Artificial intelligence replacing the PA??

[corpsman89]

Hi everyone,

 

AI and robotics have always peaked my interest because I recognize that there is an extreme threat of this new technology replacing many jobs in our economy. Technology is prevalent throughout our life, and the health care field is no exception. And while I am very glad we have this technology I am beginning to worry about how rapidly we are expanding its use and I am now beginning to see a legitimate threat to even the health care provider profession from AI and robotics.

 

Do any of you agree with this? Do you think we will (in time) see a slow elimination of PA jobs due to AI and robotics?

 

I am nervous to say that I really think that PAs and NPs will be one of the first health care providers to suffer from this rapid advancement in technology. Reason being is that a Physician is a Physician. They are always the last say in all matters and I don't see their jobs being taken by AI.  The big question is when?

 

What are your thoughts?

[jmj11]

Eventually . . . but many decades off. We have been working on the AI- Headache Specialist program that can not only diagnose headache disorders precisely, but prescribe the exact work up and treatment for that patient. I spent a year working on an algorithm for diagnosis, but going from that point to treatment is a huge jump. Adding a good exam would be another leap. The things that we worked on (and will be there in a few years) would just be a tool for primary care where a patient completes a very detailed history (my takes about an hour) and then spits out the diagnosis and treatment plan.  But that is one very narrow area of headache. Just imagine doing this for many other areas of medicine and them sewing them together. So, I think your career is safe. Now if you had a kid (young kid) who wanted to grow up and be a PA, I think it would be reasonable to think that some of their work could be replaced by AI.

[LT_Oneal_PAC]

Symptoms are only one part of making the diagnosis. Greater in some specialties, like headache, but physical exam is a big part. Considering we don't even have cars that can 100% reliably drive on the street autonomously, it's going to be a long time before they can program to physical exam, much less interpret the vastly different human responses to the tests. It will be even longer before they can be programmed to perform procedures. Look at pilots. We've long had the technology for planes to fly themselves, but there is always a pilot in that seat. Then there is the risk benefit discussion and shared decision making. Sure, some would be fine with just stats thrown at them, but many people want the answer to "would you do it?"

 

Further, I perform the exact same job as a physician in my specialty. If I go, you better believe the administrator is only a half step away from doing the same to the physician.

[Marinejiujitsu]

I a world where most patients say yes to every symptom, we are long way off.

 

Sent from my SAMSUNG-SM-N920A using Tapatalk

[MT2PA]

Look at this from a patient view - they *might* not be stoked about a robot in charge of their health and/or problems.

 

I'd love to know the stats on telemedicine (the 'Skype with a nurse and get your strep checked out!') because I'm not convinced even something that tech related is doing well.  My experience is that insurance pushes it vs patients actually wanting it.  I'd imagine the same pushback, or worse, would occur for AI.  

[Katera]

Ai no.........Telemedicine?  Oh yes most definitely.  It already is.  Check out Teledoc for example.  Out of of intense curiosity I signed up for teledoc.  It is $3/month and $45/visit.  I used them to refill some cholesterol med I take just to see how it worked.  Once I signed up online, filled out a quick medical history, the doc called me back after about 15 minutes, asked a few questions and called in my scripts.  The whole thing cost $48.

 

Right or wrong, it was VERY easy and VERY cheap.  Now this is much more likely to replace us then AI.  One provider sitting in his underwear could interact with 75 patients per day easy with virtually no overhead.  Ouch.

[MT2PA]

Ai no.........Telemedicine?  Oh yes most definitely.  It already is.  Check out Teledoc for example.  Out of of intense curiosity I signed up for teledoc.  It is $3/month and $45/visit.  I used them to refill some cholesterol med I take just to see how it worked.  Once I signed up online, filled out a quick medical history, the doc called me back after about 15 minutes, asked a few questions and called in my scripts.  The whole thing cost $48.

 

Right or wrong, it was VERY easy and VERY cheap.  Now this is much more likely to replace us then AI.  One provider sitting in his underwear could interact with 75 patients per day easy with virtually no overhead.  Ouch.

 

There's a limit to what they can do - and it if it keeps people out of the ER/UC for things that don't need to BE in the ER....sure, great.

