81–90 of 274 posts
On a related vein, do surgical procedures involve any type of quality control, where the work is independently audited? If so I wasn't aware. Quality control is such an important cogwheel of processes in many other industries, yet I don't hear much talk of it in relation to medical profession.
You can see one example here
The advent of public reporting of outcomes after intervention for myocardial infarction (heart attack) led to an increase in mortality in Massachusetts and New York.[1] 1 = https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4368858/
A better way to make this claim would be to plot the mortality rates for the states over time and show that the rates in those places requiring reporting begin to diverge as the new policy is implemented.
It's Goodhart's Law in action, this time with deadly consequences: "When a measure becomes a target, it ceases to be a good measure."[a] [a] https://en.wikipedia.org/wiki/Goodhart%27s_law
https://www.lesserwrong.com/posts/EbFABnst8LsidYs5Y/goodhart...
Sports athletes do this. For example players won't attempt end of the quarter half court shots since that will lower their percentages, or QBs taking sacks to not lower their QB ratings. What they do in sports is keep advanced stats. Instead of just giving a pure mortality rating have mortality rating, by surgery type.
It's Goodhart's Law in action, this time with deadly consequences: "When a measure becomes a target, it ceases to be a good measure."[a] [a] https://en.wikipedia.org/wiki/Goodhart%27s_law
Low mortality is a good damned measure. I don't want to be operated on some yahoo surgeon who unconditionally operates regardless of risk and has a high mortality track record as a result.
Earlier quoted context omitted.
Then they just dodge the hard cases of each specific type of surgery. It's probably still an improvement though on just avoiding a surgery type altogether.
Then you track the deferral rates.
It's Goodhart's Law in action, this time with deadly consequences: "When a measure becomes a target, it ceases to be a good measure."[a] [a] https://en.wikipedia.org/wiki/Goodhart%27s_law
Low mortality is a good damned measure. I don't want to be operated on some yahoo surgeon who unconditionally operates regardless of risk and has a high mortality track record as a result.
No surgeon wants to do it, to not get his metric lower, so instead of 30% survival chance you get effective death sentence ...
Earlier quoted context omitted.
Then they just dodge the hard cases of each specific type of surgery. It's probably still an improvement though on just avoiding a surgery type altogether.
Then you track the deferral rates.
They need to add a difficulty correlated metric to the statistics. Maybe BMI and age? I am sure IVF clinics wouldn't offer the procedure to anyone over 35 if the rather detailed statistics they are required to report didn't also include the patient's age.