How many of these are a result of doctors malpractice?
Approximately 10-20%, which probably represents their percentage of healthcare staff. So, they are likely over-represented in the error space.
Having seen situations where patients don't have good results, a lot of the time it's something like a patient not making an appointment to go see an onc when needed or a PA not interpreting results correctly. I doubt a post mortem forensic chain analysis or whatever u can call it was taken on each of the 250 - 500k scenarios.
The studies are out there -- Johns Hopkins did one of the major ones.
What you're experiencing is exactly what I said-- most people (and you in this case) would be astonished if they knew the level of error due to incompetence displayed every day in modern healthcare.
From what I've seen, a lot of patients are just plain moronic and do themselves in. They smoke, drink, eat terribly, and when the doctor can't raise their hands over the patient, shooting beams of light out over their body, and cure them like Jesus, then it's the doctors fault.
Yes, it's not just doctors. As I said earlier, it's also education (my field) and many others.
Look at all the oil filter companies that can't make simple louvers or get the specs right on their own products. Car companies that can't make a cam, oil pump, transmission, etc. that will reliably go to 100k mi.
For many reasons, incompetence and error are the human condition. I'm no different. I make a ton of mistakes all the time. Probably more than the average guy in the skilled trades. So, I'm not pumping myself up, here-- I'm just as bad as anyone (and getting worse as I age!).
The solution is to remove the human element by replacing humans with AI.
Being familiar with a spectrum of medicaid patients up to better off ones, most people are just killing themselves 5x different ways and hoping some pill fixes them when it's too late. They then say "oh man the medical system sucks!" as they drive off into the sunset in their overloaded electric shopping cart to pick out a piano box sized coffin.
Yep, patients can be equally breathtaking in their incompetence as well. They are human, as are we all.
I bet these "500k deaths a year" stat won't change much even after AI because you have to advocate for yourself and follow instructions. Yes, a surgeon will occasionally cut off the wrong leg or leave scissors in a patient, but I'm sure AIs will make mistakes too! AI won't fix stupid patients though, that's the biggest problem.
I agree that's an open question on which honest and intelligent people can disagree. I make my point for AI, we'll see someday if I'm right. Of course, I could be wrong (and will admit it if I am).
My AI can argue with your AI all day
:
"The studies that estimate 250,000 to 500,000 preventable deaths per year highlight errors that are inherently
systemic and involve non-physician staff:"
Yep, as I said above-- 10-20% is doctor error, and I don't think they make up more than that percentage of the total staff. So, likely over-represented.
What educational criteria would you feel comfortable with someone running the "AI robot" or treating you?
This is a common error in understanding. AI does not need to be "run" in the background. It is intelligent, so it can figure things out on its own. Better than a human can.
This is a whole other topic on which I could expand if you want. I used neural network theory in my research and taught very basic neural network theory in one of my classes. I'm not a huge expert, but I do know something about it.
That stat is a meme and potentially flawed.
"
Origin: The 251,000 figure comes primarily from a highly publicized 2016 analysis published in the
British Medical Journal (BMJ) by researchers from Johns Hopkins (Makary and Daniel). They calculated this number by
extrapolating rates of medical error identified in a few large-sample studies (like the earlier Institute of Medicine's report,
To Err Is Human) to the total number of US hospital admissions.
- Small Sample Size: The extrapolation was based on studies that reported a very small number of deaths, which is unreliable when applied to the entire U.S. population.
- Inter-Rater Disagreement: Determining if a death was "preventable" is highly subjective. Studies show that when multiple expert reviewers examine the same medical files, they often disagree significantly on whether a death was definitely preventable"
That's not the only study, and all research can be criticized. Having worked in clinical settings, and my wife's career-long experience in clinical medicine have shown me first-hand that there are plenty of problems.
Say you, what do you think a future AI hospital will look like?
Mostly AI robots, few humans. Those that remain will likely be parasitic management types, not needed but able to cling to the system based on spurious arguments for their "need" in oversight.
Will we also have AI plumbers and mechanics?
Absolutely, yes. Read the current thread here on flat-rate mechanic pay. There is a crisis in the trades as there is huge need and little incentive for humans to do that hard work for the low pay offered. AI robots are definitely the answer.
What will humans do, chill out all day watching Tik Tok?
You say that like it's a bad thing!
Yes, that's exactly what I think.
As a (nominally) political conservative guy, I've always been skeptical of the concept of universal basic income. Everything I've ever known about it sounded like the hair-brained "Free Stores" from the 1960s.
However, with AI, for the first time I see a legitimate path to it. Wouldn't you prefer to not have to work at all and get ~$100K check every year?
Basically, people in AK get free money from the oil reserves-- AI will be an order of magnitude more valuable and productive. Just substitute "U.S. population" for "Alaska" and "AI" for "oil," and you'll get the picture.
What's the alternative to current medicine and are you using it?
AI robots of all types. Not using now as they don't exist yet. But will as soon as feasible.
BTW, I believe your opinions on this do have some validity and a fully agree that it's not impossible that you will be right and I will be wrong on the AI future.
Maximum respect to you for taking the time to make your case to me. You did a nice job, indicating you've put some thought into this.
I appreciate that.