 

At some point (or level of disease) the risks of not actually seeing the pt will outweigh the cost savings.  As a provider I want to be on the right side of that line.  As a patient, eh.  If it does the same thing a minute clinic does and I don't have to leave my house, sure.  But it won't be replacing my need for most services.

[Aunt Val]

Fun topic.  :)  So what do you all think a PA-replacing technology would be like if and when it is developed?  A giant collection of computer algorithms similar to what jmj11 described?  A no-patient-touch scanner a la Star Trek?  A walking, talking human lookalike that could perform a physical exam and use reason to interpret data gleaned?

 

Personally, I think the Star Trek thing would be cool.

 

Realistically, though, I doubt that an actual human being could ever be replaced at the bedside.  We're lightyears away from understanding intelligence and having the technology to replicate it.  Besides, isn't part of patient care supposed to be caring for the patient?  Things like holding a hand or taking an extra minute to just listen.  Maybe computers will get extra-smart someday, but I don't think that AI will be able to care.

[corpsman89]

Nurses at the bedside are safe.

 

It could be that RNs will slowly (they already are) gain more and more autonomy and will be able to do more complex procedures and decision making processes. With the help of this, I think AI will be possible--in time of course. Basically AI will be the main decision making process, then the RN will implement this process and will do all checks necessary before final treatment is determined and made. Physicians will audit records to make sure everything is working properly and for oversight.

[LT_Oneal_PAC]

Why do you think nurses at the bedside are safe? They are the least safe. When tech gets to the point I can have a monitor alert me directly based on vital signs and an algorithm, automatically dispense meds from a vending machine with a wrist band patient scanner, and so on, I can hire a CNA to do the rest. As a former bedside nurse, they are far more replaceable than me.

[corpsman89]

Lt- Public perception. Most of our society understand that RNs are highly trained professionals. I have several family members who are nurses and some of my non-medically trained family members basically think they're doctors. RNs are gaining tremendous traction in the health care field. Just recently it was determined that a nursing degree is the equivalent to a biological sciences degree. CMS declared this to be the case and has determined that now RNs are eligible to run high complexity testing within the laboratory. This is just yet another example of their power in the health care industry. In the end, I don't think the public will be comfortable with a CNA or PCT being the ones who are primarily taking care of them.

 

Also, many RNs are at the point where they are pseudo providers in many family practice type of organizations. The military is a good example. Both active duty, and the VA utilize RNs to take care of low acuity conditions and many patients don't mind because they get seen quickly and treated effectively. They know the Doc/PA is overseeing what the RN does, but all they see, and talk to is the RN. With this type of reputation it is not farfetched to see an RN as our futures' doctor with AI making the major diagnosis and treatment calls. The RN will simply follow the AI order while also insuring that said decisions are safe and effective.

 

However, I also could see what you predicted as well. Its all about public perception. If the model you predicted were to happen it would have to be gradual to the point where the public trusted that model. I just see RNs having an advantage simply because of their title. There would be less questions asked from patients if they simply knew that they were being taken care of by an RN and not a technician.

 

I know this is very far away from happening, but that is my best prediction.

[Aunt Val]

So advancing technology could majorly transform a lot of medical professions besides just PA/NP.

 

 

 

Nurses at the bedside are safe.

 

It could be that RNs will slowly (they already are) gain more and more autonomy and will be able to do more complex procedures and decision making processes. With the help of this, I think AI will be possible--in time of course. Basically AI will be the main decision making process, then the RN will implement this process and will do all checks necessary before final treatment is determined and made. Physicians will audit records to make sure everything is working properly and for oversight.

 

An RN that is more autonomous and can do more complex procedures and decision making, kind of an "RN plus"?  Sounds like an NP to me, or at least a morph in the mid-level direction.

 

 

 

Why do you think nurses at the bedside are safe? They are the least safe. When tech gets to the point I can have a monitor alert me directly based on vital signs and an algorithm, automatically dispense meds from a vending machine with a wrist band patient scanner, and so on, I can hire a CNA to do the rest. As a former bedside nurse, they are far more replaceable than me.

 

You both have good points that there would still need to be someone overseeing the machines and interacting with patient.  Whether that's an upcycled RN or a "nurse" with less training (ie a CNA) would depend on the type of technology, I guess.

