On today's episode of WNYC's new popup advice show, tech writer Manoush Zomorodi, host of the NPR's Ted Radio Hour and author of Body Electric: The Hidden Health Costs of the Digital Age and New Science to Reclaim Your Well-Being leads a conversation about weight loss medications, longevity science and more trends in health technology, with Dr. Dhruv Khullar, practicing physician, contributing writer at The New Yorker and an assistant professor at Weill Cornell Medical College. Listeners call in with their experiences and questions about recent health and medical advancements.
Note: You are listening to the podcast version of a live radio show. The call-in number only works between 1 and 2 PM Eastern, while the show is live.
If you'd like to submit advice questions to be featured on the show, or topic suggestions for future shows, please use the form on our showpage at wnyc.org/advice
[music]
Manoush Zomorodi: Welcome to Please Advise: The Pop Up Advice show from WNYC. I'm Manoush Zomorodi, host of NPR's TED Radio Hour and author of the book Body Electric: The Hidden Health Costs of the Digital Age and New Science to Reclaim Your Well-Being. This week I am also your host on Please Advise. Thank you so much for joining us today. All week, we are giving and getting advice on how to navigate technology and innovations that impact our daily lives for good and bad. Tomorrow we'll take your questions around the challenges of AI and work. Today we are talking about our health trends and the latest in medical tech and research.
Our lines are open at 212-433-WNYC. That's 212-433-9692. Call us, we would love to hear from you. Questions like, do you feel like you trust ChatGPT more than your doctor? What has been your experience with wearable tech? Do you love tracking your sleep, or is it making you nuts? How about GLP-1s? Are you trying to decide if they're right for you? Perhaps you've heard they do more than just make people lose weight. Or maybe you're trying to make sense of all the health trends on social media, like peptide stacks. Call, let us know where you need help. 212-433-WNYC.
Or maybe you have some great advice to give. Are you a healthcare worker trying to prevent medical misinformation? What questions do your patients most ask you, and how do you support them? We want to hear how you handle the constant barrage of health trends and headlines. Call us at 212-433-9692. We are not alone, because here to help us is Dr. Dhruv Khullar. Dhruv is a physician at Weill Cornell Medicine, a health policy researcher, and a New Yorker magazine contributor who writes about healthcare. He is joining us today to answer your questions about trends in medicine and healthcare that feel confusing or frustrating, concerning or maybe exciting. Dhruv, welcome to Please Advise on WNYC.
Dhruv Khullar: So great to be here. Thanks for having me.
Manoush Zomorodi: Oh, it's my pleasure. Listeners, we invite you call in for advice on the latest issues and trends in healthcare. Maybe you're curious about how much protein you really need because boy, does it feel like I go to the grocery store and everything is labeled with more and more and more grams of protein. Do I need that? We're going to find out. What about a COVID shot? Have you signed up for one? Maybe you're wondering if you need to.
Maybe you want to understand how AI is impacting your healthcare. I went to get a mammogram the other day, and they asked me if I wanted to pay $50 extra to get AI to look at my boobs. So many things to discuss, Dhruv. You can call us at 212-433-WNYC. That's 212-433-9692. Oh, we should make a disclaimer, though, Dhruv. I want to be clear. You can give us general advice, but not specific advice without seeing someone as your patient. Folks, if you have a very specific medical question, know that you should always see your medical provider, right?
Dhruv Khullar: That's right. Exactly. I can talk generally, but in order to get personalized advice, please see your own clinician.
Manoush Zomorodi: While we wait for folks to call in with their questions or their advice, I want to talk peptide stacks, Dhruv. As a tech reporter, this is something that I've been reading about for years. It's a trend that kicked off in Silicon Valley. Then when I bring it up with normal people, they're like, "What are you even talking about? People are injecting themselves with what?" Can you explain what are peptides? What are stacks of them? Why would anyone put them in their body?
Dhruv Khullar: Well, let's take a step back and talk about what peptides are. Peptides are basically micro proteins. Most people know about proteins as strings of amino acids, and if a string of amino acids has more than, let's say, 50 or so amino acids, it's called a protein. If it has fewer, it's generally called a peptide, like a microprotein. Scientists have known about these peptides for decades. A lot of them are actually rigorously studied and have been turned into safe and effective drugs. People might be surprised to hear that insulin is actually a peptide. These days, the most famous peptide is, of course, GLP-1, glucagon-like peptide-1.
Again, these have been studied and FDA-approved, and people can use them safely. What we're talking about here are unapproved peptides. They're sometimes called gray market peptides or popular peptides. People are making all sorts of claims about them on social media. Things like better endurance, and more strength, and faster recoveries. Who wouldn't want these things? The important thing to know here is that many of these peptides that are circulating, and they have names like BPC-157 or TB-500, they sound sciency and cool.
