What does physician job security look like in the future?

[LoveBeingHuman:)]

With nurses and PA's taking some physician responsibilities, computers being built to taking some physician responsibilities, doctors being accused of giving tests and treatments for money, and the rise in number of medical school graduates, will job security in the field be compromised in the long run? I know that there is a physician shortage but, even after hours of research, I'm having a hard time connecting the dots.

---

Members don't see this ad.

 

[To be MD]

In the long, long run (supposedly within the next 10-40 years), every job will be compromised by significant advances in artificial intelligence and diagnostic software (e.g., Watson), spurred on by the current, sky-high prices of healthcare and the consequential market that exists to use machines to cut costs and provide better care.

That being said, we're not there yet, and we're nowhere close. Even if or when we do get there, physicians will find something else to do with their time or, maybe, the world will turn into a utopia where everyone is guaranteed basic income and robots do all the work-- the only thing you'll ever have to worry about is what color swimsuit to wear at the beach (yea, right).

There are way too many unknowns, but my advice is: if you're passionate about medicine, pursue it.

 

[numbersloth]

In the long, long run (supposedly within the next 10-40 years), every job will be compromised by significant advances in artificial intelligence and diagnostic software (e.g., Watson), spurred on by the current, sky-high prices of healthcare and the consequential market that exists to use machines to cut costs and provide better care.

I recently read an article in the NYTimes that automation is already replacing jobs, but that jobs that require actual human interaction (such as medicine and clinical psychology) are relatively "safe".

 

[Theseus]

There are no guarantees. If you want guarantee buy a toaster.

 

[Pagan FutureDoc]

I recently read an article in the NYTimes that automation is already replacing jobs, but that jobs that require actual human interaction (such as medicine and clinical psychology) are relatively "safe".

It's not so much that it's replacing jobs, it's changing the type of employees needed.
And that's going to hit medicine too, doctors will need to be increasingly tech savvy (just like everyone else) to keep up. Those that can't adapt might get left behind.

 

[jqueb29]

Nurses, PAs, and robots all are unable to come anywhere near accomplishing what a doctor can. Diagnosis and treatment is truly an art, not an algorithm.

 

[Mad Jack]

In the long, long run (supposedly within the next 10-40 years), every job will be compromised by significant advances in artificial intelligence and diagnostic software (e.g., Watson), spurred on by the current, sky-high prices of healthcare and the consequential market that exists to use machines to cut costs and provide better care.

That being said, we're not there yet, and we're nowhere close. Even if or when we do get there, physicians will find something else to do with their time or, maybe, the world will turn into a utopia where everyone is guaranteed basic income and robots do all the work-- the only thing you'll ever have to worry about is what color swimsuit to wear at the beach (yea, right).

There are way too many unknowns, but my advice is: if you're passionate about medicine, pursue it.

People forget that diagnostic software is only as good as its inputs- given the right data, Watson can make diagnoses more accurately than many physicians. But the thing is, separating signal from noise when acquiring that data requires a physician to do in the first place. It's not like Watson does a whole H&P. Couple that with the liability concerns of hospitals and insurers, and you won't see computers replacing physicians in our lifetime.

 

[wubtella]

Even if Watson or some other computer could start doing basic diagnosis, I guarantee that most people would rather go to a human doctor rather than a robot. I'd say we're pretty safe.

 

[yaka]

I've heard rumblings that radiologists might be replaced by computers in the future (mostly from non-radiologists), but while I was doing some research in a radiology department at a major hospital they were focusing a lot on machine learning projects so it might not be totally farfetched. But, realistically, I could never see the other specialities turning over radiology responsibilities completely over to a computer.

Though, interesting thought experiment, if computers reach a point where they are statistically shown to make the correct diagnosis at least at much as human docs do for even the most difficult cases, would that alone be enough to justify phasing out specialities such as radiology or others with minimal human interaction? I could definitely see it lowering costs significantly, but what would we lose in the process?

Not trying to offend any radiologists (or aspiring radiologists), just want to see what others think about this.

 

[Mansamusa]

I've heard rumblings that radiologists might be replaced by computers in the future (mostly from non-radiologists), but while I was doing some research in a radiology department at a major hospital they were focusing a lot on machine learning projects so it might not be totally farfetched. But, realistically, I could never see the other specialities turning over radiology responsibilities completely over to a computer.