 

 

 

One perceived benefit of AI in healthcare is that it could help ease the medical provider shortage.  There would have to be people with medical acumen involved in programing all these devices, though.  MDs might have to have advanced IT training as part of med school, even a joint MD/IT/programing degree.

 

The problem with all this is that very few people would know what to do or how to fix it if the machines quit working, electricity went out, natural disaster happened, or some other kind of crisis.

[corpsman89]

^ Solid point about the technology crapping out or a natural disaster situation.

 

The laboratory is a good example of technology advancement. The lab used to do all tests manually and had many highly trained individuals that--from what I have been told--even physicians would call down asking for advice about what tests to run in certain cases, etc. Now, practically all labs are performed with highly complex analyzers which are managed by technicians who are more skilled in troubleshooting the instrument then the testing itself. And while I do have a solid foundation of the basis of all tests that I perform I will not venture to say that I am nearly as knowledgeable as the technologists were several decades ago. This simply goes back to what you said about MDs receiving more training in IT and programming, I think that is a fair prediction for sure.

[Aunt Val]

CMS declared this to be the case and has determined that now RNs are eligible to run high complexity testing within the laboratory. This is just yet another example of their power in the health care industry.

 

Side note:  BAD IDEA.  I work in a small hospital lab now (as a phlebotomist, working closely with the techs).  There is a lot more that happens in the lab besides just putting a tube in the machine and pressing a few buttons.  The techs have to do calibrations and QC every shift, do some machine maintenance if something messes up, recognize questionable/inaccurate results and know what factors may cause them, be able to keep up with records and requirements for Joint Commission (labs get their own separate JC inspection every 3 years in addition to the general hospital inspection), and much more.  Unless nurses are trained in the equivalent of a laboratory science education (whether classroom or on the job), I think that a lot of things would be missed.  Not trying to put down nurses, but lab science isn't their field.  Most of the nurses at my hospital don't know the difference between a hemolyzed specimen and a clotted specimen and couldn't tell you the order of draw for blood tubes or why it even matters.

[Aunt Val]

^ Solid point about the technology crapping out or a natural disaster situation.

 

The laboratory is a good example of technology advancement. The lab used to do all tests manually and had many highly trained individuals that--from what I have been told--even physicians would call down asking for advice about what tests to run in certain cases, etc. Now, practically all labs are performed with highly complex analyzers which are managed by technicians who are more skilled in troubleshooting the instrument then the testing itself. And while I do have a solid foundation of the basis of all tests that I perform I will not venture to say that I am nearly as knowledgeable as the technologists were several decades ago. This simply goes back to what you said about MDs receiving more training in IT and programming, I think that is a fair prediction for sure.

 

Your reply came out while I was still typing.  :)  You're a lab tech?  Cool.  Guess I don't have to tell you about lab stuff then.  Yeah, there are a couple older techs at my lab who talk about stuff back in the day . . . they'd be invaluable in developing countries that don't have all the analyzers we have here in the U.S.

[corpsman89]

Yup, lab tech here. Former Corpsman too, but a lot of this worry about technology and AI comes from working in the laboratory. At an old laboratory I used to work at they had a Rouche analyzer that did everything. All it required was an assistant to load the sample that was just drawn and results would spit out the other side. This was amazing to the hospital administrators who didn't understand the complexity of issues that would arise. But this is the way we are heading. In terms of the lab itself I can see a future where the lab is simply a place where samples go and results come out with no constant supply of people in the lab, just field service engineers who make sure the machines work properly. Again, this is a long way a way but it seems that's the way it is heading.

 

But as many have suggested. diagnosing and treating are another animal and will take much longer for AI to master, if ever.

[BruceBanner]

I think our jobs are safe. We will see more and more telemedicine, as mentioned.

 

Kaiser is pushing this now. There are docs here who--quite literally--take TEXTS from patients and offer advice. Kind of like an RN sick line but staffed by docs.

 

We also have phone visits, which are somewhat common, and video visits.

 

The vast majority of patient visits are seen in the clinic, but telemedicine is catching on. I think it will only go so far though due to liability. Unless it's something very straightforward, most providers will reflex to "just come into the clinic."

