(My musings/thinking-aloud): One partial hypothesis/answer is that there are far fewer pilots/cabin crew than there are doctors/medical staff, and hence it's easier to treat pilots properly. But fewer people makes it
harder to let someone off if they haven't slept well so I don't really buy that.
Another is that the medical profession is much more "flexible"/fluid in its needs, unlike flights scheduled in a regular manner. However, that doesn't mean it's not possible to give doctors an 8 hour workday, and substitute/bring more doctors for other shifts. In fact, this is exactly what happens in many 24-hour manufacturing plants/operations. Foxconn doesn't let a sleep deprived employee make an iPhone, but it's okay to let a sleep deprived doctor perform life-saving/potentially deadly operations?
Another possibility is that "it would cost more" to have 3 docs instead of 1. Which might be supported by anecdotal evidence on how expensive things (including salaries) are in the medical field. However (especially if you're only familiar with the US healthcare system), this doesn't explain how doctors in say India are overworked.
The "real" reason, I think behind all this? The first part is it's because it's "easy" to compress and push an 8 hour shift to a 24 hour shift, fatigue be damned. The costs are "hidden", it's likely even with fatigue a 99% successful operation is still 95 or 90% successful. Everyone can rationalize it and go about with their day. "Oh, unfortunately people don't always make it".
Not, "there was a 20% chance they wouldn't have made it but an 80% chance it was because the doctor was fatigued."
When an aircraft crashes, 300 people may die instantly. This gets front page coverage on a newspaper. However, a few deaths "here" and "there" don't really show up, even if they tally up to thousands. This is the second factor, that "hides" and normalizes occasional slip-ups.
</End of this wall of text>
I unfortunately don't really have a solution. I'm sure there are brilliant UX designers here on HN who could help ease and streamline the admin workload.
I suppose a dedicated politician/presidential candidate somewhere could take this upon themselves to campaign for. (Seriously, the NHS was part of the discussion behind Brexit. Supporting doctors is a brilliant political strategy, the NWA didn't have a song called "Fuck the paramedics".)
There have been many instances of things catching on once found successful somewhere, be it procedures like the Heimlich or concepts like Lean Manufacturing. Therefore I don't doubt that one good Harvard Review paper showing a 20% decrease in all-cause patient mortality, with trials underway in a few hospitals, is all it will take. Question is, who will bell the cat‽