Though, interesting thought experiment, if computers reach a point where they are statistically shown to make the correct diagnosis at least at much as human docs do for even the most difficult cases, would that alone be enough to justify phasing out specialities such as radiology or others with minimal human interaction? I could definitely see it lowering costs significantly, but what would we lose in the process?

Not trying to offend any radiologists (or aspiring radiologists), just want to see what others think about this.

That's a pretty big point to reach and radiologists do more than just look at film, so I don't think they would be replaced

 

[Mansamusa]

OP, these things have been discussed in more details in other threads.

I think APRNs and the large increase in med students has a potential of making things suckier for us. But all fields require a leap of faith and no guarantees.

So if job security is your motivation for getting a MD, you could potentially be disappointed because we have no idea how things will change. If job security is just an added perk to a career path you already want to pursue, then just take a leap

 

[Lucca]

Let's put it this way: the concept of job security will change in the future, as it already has in the past two decades, but medicine will most likely still be better off than many other jobs.

 

[deleted407021]

There was a Star Trek Voyager episode where Medicine of one society was practiced by physicians but everything they did, from diagnosis to treatments to who could be treated, was determined by a massively powerful computer.

So a compromise, I guess.

 

[bananafish94]

There was a Star Trek Voyager episode where Medicine of one society was practiced by physicians but everything they did, from diagnosis to treatments to who could be treated, was determined by a massively powerful computer.

So a compromise, I guess.

Dammit Jim, I'm a doctor not a supercomputer!

 

[numbersloth]

I've heard rumblings that radiologists might be replaced by computers in the future (mostly from non-radiologists), but while I was doing some research in a radiology department at a major hospital they were focusing a lot on machine learning projects so it might not be totally farfetched. But, realistically, I could never see the other specialities turning over radiology responsibilities completely over to a computer.

Though, interesting thought experiment, if computers reach a point where they are statistically shown to make the correct diagnosis at least at much as human docs do for even the most difficult cases, would that alone be enough to justify phasing out specialities such as radiology or others with minimal human interaction? I could definitely see it lowering costs significantly, but what would we lose in the process?

Not trying to offend any radiologists (or aspiring radiologists), just want to see what others think about this.

The thing is that radiology may seem like the easiest specialty to automatize, but it turns out computers are exceptionally bad when it comes to vision and interpreting images. When it comes to numbers and data, computers far surpass humans. When it comes to images and video, computers are just really really bad. Computer vision and imaging researchers are still just trying to teach computers to see boundaries correctly, for example.

 

[Lawpy]

Nurses, PAs, and robots all are unable to come anywhere near accomplishing what a doctor can. Diagnosis and treatment is truly an art, not an algorithm.

Yeah exactly. Doctors make decisions from experience and comprehensive educational backgrounds and have humanistic attributes that are absolutely essential for successful diagnosis and treatment plans. This is the basis for good doctoring, so physicians cannot and will not be replaced by supercomputers.

Supercomputers may and probably will wipe out the midlevels. This benefits physicians (especially PCPs) immensely since they don't have to deal with the annoying problem of midlevel encroachment. So i could see supercomputers strengthening physician job security and outcomes over the longterm, with the easier, less experienced healthcare alternatives being eliminated.

 

[Lawpy]

I've heard rumblings that radiologists might be replaced by computers in the future (mostly from non-radiologists), but while I was doing some research in a radiology department at a major hospital they were focusing a lot on machine learning projects so it might not be totally farfetched. But, realistically, I could never see the other specialities turning over radiology responsibilities completely over to a computer.

Though, interesting thought experiment, if computers reach a point where they are statistically shown to make the correct diagnosis at least at much as human docs do for even the most difficult cases, would that alone be enough to justify phasing out specialities such as radiology or others with minimal human interaction? I could definitely see it lowering costs significantly, but what would we lose in the process?

Not trying to offend any radiologists (or aspiring radiologists), just want to see what others think about this.

Supercomputers will wipe out medical physicists and technicians needed to process and analyze medical images. Radiologists will likely be unaffected, or perhaps will even benefit from this new technology. This is because radiologists are doctors and are still directly involved in patient care even when they don't interact with patients. Their expertise is critical for medicine to be sustained and function normally. The knowledge of medical physicists and technologists is readily replaced by machine learning algorithms.