[UpRegulated]

Hi everyone,

 

AI and robotics have always peaked my interest because I recognize that there is an extreme threat of this new technology replacing many jobs in our economy. Technology is prevalent throughout our life, and the health care field is no exception. And while I am very glad we have this technology I am beginning to worry about how rapidly we are expanding its use and I am now beginning to see a legitimate threat to even the health care provider profession from AI and robotics.

 

Do any of you agree with this? Do you think we will (in time) see a slow elimination of PA jobs due to AI and robotics?

 

I am nervous to say that I really think that PAs and NPs will be one of the first health care providers to suffer from this rapid advancement in technology. Reason being is that a Physician is a Physician. They are always the last say in all matters and I don't see their jobs being taken by AI.  The big question is when?

 

What are your thoughts?

 

Ask Siri what she (he) thinks.  You'll have your answer. Seriously.

 

My answer?  Not in our lifetimes.  Despite all the media buzz, when your smartphone fails you miserably at least 50% of the time, how can we expect AI to make complex medical decisions when your phone AI developed by billions of dollars in research and engineered by some of the smartest people on the planet can't reliably tell you where the closest gas station is?  

 

I was in technology/computer science/computer engineering/network analysis for years.  AI is many decades away from being anything that could even remotely be used effectively in medicine as a provider replacement.  AI works by mathematical algorithms based on simple either/or Boolean logic type decision trees (and to "humanize" it, e.g. Siri, volumes of popular works of literature are scanned in to decision database to make AI sound human). Medicine is not a checklist of symptoms, nor is it a decision tree, yet that is all "AI" could do 30 years ago, and it is still all that it can do.  The basic fundamentals of AI are unchanged. 

 

AI is great for purely mathematical operations -- such as chess -- when there is a limited number of variables.  But it fails miserably when there are a large number of variables, such as predicting the weather.  The variables in medicine are quite large (larger than the weather), and there is no AI system on the planet now or anytime in the near future that can even come close to accounting for them.  So again, ask Siri (or the equivalent) something.  Not even basic questions can be answered accurately despite billions being poured into AI for these functions utilizing the greatest minds on the planet.  Now consider the patient with hyperlipidemia that fears ALL cholesterol medications and won't take them because her mother died of a heart attack despite being on statins for years (a real patient of mine).  That's a relatively simple case for the human mind!  But there is no AI, nor will there be in our lifetimes, an AI algorithm that can deal with that.

[Asclepius]

AI is great for purely mathematical operations -- such as chess -- when there is a limited number of variables.  But it fails miserably when there are a large number of variables, such as predicting the weather.  The variables in medicine are quite large (larger than the weather), and there is no AI system on the planet now or anytime in the near future that can even come close to accounting for them.  So again, ask Siri (or the equivalent) something.  Not even basic questions can be answered accurately despite billions being poured into AI for these functions utilizing the greatest minds on the planet.  Now consider the patient with hyperlipidemia that fears ALL cholesterol medications and won't take them because her mother died of a heart attack despite being on statins for years (a real patient of mine).  That's a relatively simple case for the human mind!  But there is no AI, nor will there be in our lifetimes, an AI algorithm that can deal with that.

 

I concur. Myriad of variables to take into consideration with respect to patients/medicine. However, in my opinion, the medical profession, in large measure, is technology resistant. I work almost exclusively with anesthesiologists in a large top 10 hospital, (not as a PA), and there seems to be an undercurrent of disquiet and resentment when talks of AI come into play; especially amongst physicians. Perhaps, these feelings are predicated on actually feeling expendable for a chance...(e.g. Sedasys Anesthesiology Machine). One doesn't mind the outsourcing of labor, until his/her job gets outsourced - *Although, I recently read that Johnson & Johnson ultimately decided to pull the Sedasys machine as a result of poor sales. But that's not to say the machine itself wasn't successful, it just galvanized enough push-back from anesthesiologists/American Society of Anesthesiologists that hospital admin probably didn't want to deal with. Also, for the most part, I believe that upper echelon hospital admin and their relentless ways of keeping their pockets lined, would love to see procedural AI take off. They would view it as a cost-effective means to generate more revenue...(for them).

