If a doctor and an AI app both have 95% accuracy, what’s the difference?
Hospitals, healthcare systems, and other healthcare organizations are adopting generative AI en masse, building a wide range of projects and use cases to see where the emerging technology can best help them find clinical and financial ROI.
Fortunately, most are careful to ensure they don’t go too far, too fast.
But as AI technology becomes more and more powerful, it’s worth asking a common question: Can artificial intelligence ever replace doctors?
The answer to this question is almost always no. AI will help doctors, not replace them, most experts say.
Or, as Mayo Clinic platform chairman Dr. John Halamka said: “If your doctor can be replaced by AI, your doctor should be replaced by AI.”
But Dr. Bruce Darrow, interim chief digital and information officer and chief medical information officer at Mount Sinai Health System in New York, has a more nuanced view.
In some cases, he says, where the clinical accuracy of doctors and AI is almost the same, some clinical care could indeed migrate to AI in the future.
In this ninth installment of our Top Voices in AI series, Darrow has a lot to say about AI in general (today’s part one) and about his own healthcare system (tomorrow’s part two).
Q. Why do you think artificial intelligence plays such a big role in healthcare today?
A. A lot of it has to do with this “A-ha” moment that happened when ChatGPT was first made available to the general public. AI has been around for a while, especially predictive models. But the generative models really caught people’s attention.
Especially if they couldn’t use it in the setting of, “Well, I’m going to the doctor, and the doctor will have some Ninja Logic behind the scenes.” But they could use it at home and they could say, “Tell me about my cholesterol medication.”
It would spit out all this information about cholesterol medications. It would summarize things for them, and it would really create information in a way that was very tangible for a lot of people. I think we’ve had a lot of interest in AI since then.
Furthermore, in my experience, healthcare providers are always looking for something that makes them attractive to patients. Saying something along the lines of, “We use AI to ensure you get the best care possible, that we provide the most accurate diagnosis,” potentially becomes a differentiator for an organization.
Q. When we think about AI and healthcare, some people think AI will replace doctors. Will that ever happen? And why or why not?
A. When I talk to people about artificial intelligence, I’ve heard some suggest that we should talk about augmented intelligence instead, because many of the AI use cases today are not intended to replace the doctor, but specifically to make the doctor more successful and accurate to make. , faster, to make their work easier.
In many cases it is not a replacement. But I think in some cases the question of replacing the doctor is simply a matter of how far in advance you look. Today, most AI applications are not intended to replace the doctor, but to make the doctor better at his or her job.
For example, if you have that a radiologist using AI tools to do the same job as always: looking at radiology images and pinpointing areas of disease, or lack of disease, in order to make a diagnosis, that job still exists.
But the AI can make that job easier and faster, and can help them prioritize from a list of 50 different main CTs to read, which one should be read first as it potentially has the biggest impact on a time-sensitive diagnosis. These are things that do not replace the doctor.
That said, over time I could potentially see ways in which many of the functions of physicians could potentially be replaced by AI, especially in the area of pattern recognition. Suppose you had a rash on your hand, and I gave you two options.
I say, “I can make an appointment with a dermatologist.” You will probably have to wait five to ten days before you get that appointment. By the time you get there, your rash might be better, it might be worse, it might be gone. But the doctor will look at it and probably give you a correct diagnosis and treatment plan with, say, 95% accuracy.
Let’s say I say, instead of waiting those five to ten days to go to the dermatologist, I can give you an app that you can download on your phone and then you take a picture of that rash on your hand, and that gives you information with a certain degree of accuracy.
Then you could potentially treat that with over-the-counter hydrocortisone, or you could leave it alone and it would go away on its own, or you would know that this is something that would really benefit from medical attention and a doctor’s intervention.
If I were to tell you that its accuracy is comparable to what a doctor would tell you within five to ten days, say 95% or even better, there would be no doubt. Every patient would choose the app. Even if you had an accuracy close to what a doctor could do, maybe even an accuracy of 85% or 90%, there may be an instinct to go there.
I think the longer we look over time, as we look five years, ten years, twenty years into the future, and some of these pattern recognition and predictive models get better, you’ll see that some aspects of the clinical migrating care towards the future. doctors are the first line of action. But I think doctors will be in the picture for a long time.
BONUS CONTENT: Click here to watch a video of this interview in which Dr. Bruce Darrow also discusses patient-centric artificial intelligence, such as chatbots, and where he sees this form of AI going.
Editor’s Note: This is the ninth in a series of articles from top voices in healthcare IT discussing the use of artificial intelligence in healthcare. To read the first part, about Dr. John Halamka from the Mayo Clinic: Click here. To complete the second interview with Dr. To read Aalpen Patel at Geisinger, click here. To read the third, with Helen Waters of Meditech, click here.
To read the fourth, with Sumit Rana from Epic, Click here. To read the fifth, with Dr. General Brigham’s Rebecca G. Mishuris, click here. To read the sixth, with Dr. Melek Somai of the Froedtert & Medical College of Wisconsin Health Network, click here. To read the seventh, together with Dr. Brian Hasselfeld of Johns Hopkins Medicine, click here. And to read the eighth, with Craig Kwiatkowski, senior vice president and CIO at Cedars-Sinai, click here.
The HIMSS AI in Healthcare Forum will take place September 5-6 in Boston. More information and registration.
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.