Doctors cannot compete with machines

Burrell College of Osteopathic Medicine
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
DermViser

DermViser

Full Member
10+ Year Member
Joined
Apr 4, 2009
Messages
17,237
Reaction score
13,034
Points
7,076
  1. Resident [Any Field]
of course doctors can compete with machines

we have the ability to tolerate BS from our patients and take the blame when things go wrong

data science will revolutionize our field though

i hope one day we can pull up the medical charts of every patient who has the same history, P.E., complications, and labs as our patient and get some statistical analysis of the diagnosis and efficacy of treatment.

for now i will settle for checklists and electronic alerts for any orders written by the doctor as well as automated presentations that will inform the patient.

imagine how much more efficient it would be if you can just use video presentations for diet and exercise, informing the patient how/when to take their medications, signs and symptoms that require immediate medical attention.
 
diseases.jpg

IdiocracyHospital.jpg

It wasn't a movie. It was a prophecy.
 
This guy has no idea what he's talking about

“Sufficient data used properly and reduced to the right insights does in fact make up for errors,” Khosla said. “I would rather have 1,500 EKGs (electrocardiograms, a test that checks for problems with the electrical activity of the heart) done much more poorly than two EKGs done a year very well, because the sources of errors in the current system are just too large. When I have two EKGs a year, I may not be symptomatic. I’m not arguing that these systems don’t have errors. I’m saying the volume of the data, properly applied, makes up for it.”

Who is going to pay for these 1,500 EKGs? Who is going to set them up and interpret them? Who has the time to sit there while they get 1,500 EKGs lol. Overtesting leads to incidentalomas, leads to overtreatment, leads to harm.


"In particular, he said, wearable medical sensors, like Fitbit, will give patients power to make informed health and health-related decisions on their own."

Patients can't even do simple things like taking a medication every day, quit smoking or start eating healthy foods. Some of them can barely spell their own names. How are they equipped to making complex decisions about their health without the slightest bit of medical knowledge?


“Humans are not good when 500 variables affect a disease. We can handle three to five to seven, maybe,” he said. “We are guided too much by opinions, not by statistical science.”

Guided by experience, not opinions. Statistics are irrelevant when your patient is sitting right there in front of you. Who cares if a drug works for 95% of people in a trial if it doesn't work for your patient? Ridiculous.
 
Psai said:
This guy has no idea what he's talking about

“Sufficient data used properly and reduced to the right insights does in fact make up for errors,” Khosla said. “I would rather have 1,500 EKGs (electrocardiograms, a test that checks for problems with the electrical activity of the heart) done much more poorly than two EKGs done a year very well, because the sources of errors in the current system are just too large. When I have two EKGs a year, I may not be symptomatic. I’m not arguing that these systems don’t have errors. I’m saying the volume of the data, properly applied, makes up for it.”

Who is going to pay for these 1,500 EKGs? Who is going to set them up and interpret them? Who has the time to sit there while they get 1,500 EKGs lol. Overtesting leads to incidentalomas, leads to overtreatment, leads to harm.


"In particular, he said, wearable medical sensors, like Fitbit, will give patients power to make informed health and health-related decisions on their own."

Patients can't even do simple things like taking a medication every day, quit smoking or start eating healthy foods. Some of them can barely spell their own names. How are they equipped to making complex decisions about their health without the slightest bit of medical knowledge?


“Humans are not good when 500 variables affect a disease. We can handle three to five to seven, maybe,” he said. “We are guided too much by opinions, not by statistical science.”

Guided by experience, not opinions. Statistics are irrelevant when your patient is sitting right there in front of you. Who cares if a drug works for 95% of people in a trial if it doesn't work for your patient? Ridiculous.
Click to expand...
Personally I'd rather have two EKGs that are right in a year than 1500 that have a 25% chance of telling me I'm in vtach or vfib or whatever. Will they just set the software up to ignore readings that are likely erroneous? In this case, what happens when the device ignores something serious and passes it off as noise?

Machines can aid physicians, certainly, but it will be a great deal of time before they can replace them, surely longer than any of us will be alive.
 