[corpsman89]

So now that we have determined that we likely will not see AI in our lifetimes, and that we will likely see the expansion of telemedicine . . .

 

What role does a PA play in the future of telemedicine?

[Reality Check 2]

I don't want to play a role in telemedicine.

 

I can see its role in incredibly rural situations or with mostly nontangible fields such as psychiatry.

 

However, it has already been tainted as a control tool to keep folks from leaving work and as a perceived cash cow.

 

I have zero intention of ever participating in it as either patient or provider.

 

Hands on, gut feelings, intuition and real vitals cannot be replaced by AI or telemedicine.

 

The subtleties of rhonchi, wheezes, skin tone, eardrums, posterior pharynx simply cannot be simulated or discovered on an iphone screen or even ipad. Strep smells, mono smells, illnesses have subtle colors and odors.

 

If faced with a patient with symptoms begging a differential of CHF or pneumonia - I would say I could not diagnose effectively and the patient should be seen in a real clinic. The patient still gets charged and is no better and treatment has been delayed. What a waste.

 

So, telemedicine is not for me

[corpsman89]

While I am sure that many providers will agree with not wanting to work in telemedicine, wouldn't it be wise to embrace it? Maybe we can work to make it better? One of the things that keeps many companies and industries going strong is being able to peer into the future, view trends, and act upon those indicators to create something new. The ones who fail to do this . . . fail.

 

The obvious indicator is that telemedicine will be expanded and embraced by not only patients, but politicians as well, which we all know is the true driver in health care, and that wont change. (sadly).

[Reality Check 2]

At this stage in my career - no, I do not want to embrace telemedicine.

 

I do not have a compelling need to do so for my job.

 

I also can't think of any way to make it better except to make it an actual clinic and then no need for the tele part.

 

It is a belief I hold deeply and have seen it be a disaster so far in urban use.

 

Others may want to advance the cause. I will stay on my horse and buggy....

[Fernwood]

I was told in a meeting last week that telemedicine is the future, brick and mortar medical offices are going away. This came from our director of clinic operations. This is the most absurd idea I have ever heard. I currently work in urgent care and we have telemedicine visits through the urgent care clinics. How is this good medicine? When the importance of the physical exam simply become a negligible aspect of a patient encounter. I am drafting a letter to our director and will simply ask why it is we bother with all the intake vitals? Why bother with exam tables, otoscopes and stethoscopes? All that silliness of running strep, flu, and urine tests? Just park them in a room , I’ll ask a few questions, they can give their best guess and I’ll prescribe something. I have read articles actually claiming that telemedicine is providing “better care”. Who is buying this load of crap. This parallels what the ACA did with health insurance. The proponents of the ACA love to tell you about the thousands of additional Americans who now have health insurance. The fact is there may be more Americans running around with health insurance but they don’t have any healthcare. Many of the ACA approved healthcare insurance policies provide little in the way of true healthcare benefits. They have high copays and high deductibles. Many offices don’t accept these sham insurances. Telemedicine boasts “increased access”, that’s fantastic but access to what? Our telemedicine is simply a Skype visit. That’s absolute crap. I had a doctor refuse to refill a kids albuteral nebulizer medication because he wanted to see the child. The child was in no distress, having no exacerbation of his mild asthma, completely asymptomatic. The physician had no clinic time available so he instructs the child’s mother to take the child to urgent care? What?this physician works within the same healthcare system I do. This is the same healthcare system that is promoting telemedicine. So it’s not okay to refill a script for an established patient who is asymptomatic but it’s completely fine for me to Skype with a new patient who is having difficulty breathing and prescribe an albuterol inhaler over the phone. This is simply the most idiotic thing I have ever heard. With this sort of thinking artificial intelligence will be replacing medical providers. They will use google derived algorithms to diagnose and treat. The upside is they can update your Netflix subscription at the same time.

I suspect my time with this healthcare organization will be short lived as I have actually laughed at the management individuals who are pushing this absurdity. These individuals are of course physicians who coincidentally don’t actually practice medicine anymore.

[katieo]

Telemedicine is a god send for UTI. If it fills this niche alone I think it’s a success.