[mankyd]

> Not sure if they accounted for delayed surgeries in the study.

That's sort of what I am wondering. Perhaps it just delays the inevitable - the patient is gravely ill is is going to die if they don't perform _potentially_ life-saving surgery. The surgery, is of course risky.

The conference delays the surgery, so the patient's surgery or other high risk procedures are delayed. This gives the patient a few more days of being ill, but doesn't probabilistically change the outcome of actually undergoing the procedure.

[Swizec]

> The conference delays the surgery, so the patient’s surgery or other high risk procedures are delayed.

My understanding from listening to the author’s podcast, is that this is the proposed mechanism. There is a percentage of patients who were going to get better on their own anyway. But if they receive urgent care, it may cause harm.

The conclusion seems to be that there is a measurable percentage of patients who got surgery but didn’t need it and thus suffered greater harm than if they had been left alone. Because heart attacks are so critical, medical staff errs on the side of action instead of waiting. This seems reasonable, but may in fact be bad.

it’s a good podcast: https://freakonomics.com/podcast/what-happens-to-patients-wh...

[MattRix]

They appear to be already accounting for that since they are measuring 30-day mortality for acute conditions. They’re saying it’s possible the reduced mortality is due to the high risk procedures actually being unnecessary.

[oaktrout]

As an argument to illustrate why 30 day mortality isnt a long enough period, imagine this scenario: you have a cancer that has a 90% chance of killing you in one year. You will be cured if you get the surgery tomorrow, but the surgery has a 30% mortality rate. In one month, 30% or those who got the surgery will die, where all those who didn't get the surgery will still be alive. In one year, 70% of those who got the surgery live, whereas only 10% who didn't get the surgery are alive.

[prepend]

But they don’t account for patients going to other hospitals and dying there instead.

[yellowapple]

Now I'm curious about mortality rates in the weeks following a conference. "Lemme try this neat trick I learned..."