Psai said:
This guy has no idea what he's talking about

“Sufficient data used properly and reduced to the right insights does in fact make up for errors,” Khosla said. “I would rather have 1,500 EKGs (electrocardiograms, a test that checks for problems with the electrical activity of the heart) done much more poorly than two EKGs done a year very well, because the sources of errors in the current system are just too large. When I have two EKGs a year, I may not be symptomatic. I’m not arguing that these systems don’t have errors. I’m saying the volume of the data, properly applied, makes up for it.”

Who is going to pay for these 1,500 EKGs? Who is going to set them up and interpret them? Who has the time to sit there while they get 1,500 EKGs lol. Overtesting leads to incidentalomas, leads to overtreatment, leads to harm.


"In particular, he said, wearable medical sensors, like Fitbit, will give patients power to make informed health and health-related decisions on their own."

Patients can't even do simple things like taking a medication every day, quit smoking or start eating healthy foods. Some of them can barely spell their own names. How are they equipped to making complex decisions about their health without the slightest bit of medical knowledge?


“Humans are not good when 500 variables affect a disease. We can handle three to five to seven, maybe,” he said. “We are guided too much by opinions, not by statistical science.”

Guided by experience, not opinions. Statistics are irrelevant when your patient is sitting right there in front of you. Who cares if a drug works for 95% of people in a trial if it doesn't work for your patient? Ridiculous.
Click to expand...

I actually calculated this. There are 75 million ECGs performed per year in the U.S. (as of 2001; probably more now). An ECG costs 10-15 dollars without interpretation.

Let's assume that those 75 million include some repeat ECGs, unnecessary ECGs, etc and halve it (I'm being generous) and say that only 37.5 million unique ECGs are performed each year. 37.5 million x 1500 (Vinod Khosla's magic number) x $10/ECG (medicair reimbursement rate w/o interpretation)= $562,500,000,000. Like you said... WHO IS GOING TO PAY FOR THIS???!?!?
 
How long before patients are replaced by machines? I like going to Vegas and play'n the slot machines.
 
Advertisement - Members don't see this ad
Delayed2017 said:
I actually calculated this. There are 75 million ECGs performed per year in the U.S. (as of 2001; probably more now). An ECG costs 10-15 dollars without interpretation.

Let's assume that those 75 million include some repeat ECGs, unnecessary ECGs, etc and halve it (I'm being generous) and say that only 37.5 million unique ECGs are performed each year. 37.5 million x 1500 (Vinod Khosla's magic number) x $10/ECG (medicair reimbursement rate w/o interpretation)= $562,500,000,000. Like you said... WHO IS GOING TO PAY FOR THIS???!?!?
Click to expand...

pretty sure it won't be an EKG, but some monitor you would wear which would do real-time EKG monitoring

someone should explain to him sensitivity and specificity
 
Oh look, another non-MD saying how MDs are going to be replaced by machines. A venture capitalist focused partially on information technology advances discussing how information technology in the future is going to replace doctors? Sorry if I don't take him supersrs 🙄
 
Winged Scapula said:
LOL -- my pet peeve: Pre Op testing nurses who call me about "abnormal EKGs".
Click to expand...

As someone who used to make some of those calls, part of it is policy. If we didn't document that the physician was notified and then someone does a chart audit and catches it, well...

Most of the time where I worked there were anesthesia residents or an anesthesia attending nearby, so it was easier to let them know. It feels dumb to call when it's an issue that's been stable for years, but if all the boxes aren't off, badness is sure to follow.

I miss the days when common sense and clinical judgement prevailed.
 
fab4fan said:
As someone who used to make some of those calls, part of it is policy. If we didn't document that the physician was notified and then someone does a chart audit and catches it, well...

Most of the time where I worked there were anesthesia residents or an anesthesia attending nearby, so it was easier to let them know. It feels dumb to call when it's an issue that's been stable for years, but if all the boxes aren't off, badness is sure to follow.

