There have been some articles here in the past on the thoughts of Leland Hood who often discusses the future of medicine, the advances he expects and the changes he sees coming, especially changes in the physician's role. For example he sees the medical office of the future as a conference room with the patient, the physician, a biologist and a statistician.
This is A Guest Blog from Ned:
The income of physicians is going to have to decrease by very, very large amounts in the future. The math just doesn't add up, and that's the lowest hanging fruit to carve out. Attendings make freakish salaries, and, as physicians get pushed towards employee-employer contracts with hospitals, the hospitals will have more negotiating leverage over their new employees. When docs start to complain that their salaries are getting cut by 25%-50%, the general public will crucify them. And physicians will lose the PR battle.
The other problem is more subtle. Currently, a physician's value comes in two main forms (1) diagnostic interpretation and (2) surgical intervention. The value of a diagnosis is decreasing. Information is everywhere, and it will only be a matter of time until a computer outperforms a physician in terms of diagnostic accuracy, volume, and speed (think Watson). If you disagree, then you are thinking like those chess aficionados in the 80s who declared that a computer algorithm could never triumph over a grand master's experience and judgement.
Except it can. In fact, the last time anyone beat a computer was in the 90s. A computer hasn't been defeated by a grand master since I was a child.
And, in truth, arriving at a diagnosis is closer to chess than most people think. And there are multiple examples where computers have outperformed physicians. would think. Both are essentially probability calculations given appropriate inputs. But while chess has only 64 squares to decide on, a diagnosis has gazillions. So computers aren't there yet. But they will be.
The other issue facing physicians is the prevalence of substitutes. Would I rather go and see a physical therapist for 30 mins of 1x1 time to help me understand my hip injury, or would I rather see an orthopedist who will make me wait for 45 minutes, make me see a clueless resident for 25, and then see me for 5 minutes? No brainer. Will my physical therapist be just as good at diagnosing me? Yes. Will it be cheaper? Yes. Will I get more out of the visit? Yes. Then how can a physician command a premium price? I don't think they can.
Another example of substitute services. Could your PA do as good of a job as you do from a diagnostic standpoint? Probably pretty close. So what if the laws were changed to allow a PA to make diagnoses? To see patients? Could you stay in business?
Final point I'll leave you with. In medicine, the coin of the realm is data. And docs don't understand data. Once upon a time, docs were in charge and told people like me what to do. Now, data scientists are the ones charting the course. And this isn't isolated to medicine. Experienced business people who go with their gut are getting outperformed by soft spoken, awkward guys who understand the data. This means that the "cowboys" of the business world who say "You know what! I know what we should do! Everyone! Follow me!" are getting fired.
Everyone cares about the next big thing to save a bunch of money, but the truth is that you could save tons of cash by making physicians wash their hands as often as they should. Infection rates would drop like crazy. But they don't do it. And this alludes to a greater point. Physicians are not really scientists. They're trained to think scientifically, sure. But they don't really understand science as intimately as TV would have you believe. This prevents them designing appropriate studies, collecting data, analyzing it appropriately, and arriving at appropriate conclusions. This is why so many studies are not reproducible. So how can you put them in charge of the next technical revolution? You can't.
Physicians are about to become dinosaurs. They're scouts in the A's clubhouse, pissed off that statisticians are doing the jobs they used to do. The times are about to change, their value is going to drop in 15 years.
This is A Guest Blog from Ned:
The income of physicians is going to have to decrease by very, very large amounts in the future. The math just doesn't add up, and that's the lowest hanging fruit to carve out. Attendings make freakish salaries, and, as physicians get pushed towards employee-employer contracts with hospitals, the hospitals will have more negotiating leverage over their new employees. When docs start to complain that their salaries are getting cut by 25%-50%, the general public will crucify them. And physicians will lose the PR battle.
The other problem is more subtle. Currently, a physician's value comes in two main forms (1) diagnostic interpretation and (2) surgical intervention. The value of a diagnosis is decreasing. Information is everywhere, and it will only be a matter of time until a computer outperforms a physician in terms of diagnostic accuracy, volume, and speed (think Watson). If you disagree, then you are thinking like those chess aficionados in the 80s who declared that a computer algorithm could never triumph over a grand master's experience and judgement.
Except it can. In fact, the last time anyone beat a computer was in the 90s. A computer hasn't been defeated by a grand master since I was a child.
And, in truth, arriving at a diagnosis is closer to chess than most people think. And there are multiple examples where computers have outperformed physicians. would think. Both are essentially probability calculations given appropriate inputs. But while chess has only 64 squares to decide on, a diagnosis has gazillions. So computers aren't there yet. But they will be.
The other issue facing physicians is the prevalence of substitutes. Would I rather go and see a physical therapist for 30 mins of 1x1 time to help me understand my hip injury, or would I rather see an orthopedist who will make me wait for 45 minutes, make me see a clueless resident for 25, and then see me for 5 minutes? No brainer. Will my physical therapist be just as good at diagnosing me? Yes. Will it be cheaper? Yes. Will I get more out of the visit? Yes. Then how can a physician command a premium price? I don't think they can.
Another example of substitute services. Could your PA do as good of a job as you do from a diagnostic standpoint? Probably pretty close. So what if the laws were changed to allow a PA to make diagnoses? To see patients? Could you stay in business?
Final point I'll leave you with. In medicine, the coin of the realm is data. And docs don't understand data. Once upon a time, docs were in charge and told people like me what to do. Now, data scientists are the ones charting the course. And this isn't isolated to medicine. Experienced business people who go with their gut are getting outperformed by soft spoken, awkward guys who understand the data. This means that the "cowboys" of the business world who say "You know what! I know what we should do! Everyone! Follow me!" are getting fired.
Everyone cares about the next big thing to save a bunch of money, but the truth is that you could save tons of cash by making physicians wash their hands as often as they should. Infection rates would drop like crazy. But they don't do it. And this alludes to a greater point. Physicians are not really scientists. They're trained to think scientifically, sure. But they don't really understand science as intimately as TV would have you believe. This prevents them designing appropriate studies, collecting data, analyzing it appropriately, and arriving at appropriate conclusions. This is why so many studies are not reproducible. So how can you put them in charge of the next technical revolution? You can't.
Physicians are about to become dinosaurs. They're scouts in the A's clubhouse, pissed off that statisticians are doing the jobs they used to do. The times are about to change, their value is going to drop in 15 years.
No comments:
Post a Comment