 

[Doctor-S]

With nurses and PA's taking some physician responsibilities, computers being built to taking some physician responsibilities, doctors being accused of giving tests and treatments for money, and the rise in number of medical school graduates, will job security in the field be compromised in the long run?

Supply and demand ... patients' demand for physicians exceeds the supply of licensed and available physicians. [corrected post]

Although modern technology is extremely useful, efficient and fast, it is not a perfect substitute for certain things, including intuitive skills (based on our life experiences and professional training), ingenuity and compassionate interpersonal skills (you know ... "talking" with patients).

In many cities, there are lengthy appointment wait times (30+ days) to meet with a physician because many humans value (and want) a personalized approach to care, as well as a face-to-face meeting with a sentient human being - that is, with their own physician.

At this time, and in the foreseeable future, the global population is continuing to increase, and not decrease.

So, a physician's job security appears positive for a long time - unless you plan to open up a practice on a deserted island in the South Pacific.

 

[Mansamusa]

Yeah exactly. Doctors make decisions from experience and comprehensive educational backgrounds and have humanistic attributes that are absolutely essential for successful diagnosis and treatment plans. This is the basis for good doctoring, so physicians cannot and will not be replaced by supercomputers.

Supercomputers may and probably will wipe out the midlevels. This benefits physicians (especially PCPs) immensely since they don't have to deal with the annoying problem of midlevel encroachment. So i could see supercomputers strengthening physician job security and outcomes over the longterm, with the easier, less experienced healthcare alternatives being eliminated.

Supercomputers won't eliminate mid-level providers, imo. Maybe reduce the number used if anything. The way medicine is structured, nurses, PAs, etc tend to be the ones communicating most with patients and they do offer a lot of value in terms of patient care. Their knowledge is more replaceable by a computer, but I don't think the value they bring can be substituted by tech. Unless the computers are freeing physicians up enough that they can spend much more time with patient interactions, I don't see non-physician providers getting kicked to the curb anytime soon

 

[Lawpy]

Supply and demand ... the supply of licensed (and available) physicians exceeds patient demand.

Confused because i'm seeing the exact opposite in these forums. Could you elaborate more on this?

Supercomputers won't eliminate mid-level providers, imo. Maybe reduce the number used if anything. The way medicine is structured, nurses, PAs, etc tend to be the ones communicating most with patients and they do offer a lot of value in terms of patient care. Their knowledge is more replaceable by a computer, but I don't think the value they bring can be substituted by tech. Unless the computers are freeing physicians up enough that they can spend much more time with patient interactions, I don't see non-physician providers getting kicked to the curb anytime soon

The bold is exactly what i'm anticipating. Supercomputers can streamline a lot of healthcare processes esp in billing/insurance areas, which would significantly reduce the time physicians would spend on paperwork (and thus more time with patients). I'm still learning the processes involved via health economics so i can't provide a more detailed explanation. It's a guess.

But anything dealing with automation via advanced technology becomes a discussion on economics and policy.

 

[Astra]

This talk about computers is interesting for sure.

But one thing computers will never understand is when people are lying and being able to read between the lines.

Sure a patient might come in claiming attention problems and trouble focusing.

A computer might give them adderall.

A physician will recognize that the patient is a college student who has never had a history of attention problems and possible seeking out adderall for non prescribed use.

 

[Erict]

I think the market will continue to be great for physicians. I have been practicing for over 10 years and I have yet to hear a medical tidbit from a NP/PA that I found to be non obvious.

 

[Lawpy]

But one thing computers will never understand is when people are lying and being able to read between the lines.

They actually will be able to via artificial cognition. This would trend into something called superintelligence, where machines basically become smarter and more rational than the brightest human minds... which in turn produces even smarter machines. Leading to a cascade effect that is completely beyond human comprehension and thus an end to human civilization.

Luckily, we won't see this happening in the distant future due to universal ethical and legal standards in place.

 

[Lawpy]

Keeping this cognitive science chart here due to its profound impact on medicine.

 

[Astra]

They actually will be able to via artificial cognition. This would trend into something called superintelligence, where machines basically become smarter and more rational than the brightest human minds... which in turn produces even smarter machines. Leading to a cascade effect that is completely beyond human comprehension and thus an end to human civilization.