They may not have a single rigorous study in humans to support their safety and efficacy. Before we put these things in our body and people can order them online, they're getting from compounding pharmacies, from certain types of longevity clinics. It's important to know that basically, we have no data on what these things are doing inside the body, inside humans. I really advise a lot of caution before people choose to inject them into themselves.
Manoush Zomorodi: It's interesting because in your essay, your New Yorker essay on the recent popularity of peptides, you wrote that if the healthcare system fails to meet the needs of patients, people will look for their own solutions. What does the popularity of peptides tell you about where we are right now? I spoke to someone who was taking peptides, and he said to me, "Look, we're living longer. Science has not caught up with what is going on and what people want out of treatments."
Dhruv Khullar: Well, the first thing to say is that I think it's fantastic that people are excited about improving their health span, improving the ways that they can take control of their health. These types of things should be encouraged. It's a really positive sign that we are in a place where people are really excited about these types of things. Because of that, people deserve good information about what these types of products are doing. In my view, a lot of the rise of these peptides has more to do with branding and claims of easy fixes on social media.
Again, we don't have the science to support their use right now. That's not to say that we might not in the future. I do think that some of these peptides could have some use in some people for some things. Filling in those blanks is going to take a lot of study to figure out who might benefit and who might actually be harmed. If you think about even the safest medications, the best medications that you think are generally well tolerated, take Tylenol. Even Tylenol has side effects.
Even it has a maximum dose over which it could be toxic and a minimum dose under which it might not work. We don't know any of those types of things for these peptides. People are choosing to inject them inside their body based on what they might read on Reddit or hear from a friend. That's just not the way that we're going to go about figuring out whether these things work and for whom they might be effective.
Manoush Zomorodi: I'm Manoush Zomorodi. You're listening to Please Advise on WNYC. My guest is Dr. Dhruv Khullar. He's a physician at Weill Cornell here in the city, a health policy researcher, and a New Yorker contributor. If you have a question for him, please give us a call. We're at 212-433-WNYC. That's 212-433-9692. Let's go to Steve in Basking Ridge. Steve, I think you have a question about something else that people have been taking a lot of these days that does have some research behind it. GLP-1s?
Steve: Yes. Hi, how are you?
Manoush Zomorodi: Hi. Good.
Steve: Yes, I was wondering if GLP-1s if they can reduce cravings for sugar dependencies.
Manoush Zomorodi: That's a great question. GLP-1s, what do we know right now? Can they reduce cravings for sugar dependencies, Dhruv, and what else can they do?
Dhruv Khullar: Well, GLP-1s, unlike what we just talked about, have a tremendous amount of evidence, not just for diabetes, where they were first studied, but of course, for obesity and cardiovascular disease and stroke risk and reversal of fatty liver disease and the slowing of chronic kidney disease. These are medications that are truly for people who can tolerate them, game changers in a lot of ways. They do a number of things. They get your pancreas to secrete more insulin. They slow the progression of food through the stomach. They can even, we're learning now, alter the reward system inside the brain.
They are actually very effective at reducing people's cravings for a number of things. Sugar may be one of them, calories more generally. Something that I've been thinking a lot about is the fact that many people now report that they have reduced cravings for things like opioids or alcohol or other types of drugs. We are learning a lot about these medications. Part of that is because so many people are now taking them. By some estimates, one in every eight Americans has taken-
Manoush Zomorodi: Wow.
Dhruv Khullar: -a GLP-1, tens of millions of people. We're learning a lot more than we could learn when you just had a few hundred people in these rigorous clinical trials. Now that they've entered the real world in this massive way, we are learning all the different ways that these medications, mostly for the good, are affecting our bodies.
Manoush Zomorodi: To be clear, you've also done some reporting and seen, also in patients anhedonia, this idea that they can turn off the pleasure systems for things that maybe aren't dangerous to you. Getting outside and being a person out in the world that some people respond in depressive ways to GLP-1s. Is that right?
Dhruv Khullar: That's right. Just because they're incredibly effective medications doesn't mean they're necessarily going to work for everyone and that people won't have side effects. The side effects that we have seen, people, the most common is probably GI upset, things like nausea and constipation, vomiting even. What we're talking about here is also the potential psychoactive side effects. A minority of people, I don't want to say that this is the dominant feeling that people experience. Some people, because their reward system is being altered in certain ways, may experience a lack of interest in things that they used to derive a lot of pleasure from.
I've spoken with patients, for instance, who used to love gardening. After they started a GLP-1, they really didn't have the interest in that, or they used to love running, or they had other activities that they really derived a lot of pleasure from that they no longer felt like doing. Some people describe this as just having a blah mood all the time. Again, probably not a dominant concern, but for some small fraction of people, this is something that we should study more and keep an eye on as they start the medications.
Manoush Zomorodi: We still don't understand so much about the brain. It is fascinating. Okay, let's go to Alex in Jersey City. Alex, I think you have been observing that medspas that they're everywhere.
