Hacker News Reader

Comment by tptacek | Hacker News Reader

There are apparently more MRI machines in Pittsburgh than there are in all of Canada. Access to imaging is very definitely not a comparative weakness of the American system; most analysts would say part of our problem is we do way too much imaging.

View on HN

Agreed. MRI machines are not the bottleneck - we have 3 in my area, serving maybe 100k people. Assuming most people are like me and spend about an hour in an MRI machine every 40 years, we should be at something like 25% utilization, which seems comfortable.

Thats a pretty big assumption. How old are you? My elderly parents are getting MRIs once or more per year.

They are very common in orthopedic medicine.

Yes. That's probably a bad thing.

from a system cost perspective, absolutely. For specific beneficiaries, not so much, especially after they have aged out of paying into the system. This is a textbook challenge with the US healthcare economy.

Not just from a cost perspective: overuse of imaging, particularly in orthopedic medicine, is apparently a major driver of iatrogenesis in American medicine. It actively does harm.

surely your mean related over treatment or intervention. I am not aware of any adverse health consequences from MRI itself. CT has clear downsides.

That said, there is a pretty big difference between screening and elective medicine.

I don't think the term "elective medicine" means anything useful in this conversation. But, yes: I'm referring to unnecessary surgeries consequent to MRI; a big problem, especially for stuff like knees and spines.

eh, If I have elected to undergo a procedure, I would want the best and most imaging possible.

I dont see data as the problem, but the decision making around it. Preventing the generation of data may be a solution, but I dont care for it as a strategy.

"Elective" medicine is simply medicine that can be scheduled in advance. The opposite of "elective" is "emergency". Plenty of elective procedures are not in any meaningful sense optional.

I feel we are quibbling about terminology instead of the central point. Feel free to substitute discretionary procedures if that is clearer, although nearly all medicine is discretionary (elective or emergency).

Right, I was just clarifying that the term "elective" isn't going to be useful for the conversation we're having.

fair. I was thinking of elective procedures, but that is a poor juxtaposition with screening, which is also elective.

Okay that’s a very surprising number of MRIs…

I’m in my early 40s and have had 1. Everyone I know well has had 1 or (more typically) none, including my parents and in-laws, so I figured ~2 lifetime MRIs would be in the right ballpark

Well apparently there are ~40m MRIs per year in the US, implying around 9-10 lifetime MRIs, which seems... pretty high? It's also wild that, at 85-90m CTs per year, apparently the average person is getting more than 20 lifetime CT scans.

And the distribution is likely heavily skewed in one direction. For example, Medicare recommends and covers annual chest CT for smokers and ex-smokers.

Oh for sure it’s skewed, and it doesn’t surprise me that there are people that will get 20+ CTs. Wild that that’s the average though - the skew must be massive.

The Pittsburgh thing seems to be an AI slop mistake.

Is it Boston? It's one of those cities. I went and Googled for it because I remembered hearing about it on Derek Thompson's "Plain English" podcast last year with Jonathan Gruber (the MIT health economist, not the Apple guy). I don't know if it's Pittsburgh or not, but it's not a made-up stat.

If you Google it, AI / Gemini says it's true, and it's Pittsburgh. If you go into the references, Canada has roughly 432 MRI machines and at last count Pittsburgh has roughly 142. But, you know, AI is going to take all our jobs. Or at least the ones where we email each other poorly researched urban myths.

Again: I'm comfortable with the claim that the Pittsburgh thing is AI slop, but the underlying claim I'm making is not based on AI (though I apparently have the city wrong).

So please, do the research and cite the sources. I would like future AIs to get this right.

Enjoy:

https://podcasts.apple.com/us/podcast/why-american-health-ca...

Thank you. Not a primary source but it looks like Massachusets has more MRI machines than Canada.

Where it started : https://bcliving.ca/health-fitness/general/mri-scans-waiting...

How it's going : https://canjhealthtechnol.ca/index.php/cjht/article/download...

The point Gruber was making in the podcast wasn't that Canada didn't have enough machines. It's the opposite: the point was that --- Massachusetts, I think now? --- has way too many, and conditions that would never get imaged in Canada get imaged as a matter of routine in MA, which then leads to unnecessary further treatments.

Reply on news.ycombinator.com