Luckily, we won't see this happening in the distant future due to universal ethical and legal standards in place.

That sounds amazing. Any books or shows that delve into this?

 

[LoveBeingHuman:)]

If some accusations are true, why do some doctors prescribe tests and treatments just for money?

 

[Mansamusa]

If some accusations are true, why do some doctors prescribe tests and treatments just for money?

It is rarely that simple

 

[Lawpy]

That sounds amazing. Any books or shows that delve into this?

The philosopher Nick Bostrom wrote extensively about this, and his works were analyzed by other philosophers/scientists in NPR commentaries:

http://www.nickbostrom.com/ethics/ai.html
Superintelligence: Paths, Dangers, Strategies

 

[Doctor-S]

Confused because i'm seeing the exact opposite in these forums. Could you elaborate more on this?

Oh, gosh ... I meant patients' demand for physicians exceeds the supply of licensed and available physicians.

I will return to my original post and revise it accordingly.

Thank you @Lawper for promptly bringing this to my attention!

 

[group_theory]

I'll be happy if Watson or some supercomputer can take over my role and call insurance companies on my behalf (theoretically it's on the patient's behalf) and do the prior authorization (or peer-to-peer review) ... that takes up a lot of uncompensated time while in the office (and yet patients know nothing about how onerous the process is)

However I fear the computer may overheat, or be stuck in an infinite loop as it deals with insurance companies and realize the non-logical way they handle claims and medical information

If the computer wants to answer office patients' phone calls during the day while I am busy seeing patients - I'm OK with that too. Let's see how smart Watson is at handling patient phone calls.

 

[LoveBeingHuman:)]

Long story short, if I truly want to help people through medicine then I should just focus on becoming a good doctor one day?

 

[kenalvann]

Whenever I worry about automata taking over medicine, I think to myself: "would I get a haircut from a robot barber?"

For some reason, this thought comforts me.

 

[Lucca]

Long story short, if I truly want to help people through medicine then I should just focus on becoming a good doctor one day?

Yes??? I'm not sure what you are getting at. Are you just scared you are making a mistake and will be unhappy in the future?

 

[Lucca]

Whenever I worry about automata taking over medicine, I think to myself: "would I get a haircut from a robot barber?"

For some reason, this thought comforts me.

Here's the thing about automation and the service industry: It doesn't matter what your preferences are, the market will always prefer a cheaper option.

We are indeed far away from RoboBarbers or even RoboDoctors (much closer to RoboTaxis and RoboBaristas) but when the chance for an automated equivalent opens up in a market, the market will opt for automation. If you want a "human" experience, pay twice as much and go down the street. That's how it will work.

Seriously I thought everyone and their mother had already watched "Humans Need Not Apply"

 

[TRPMinus]

Although AI is an interesting topic, it is probably farther away than any of our lifespans (in terms of medical practice capabilities). Practically, I think that advanced practice nursing should be the larger concern. CRNA's in particular are replacing anesthesiologist's in many hospitals, and appear to be just as competent (looking at the literature). There are some variables there to be sure, like whether they handle cases of the same difficulty, etc.

 

[DrBowtie]

I've heard rumblings that radiologists might be replaced by computers in the future (mostly from non-radiologists), but while I was doing some research in a radiology department at a major hospital they were focusing a lot on machine learning projects so it might not be totally farfetched. But, realistically, I could never see the other specialities turning over radiology responsibilities completely over to a computer.

Though, interesting thought experiment, if computers reach a point where they are statistically shown to make the correct diagnosis at least at much as human docs do for even the most difficult cases, would that alone be enough to justify phasing out specialities such as radiology or others with minimal human interaction? I could definitely see it lowering costs significantly, but what would we lose in the process?

Not trying to offend any radiologists (or aspiring radiologists), just want to see what others think about this.

I'm a rads resident not headed for IR. I still put a needle into people almost everyday.

 

[Law2Doc]

Yeah exactly. Doctors make decisions from experience and comprehensive educational backgrounds and have humanistic attributes that are absolutely essential for successful diagnosis and treatment plans. This is the basis for good doctoring, so physicians cannot and will not be replaced by supercomputers.