I miss the days when common sense and clinical judgement prevailed.
Click to expand...
But don't you just feel better and a great sense of accomplishment by checking off boxes, documenting, meaningful use, etc.? (NOT)
 
Advertisement - Members don't see this ad
fab4fan said:
While I'm at it, allow me to wax rhapsodic over that marvelous invention, the EHR. (So not.)
Click to expand...
Yup, an invention that apparently saves the system money, which we later find out, surprise, surprise it actually doesn't, it increases spending. Pair that with spending decreased time with the patient. It's a complete lose-lose situation for patients. Great way to drive out private practice, though, which I suppose is the point. I'm sure the owner of EPIC is loving it, being an Obama donor and all.
 
Last edited:
DermViser said:
Yup, an invention that apparently saves the system money, which we later find out, surprise, surprise it actually doesn't, it increases spending.
Pair that with spending decreased time with the patient. It's a complete lose-lose situation for patients. Great way to drive out private practice, though, which I suppose is the point. I'm sure the owner of EPIC is loving it though, being an Obama donor and all.
Click to expand...

I see a specialist who has essentially dug in her heels and refused to go with EHR. It works for her since she's in a niche practice and doesn't have partners. It's great to have a conversation with a doctor who isn't transfixed on a computer screen, because the prompts from the EHR make him/her have to go to page after page to document.
 
fab4fan said:
As someone who used to make some of those calls, part of it is policy. If we didn't document that the physician was notified and then someone does a chart audit and catches it, well...

Most of the time where I worked there were anesthesia residents or an anesthesia attending nearby, so it was easier to let them know. It feels dumb to call when it's an issue that's been stable for years, but if all the boxes aren't off, badness is sure to follow.

I miss the days when common sense and clinical judgement prevailed.
Click to expand...

cant NP or PA do it?
 
I don't understand, why can't machines be programmed to read EKGs? Given enough data, they should be superior to us at pattern recognition.
 
Brain Bucket said:
I don't understand, why can't machines be programmed to read EKGs? Given enough data, they should be superior to us at pattern recognition.
Click to expand...
If the leads aren't placed right or the patient is moving, you end up with artifact. Sometimes this artifact even looks pathological in nature. Machines are completely terrible at determining what is a BS reading and what is legit in this regard. EKGs aren't something you want wrong 10% of the time.
 
Mad Jack said:
If the leads aren't placed right or the patient is moving, you end up with artifact. Sometimes this artifact even looks pathological in nature. Machines are completely terrible at determining what is a BS reading and what is legit in this regard. EKGs aren't something you want wrong 10% of the time.
Click to expand...

Ah. Thanks.
 
Advertisement - Members don't see this ad
Psai said:
This guy has no idea what he's talking about

“Sufficient data used properly and reduced to the right insights does in fact make up for errors,” Khosla said. “I would rather have 1,500 EKGs (electrocardiograms, a test that checks for problems with the electrical activity of the heart) done much more poorly than two EKGs done a year very well, because the sources of errors in the current system are just too large. When I have two EKGs a year, I may not be symptomatic. I’m not arguing that these systems don’t have errors. I’m saying the volume of the data, properly applied, makes up for it.”
Click to expand...

Me thinks someone needs a remedial lesson in overscreening/overtreatment. The mammography debate would be an excellent starting point, if this person had the chops to understand any of it.
 
Same goes for EEG reading.

Let the entrepreneurs have their say. We all know it's bogus. What's next, independent robotic brain surgery? Yeah.
 
Doctors have nothing to fear from machines.

The only thing we have to fear is patients eating healthy, stopping smoking, and exercising. And I will bang Natalie Dormer before that happens.
 
fab4fan said:
As someone who used to make some of those calls, part of it is policy. If we didn't document that the physician was notified and then someone does a chart audit and catches it, well...

Most of the time where I worked there were anesthesia residents or an anesthesia attending nearby, so it was easier to let them know. It feels dumb to call when it's an issue that's been stable for years, but if all the boxes aren't off, badness is sure to follow.

