Hospitals, health systems and other healthcare organizations are increasingly embracing the apply of generative AI, pursuing a wide range of projects and apply cases to see where emerging technology can best lend a hand them achieve clinical and financial return on investment.
Fortunately, most are cautious and don’t go too far or too quick.
But as AI technology becomes more powerful, it’s worth asking a common question: Can AI ever replace doctors?
The answer to this question is almost always no. According to most experts, artificial intelligence will lend a hand doctors, not replace them.
Or as Mayo Clinic platform president Dr. John Halamka said: “If your doctor can be replaced by artificial intelligence, your doctor should be replaced by artificial intelligence.”
But Dr. Bruce Darrow, interim chief digital and information technology officer and chief medical information officer at Mount Sinai Health System in Up-to-date York, has a more sophisticated view.
He says that in some cases, where the clinical accuracy of doctors and AI is almost equal, some clinical care could actually shift to AI in the future.
Q: Why do you think artificial intelligence is having such a substantial moment in healthcare today?
AND. Much of this has to do with the “Aha” moment that occurred when ChatGPT was first released to the general public. Artificial intelligence has been around for a while, especially predictive models. But generative models have really caught people’s attention.
Especially when they couldn’t apply it in a situation of, “Well, I’m going to the doctor, and the doctor will have some ninja logic behind the scenes.” But they could apply it at home and they could say, “Tell me about my cholesterol medication.”
It would spit out all the information about cholesterol medications. It would summarize everything for them and really create the information in a way that was really actual for a lot of people. I think we’ve been very interested in artificial intelligence since then.
My experience also shows that service organizations are always looking for something that will make them attractive to patients. So saying something like, “We use artificial intelligence to provide you with the best possible care and the most accurate diagnosis,” becomes a potential differentiator for an organization.
Q. When artificial intelligence and healthcare come to mind, some people believe that artificial intelligence will replace doctors. Will it ever happen? And why or why not?
AND. When I talk to people about artificial intelligence, I have heard some suggest that we talk about augmented intelligence instead, because many apply cases for artificial intelligence today are not about replacing the doctor, but specifically about making the doctor more successful and more true faster to make their work easier.
In many cases it is not a replacement. However, I think in some cases the issue of replacing a doctor may simply be a matter of how far in advance you look. Currently, most applications of artificial intelligence are not intended to replace doctors, but to make them better at their job.
For example, if you have a radiologist who uses artificial intelligence tools to do the same job they have always done, which is to look at radiological images and pinpoint areas of disease or lack thereof to make a diagnosis, this profession still exists.
However, artificial intelligence can make this task easier and faster, it can lend a hand them prioritize from a list of 50 different head CT scans to read, which one should be read first, as this has the potential to have the greatest impact on diagnosis, the time of which depends on time. These are things that cannot replace a doctor.
Having said that, over time I could potentially see how many functions of physicians could be replaced by AI, especially in the areas of pattern recognition. Let’s say you had a rash on your hand and I gave you two options.
I say, “I can make an appointment for you with a dermatologist.” You will likely have to wait five to ten days to schedule an appointment. By the time you get there, your rash may have cleared up, maybe it’s gotten worse, or maybe it’s gone. But your doctor will check it and will likely give you the correct diagnosis and treatment plan with, say, 95% accuracy.
Let’s say that instead of having to wait those five to ten days to see a dermatologist, I can give you an app that you can download to your phone and take a picture of that rash on your hand and you’ll get the information with some degree of accuracy.
Then you could potentially treat it with over-the-counter hydrocortisone, or you could leave it alone and it would go away on its own, or you could know that this is something that would really benefit from medical attention and medical intervention.
If I said that the accuracy of this test is comparable to what the doctor will say in five to ten days, say 95% or even greater accuracy, there would be no doubt. Every patient would choose the app. Even if the accuracy was close to what a doctor could achieve, maybe even 85% or 90%, there might be an instinct to go there.
I think the longer we look back over time, five, 10, 20 years into the future and see that some of these pattern recognition models and predictive models are getting better and better, you can see certain aspects of clinical care moving away from physicians being the first line approaches. But I think doctors will continue to be in this position for a long time.
BONUS CONTENT: Click here to watch a video of the interview where Dr. Bruce Darrow also discusses patient-facing AI such as chatbots and where he sees this form of AI.