Once you learn about AI in medicine and actually use it, you realize how horribly useless and bad it is.

Then you see how much the economy is betting on it not being so.

Then the horror of it all hits you.

Do you anticipate lowering reward processing like it does in substance use+increased satiety? Or some other mechanism? It was incidentally tried in one of my diabetic HD patients and actually worsened things. IDWG went down a little but weight dropped more. So IDWG/ weight % looked horrible

And why did finerenone get away with it? Kerendia is being prescribed right and left now because they had the RCT data and FDA approval. But it’s very clear to me that old MRAs r just as protective, just lacked the RCT data for non medical circumstances. An unpopular opinion, but I own it.

I agree 100%. May not be a very unpopular opinion.

IgA is pronounced eye-gee-ay

The reason the disease should be called "eye-gee-ay nephropathy" is because IgA is central to the pathogenic mechanism.

Moving to eye-gan is moving away from describing the disease by its pathogenic mechanism.

What puzzles me is how people make the point to fund more vitamin D studies and get through review. They can’t possibly make Cartman’s point

Nicely analyzed!

Why not? Looks like Apple made a monitoring device, not a screening device. Gets away from the entire debate about false positives. Once you have hypertension, use this highly specific device to detect uncontrolled bp averaged over time. Should be good for masked hypertension and give decent abpm

The drug costs about $450,000- 550,000 per year. That proteinuria reduction better translate into dialysis avoidance forever or immense improvement in QALYs.

ABIM response: /s

Cayenne/ jalepeno pepper juice + water + a spoon of dish wash detergent to increase spread on surfaces. Spray liberally and don’t rub your eyes after touching your furniture

Never underestimate the underdog.

CNBC bro says this company and their cloud subsidiary: Fully Automated Response Technologies are going to engulf the AI market… like a supernova… or a black hole.

Then the AI quality measuring tool will give it a perfect score in access to care. The prescriber and insurance company will be put in the preferred tier for appropriate and expedited steps. Do it enough times they get a. AI award for care. Snowball effect, never say “oops I made a mistake”.

Just the other day, my DAX Copilot hallucinated and put in my note that my suspected IgA patient going for biopsy received a kidney transplant already.
With the advent of AI prescribers, that patient will get a nice prescription of tacrolimus as well. That will be another acceptable error.

Makers of GenAI LLMs have disclaimers saying “expect 2-3 serious errors per document generated”. This person was the “acceptable” serious error.
Stop trying to question progress, bro!!

I had one patient who was resistant to Cyclophosphamide but responded to the 1 gram rituximab 2 weeks apart regimen. Went into complete remission. However, he was caught early and had lots of kidney to save. Elderly anuric patients without pulm symptoms may have nothing to gain and lots of risk.

But the high creatinine <check engine> light was in red with not one, but TWO upwards arrows!!