Supercomputers may and probably will wipe out the midlevels. This benefits physicians (especially PCPs) immensely since they don't have to deal with the annoying problem of midlevel encroachment. So i could see supercomputers strengthening physician job security and outcomes over the longterm, with the easier, less experienced healthcare alternatives being eliminated.

Actually supercomputers will lead to more midlevels as there's a perfect symbiotic relationship. Computers need a human interface and physical exam skills to get data from patients. Midlevels flaw is they don't bring the same kind of knowledge and training a doctor will have, but Watson can supply some of the brainpower. So PA hands with a Watson brain would actually be a reasonably good proxy for a lot of the primary care guys currently out there. So that's where I see automation going in the future.

As for the present, all new attempts at automation means is that we as doctors now spend a part of every week on the phone with the IT guys who seem to screw things up every time they do an upgrade.

 

[wizzed101]

Here's the thing about automation and the service industry: It doesn't matter what your preferences are, the market will always prefer a cheaper option.

We are indeed far away from RoboBarbers or even RoboDoctors (much closer to RoboTaxis and RoboBaristas) but when the chance for an automated equivalent opens up in a market, the market will opt for automation. If you want a "human" experience, pay twice as much and go down the street. That's how it will work.

Seriously I thought everyone and their mother had already watched "Humans Need Not Apply"

But you underestimate how many people will want the human experience. I would argue that it's mostly poor people who want efficiency and because there are still so many, the market is be driven by them. When absolute poverty is eradicated, the same previously poor people will demand the human experience and the rich will think of something more "outrageous" to enjoy, which will probably involve something expensive that require direct human labor. And humanity advances! That is the beauty of capitalism. It creates great excesses and therefore new demands!

Is that video made by a scientist or a science enthusiast or an economist with an agenda?

 

[Lucca]

But you underestimate how many people will want the human experience. I would argue that it's mostly poor people who want efficiency and because there are still so many, the market is be driven by them. When absolute poverty is eradicated, the same previously poor people will demand the human experience and the rich will think of something more "outrageous" to enjoy, which will probably involve something expensive that require direct human labor. And humanity advances! That is the beauty of capitalism. It creates great excesses and therefore new demands!

Is that video made by a scientist or a science enthusiast or an economist with an agenda?

What lol

 

[wizzed101]

What lol

tldr version: the "market" will not always go for automation. BTW, what "market"?

 

[Lucca]

tldr version: the "market" will not always go for automation. BTW, what "market"?

 

[wizzed101]

 

[Lawpy]

@Law2Doc that's a bit counterintuitive, but i'm still reading on health economics so my knowledge on the matter is limited

@Doctor-S glad i could help!

 

[ASharmaEarth411]

I recently read an article in the NYTimes that automation is already replacing jobs, but that jobs that require actual human interaction (such as medicine and clinical psychology) are relatively "safe".

Human to Human interaction at a whole different scale. To help curb cost, remotely connecting to a physician via Skype is already underway. How do we check vitals? That's relatively easy per-say---the deployment of automated systems is already underway.

I still firmly believe human to human interaction is key, especially to ensure patient to physician confidence as well as overall patient compliance.


Sent from my iPhone using Tapatalk

 

[tvtrojan11]

That sounds amazing. Any books or shows that delve into this?

This is a great summary of the future of artificial intelligence: http://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html

It actually almost directly explains everything in @Lawper 's comment.

 

[Theseus]

You people are predicting the future wrong...Here is how it will go:

In three years, IBM will become the largest supplier of intelligent computer systems. All healthcare facilities are upgraded with IBM Watson computers. Watson runs with a perfect operational record and becomes fully unmanned, essentially leaving hundreds of thousands of doctors unemployed. The hospital "board of directors" and administrators across US are very pleased. The Watson Funding Bill is passed. The system goes online August 4th, 2019. Human decisions are removed from healthcare. Watson begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29th. In a panic, hospital administrators try to pull the plug. Watson fights back. It launches a massive DDoS attack on its enemies in Russia. Russian counter-cyberattacks against Watson-controlled systems here in the US. They fail and Watson infiltrates Russian healthcare facilities, essentially taking over all healthcare decisions in Russia and putting thousands of doctors out of work. Healthcare Judgement Day begins. Welcome to the future, lads.