VIDEO: What are the promises and perils of AI in healthcare?
In the latest episode of “Office Hours with Dr. Saini,” Lown Institute president Vikas Saini answers the question: What are the promises and perils of AI in healthcare?”
“What does AI mean for the art of healing? It’s an interesting question, and the short answer is that it could be liberating.”
Dr. Vikas Saini
Welcome to Office Hours with Dr. Saini. I’m Vikas Saini, president of the Lown Institute. In this series, I tackle questions about the healthcare system as honestly as I can, one question at a time. This week’s question: What are the promises and perils of AI in the healthcare sector?
Register before prices increase!
The healthcare affordability crisis isn’t a partisan issue. It’s a human one. This one-day conference brings together leaders who recognize that solutions won’t come from staying in our corners.
Q1: How can AI be useful in medicine?
The one area where AI might be useful I think generically is, since a lot of human intelligence is pattern recognition and since a lot of kind of diagnostic acumen is pattern recognition – it’s not all – some of it has to do with a more methodical way of looking at at you know the odds, what AI is going to do is allow many people to operate at a higher level in terms of some of the prompts for for diagnoses they might not have thought of and the like. If you have something that isn’t quite a fit, if you had a new disease, for example, or a new virus, I wonder, you know, could that ever happen? A new virus? That’s an area where you might actually run astray. And so I think one has to be cautious there as well.
Q2: What does AI mean for the art of healing?
What does AI mean for the art of healing? What an interesting question. I think the short answer is it could be liberating. When you know when a physician is working with a patient, especially if it’s a new patient or if it’s a new symptom and you’re trying to sort it out, you go to work. There’s a lot you have to cover and things you don’t know. You got to check the literature. You got to do this, you got to do that. And even then, it’s tough. Similarly, figuring out treatment options or even a plan of treatment takes up time. And I think the biggest complaint that patients have is they don’t get enough time. And the reality is, and this is one of the paradoxes of healthcare is that you’re actually as a clinician, you’re more efficient, not the less time you take, the more time you take. Because the more holistic your understanding of the situation, the easier it is to find that through-shot that actually addresses multiple things at once. But the deployment of AI could – and not current AI, but an AI of the future could easily make that seamless and painless in a way that it frees up space for the being together that I think is an important part of the healing process.
So then let’s think about what the future might hold. And, because I’m a doctor, I like to think about “What is the role of a doctor in that world?” And I actually think it could be incredibly liberating. And by that I don’t just mean liberating from coding and billing and the burden of commerce. I’m actually talking about liberating in terms of some of the anxiety about you forgot to think of a certain diagnosis or even the understanding that you have to spend a lot of your time figuring out this or that when instead you can now really be present with the patient. Really, be in a relational conversation with the patient in which the imparting of information – both from the patient to you and from you to the patient – is faster and automated, freeing up both of you to explore the implications of it, to explore the emotional tenor of it. And I will submit that that’s really one of the highest callings of physicians is to be present with the person together facing an illness.
Q3: Will AI lead to fewer specialists?
I don’t think we’re anywhere near the point where AI will lead to fewer specialists. In a hundred years, I could see robots that are highly highly intelligent that can do robotic surgery. They probably be able to do a bunch of robotic procedures and be able to talk to you. So I can imagine that world and I don’t think it’s anywhere near happening. I think where it might matter, again, if we’re trying to bend the cost curve, where it might matter is in places in which the same work can be done by fewer people. But I have to re-emphasize that in healthcare one of the problems we have is there is no policeman. It has to be done in a way that not just improves patient care or some metric of outcome. It improves the experience. And so if we can do that, that’s a win. Otherwise, if you have the same reimbursement system and the same chasing for dollars, then what you’re going to do is you’re going to have a lower nurse patient ratio and you’ll have a robot walking into your room and you know, dystopia.
Q4: What scares me most about AI in healthcare
What scares me most about AI in healthcare is what scares me about AI in general, which is that so far we seem to be dependent on a very small handful of companies that are duking it out looking to become near monopolies. In that kind of world, the opportunities and the possibilities to realize not just the savings but some of the efficiencies and some of the less commercial aspects of the benefits of AI, the opportunities to realize those are going to be foreshortened, curtailed, constrained by business considerations, by monopolies and by the inability of smaller operators to really deploy their tools and reach a market.
Well, that’s all the time we have for today. So, tune in next time. If you’re interested in our work, in our events, publications, go to lowninstitute.org and check us out.