I miss the days when common sense and clinical judgement prevailed.
Click to expand...
OH yes, I understand the need for documentation and why its being done. Its just that I would like a little input and "common sense", as in: "sinus bradycardia" - not abnormal in a 29 year marathon runner (not "do you want her to see a cardiologist and cancel surgery for tomorrow?") or "evidence of old infarct": unchanged from prior EKGs dating back 3 years and we have clearance from cardiologist/ECHO in file. I KNOW they are just reading the report from the machine to me rather than reading the report and then providing more information/discussing it with me. Its just a reaction without thinking it through that bothers me
 
fab4fan said:
I see a specialist who has essentially dug in her heels and refused to go with EHR. It works for her since she's in a niche practice and doesn't have partners. It's great to have a conversation with a doctor who isn't transfixed on a computer screen, because the prompts from the EHR make him/her have to go to page after page to document.
Click to expand...
I take it she doesn't take Medicare or is cash only? Otherwise the use of EHR is mandated by 2015.

<sigh> What a dream that would be. Good for her!
 
Advertisement - Members don't see this ad
fab4fan said:
As someone who used to make some of those calls, part of it is policy. If we didn't document that the physician was notified and then someone does a chart audit and catches it, well...

I miss the days when common sense and clinical judgement prevailed.
Click to expand...
While your common sense and clinical judgment may be up to par, if everyone had it...we'd never bother with the boxes. Protocol is there to help, even if it isn't always efficient. 🙂
 
Brain Bucket said:
Ah. Thanks.
Click to expand...

You will likely see this for yourself when you get on the wards. Here, at least, an automated interpretation is provided on the EKG print-out. I don't even bother looking at it anymore. The number of times I've seen **VTACH 250>100 or **ABNORMAL ST CHANGES on a completely asymptomatic patient is magnitudes more than the number of times an automated report has actually provided useful information.


Sent from my iPhone using Tapatalk
 
NickNaylor said:
You will likely see this for yourself when you get on the wards. Here, at least, an automated interpretation is provided on the EKG print-out. I don't even bother looking at it anymore. The number of times I've seen **VTACH 250>100 or **ABNORMAL ST CHANGES on a completely asymptomatic patient is magnitudes more than the number of times an automated report has actually provided useful information.


Sent from my iPhone using Tapatalk
Click to expand...
One reason any type of tech advancement won't take over anything docs do (at least not the **** of significance): malpractice. No hospital or doctor will put their malpractice on the line and rely on a machine. Heck, you can't even quote vitals in an EMR to attendings without getting flack for not measuring it yourself, sometimes.
 
Winged Scapula said:
OH yes, I understand the need for documentation and why its being done. Its just that I would like a little input and "common sense", as in: "sinus bradycardia" - not abnormal in a 29 year marathon runner (not "do you want her to see a cardiologist and cancel surgery for tomorrow?") or "evidence of old infarct": unchanged from prior EKGs dating back 3 years and we have clearance from cardiologist/ECHO in file. I KNOW they are just reading the report from the machine to me rather than reading the report and then providing more information/discussing it with me. Its just a reaction without thinking it through that bothers me
Click to expand...


I don't disagree with you. Unfortunately, people writing policies hamstring those of us at the bedside. If we didn't report it, we heard about it from management.
Winged Scapula said:
I take it she doesn't take Medicare or is cash only? Otherwise the use of EHR is mandated by 2015.

<sigh> What a dream that would be. Good for her!
Click to expand...


Actually, she does take Medicare. I'm not sure how she's avoided EHR so far. It's a small practice, so IDK if that makes any difference, e.g. if there is an exception for practices below a certain number of patients. I have to admit I'm not up on all the details of that issue. Other stuff demanding my time and attention.
 
Advertisement - Members don't see this ad
You must log in or register to reply here.

Similar threads

SurfingDoctor
Young Doctors Want Work-Life Balance. Older Doctors Say That’s Not the Job.
Replies
14
Views
5K
deleted1192955
D
D
Need to remediate -- cannot participate in extracurriculars
Replies
7
Views
2K
deleted1089232
D
Goro
MD & DO Doctors behaving badly
Replies
1
Views
1K
DrStephenStrange
DrStephenStrange
magician7772222
M1 feeling overwhelmed with anki and lost with study strategies
Replies
9
Views
3K
doc_J61
doc_J61
D
Clerkship Order with Class Overlap
Replies
3
Views
700
differentiating
differentiating
Top Bottom