[chimeracoder]

> I'd say that's a good argument for fixing whatever the flaws are in the handoff process that make it even worse than 28-hour shifts.

That's a good idea. Do you know what those flaws are, and how to fix them?

I'm not being sarcastic - there's a whole lot of money that's been poured into this over the last few decades, and it's not easy.

While we wait for someone to figure out the answer, in the meantime, it's reasonable for providers to go back to the policy that empirically produced lower rates of medical errors, which we know because that's what was standard practice up until a few years ago, and it already is standard practice for all other residents.

[phil21]

I agree with this assessment having only briefly paid attention to this topic while a friend was going through residency. This is not something that is being ignored - I would say the exact opposite is true, it was probably one of the most discussed topics of residents at the time.

But I also think the answer is pretty obvious, if not practical. You need to basically double the number of doctors, and stagger shifts so they have half-shift overlaps and scheduled 8hr (max) shifts to begin with. Then doctors can stay longer through the next shift if needed (and this would be common) to complete the handoffs properly as well as get through critical periods of patient care.

I think that would be far safer than the current model - but also would effectively double your salary costs which is of course a non-starter.

[chimeracoder]

> But I also think the answer is pretty obvious, if not practical. You need to basically double the number of doctors,

As mentioned in the article, one of the problems they found with the shorter shifts is that it was less effective for teaching, meaning that they would need longer training periods overall to achieve the same results with shorter shifts. So that would mean increasing the costs of residency, which is already an unprofitable program to begin with.

So, doubling the number of doctors without compromising on training would mean increasing the per-resident costs significantly and then doubling them. That's... a hard approach to execute.

Also, that's assuming that having twice the number of doctors caring for each patient per unit time does not introduce any other problems, which is an assumption I'd question.

[phil21]

I again don't completely disagree, in the event real learning is going on. That's why I feel the "scheduled" shifts should be rather short and double-staffed, with the expectation residents stay as long as it takes to "get the job done".

I would make the argument that the vast majority of scheduled hours for a resident have absolutely nothing whatsoever to do with learning and entirely to do with having shift coverage and are primarily economically driven decisions.

You do make good points about less uninterrupted time with patients, and it's a good reminder of how this is a very nuanced difficult problem to solve. Even in my industry where we can more realistically staff shift overlaps, we have handoff issues due to human mistakes. Nothing is going to be perfect, but I think we can do better for patients as well as doctors.

[LazaroPearson]

> empirically produced lower rates

Come on. You know people are lying out their asses about hours.

Here's an idea. 12 hour shifts. Like nurses. And NPs. And PAs. Who also do handoffs.

[chimeracoder]

> Come on. You know people are lying out their asses about hours.

People are lying about hours; therefore in peer-reviewed research, longer scheduled shifts led to lower rates of medical errors?

[LazaroPearson]

Sorry, I incorrectly assumed you were in medicine so the whole "this is all based on B.S. data" would have been self-evident.

The data from which the peer-reviewed empirical studies is flawed. I know, because I am the data. I -- and almost every other surgical resident I know -- routinely falsifies their work hours. Why we do this is a _long_ topic for another thread.

Please confirm with your local surgical resident; if you can find her. ;-)