There is no logic in that, i'm not sure why you asserted that. I noted that long shifts reduce handoff errors.

Consider two scenarios: Scenario 1: 8 hour shifts: Doctor 1 hands off to Doctor 2 at end of 8 hrs Doctor 2 hands off to Doctor 3 at end of 16 hrs Doctor 3 hands off to Doctor 4 (or perhaps 1) at end of 24hrs

Each handoff is a game of telephone. In theory it should not be, but in reality it is. Details are missed, directions mangled, and risk accrued. Just like a technical design document going from a product manager to the development lead to the developer...things go wrong in the process.

Just like with tech teams where each added person creates overhead and waste from communications overhead and mis-aligned understanding of requirements, hospitals face the same issue. Unlike technology, you dont just push a fix and rebuild -- an error such as forgetting to record a symptom on the binder/EMR could mean mis-diagnosis or death. Shifting to 12-hours shifts means just two handoffs instead of three. You dont eliminate errors, but you reduce them. In some sense, it may be better to have 4 days of 12-hour shifts than 5 days of 8 hour shifts.

Of course, this is just one aspect of medical errors. There are others which may be brought on by increased shift lengths. We'd need to look at the total sum.

Regarding your other point: > In every country except maybe Switzerland, residency programs last 3-5 years, whatever the number of hours worked per week,

That makes no sense. If that were the case, residents would choose to work an hour a week and easily graduate in 3-5 years. But that would make for a poor doctor, as they need a certain # hours of training.

> There is no logic in that, i'm not sure why you asserted that. I noted that long shifts reduce handoff errors. Consider two scenarios: Scenario 1: 8 hour shifts: Doctor 1 hands off to Doctor 2 at end of 8 hrs Doctor 2 hands off to Doctor 3 at end of 16 hrs Doctor 3 hands off to Doctor 4 (or perhaps 1) at end of 24hrs Each handoff is a game of telephone.

This is not how this works. You either have acute patients who would indeed need handoffs between day and night teams, and you have programmed patients who usually do not need any handoff. Acute patients are usually managed by very specialized teams both in terms of managing physicians, nurses, and usually residents are not alone at night. Those patients are also indirectly managed by other specialists, wether it be the surgical teams, or the imagery team on call. They are also hooked to at least an EKG machines, and depending on the condition will have more invasive surveillance system (arterial catether, articial life support, dialysis etc...). Those patients might die from a medical error, but actually the handoff are not the problem, those are repeated, digitalized, supported and surveillance is maximal. Errors are due to the team's reactivity and knowledge in front of the severity and extremely fast time course of their condition. That is the deciding factor, of course if the doctors are exhausted, they forget things, they miss the jugular vein, they take more time cauterizing the bleed, or closing the wound, they are afraid to wake up their senior because they know he has not slept since yesterday.

On the other hand you have chronic patients, who are under light surveillance by the nurses at night, who sometime call the doctor on call who knows absolutly nothing of the patient, and is sometime not even of the speciality. If things get rough, the patient becomes "acute" and is managed by the acute team. At night for those patients, handoffs WOULD have been nice, be they were none, not because the residents did not do their 12-15 hour day but because those patients are not supposed to have handoffs to the night team, and because no doctors are paid to stay at night in all chronic services.

> Regarding your other point: > In every country except maybe Switzerland, residency programs last 3-5 years, whatever the number of hours worked per week, That makes no sense. If that were the case, residents would choose to work an hour a week and easily graduate in 3-5 years. But that would make for a poor doctor, as they need a certain # hours of training.

So they do finish in 3-5 years, and make for as good doctors as US doctors, although they work more reasonable hours which is around 60 hours a week. Although I believe that is still too much to insure quality of care.

You're correct that handoffs are a common point-of-failure, and also that residency programs are already near the tipping point of financial non-viability for the non-wealthy.

However, to your last point:

> (btw, w/r/t/ shortages, it is probably due to the doctors themselves -- specialty boards such as the Board of Radiology essentially throttle the number of available training positions (residencies) per year creating artificial scarcity and keeping up their own pay.)

The actual bottleneck on residency positions is not artificially imposed by any external group - the bottleneck is the funding for residency positions themselves. Residency programs run at a loss, which is why Medicare has had to step in to fund GME. There are a very small number of non-Medicare funded residency slots, but they're also funded through external means.