Alex: Yes, that's right.
Manoush Zomorodi: Tell me more. Have you been?
Alex: I have not been, but I do actually have a personal connection. My mom, she's a doctor of nursing practice, a DNP. She got her degree about 10 years ago, and she's always had a dream of being a small business owner. She actually saw this coming up, and she decided, "I'm going to try and get in on this." She's trying to get a medspa up and running off the ground right now. Now my stepdad is retiring. He's a physician. She's trying to get him-- not trying.
She has him on staff as supervisory role for a lot of what she's trying to offer as services. She's finding that it's becoming really difficult to compete at the local level with all these other medspas and that a lot of them don't have the same staffing by medical professionals. I just wondered if the guest might be able to speak to the prevalence of the medspas these days and what they offer and what should be taken as medical care from them.
Manoush Zomorodi: Excellent question. Thank you, Alex. I didn't even realize, but I have this question too, because they are everywhere. Are there doctors in these places? What are they offering? What do you have to get in terms of a license to call yourself a medspa, Dhruv?
Dhruv Khullar: There are now thousands of medspas across the country, and they generally combine some elements of a medical practice with other elements that you might think of more as a day spa type of place. There's a lot of variation across these. It's not like every medspa is the same or they offer the same types of services. Some of them have more evidence-based treatments and therapies that they're offering, and some of them have less so.
Typically, you might have some type of licensed physician or other type of clinician. There may be some nurses, there may be people who are offering other types of services, massages, and so on. The types of treatments that are offered run the gamut. Laser and light therapies, injectable Botox, body contouring, all sorts of things. I think that people should take each of the services that are offered and understand exactly what the evidence is behind those services.
Some of them may carry minimal risks, and it may not require a whole lot of evidence or data behind it in order to ensure its safety. Others are invasive procedures, or they are medications or other types of treatments like we've talked about, that may actually carry some really important biological effects. These are, as you mentioned, they're a growing part now of the medical system. I think we're going to see a lot more care being delivered through them in the coming years.
Manoush Zomorodi: How much is that because doctors are struggling to deal with health insurance, that they need a side hustle. I went to see my gynecologist for my annual checkup, but he's also offering Botox on the side and other sorts of services that insurance would never cover. He's almost like, like the caller said, trying to drum up business. Where do the economics play into going after this market?
Dhruv Khullar: Well, I think that's part of it. I think we're seeing it not just with things like medspas, but things like concierge medical practice and direct-to-consumer care. A lot of people, both patients, but also clinicians are fed up with some of the administration that they have to deal with when it comes to health insurance or some of the other overhead and burdens, documentation and otherwise, that they're having to contend with. There's a temptation to go outside of the traditional medical system. A lot of people find real value in that.
I have colleagues who have made that transition from a traditional medical practice to a concierge model, and they feel like they're able to care for patients and give them the time and attention that they always wanted to. Of course, the concern is that almost by definition, you're caring for a smaller number of people. We already have huge primary care shortages in this country. People have difficulty with accessing medical care, which can be inconvenient and expensive. On the one hand, it's very natural for people to adopt this type of concierge model. On the other hand, it can create more systemic problems because of the workforce shortages that we have.
Manoush Zomorodi: I'm Manoush Zomorodi, and you are listening to Please Advise on WNYC. We are talking about tech trends and people taking their health into their own hands. My guest is Dr. Dhruv Khullar. He's a physician at Weill Cornell here in New York City. He's also a health policy researcher and a New Yorker contributor. Let's go to KJ in Morris County, New Jersey. KJ, I understand you have been tracking some of your movement. What have you been up to?
KJ: Hi, yes. I just got a Garmin watch about two weeks ago, and it was a very intentional purposes. I did not choose to get a smartwatch for years because, if anything, I'm trying to cut back on my phone usage and not make myself constantly aware of whether I have a text or an email or a work email, et cetera. I've been jogging for exercise for about eight years and over the summer, completely randomly, after years of being a really slow runner. By that I mean like 11-minute miles. I dropped about a minute off of my mile time.
One of my teenagers suggested I get a Garmin watch and look at my heart rate. I thought, "I'm 50, and maybe that's not a bad idea." Anecdotally, I actually know two people who went into AFib and were notified by their smartwatch. I thought, "Okay, let's do this, and I'll make it so it doesn't alert me to my text messages." I did that, and I got to see what my heart rate was while I was running. I would say I'm enjoying it. I like to geek out over those stats. It also is tracking my sleep. I could take it off, but I don't.
Then I find myself losing time into, "Well, what does it mean that I had restless moments in my sleep? How does it know if I'm awake?" I think I remember being awake. It doesn't say I was awake at that juncture last night. I could end-- I took a nap on Sunday, and I told my watch I was going to take a nap. Then I told my watch I was done napping. I just wondered, in addition to being a little ridiculous, whether that behavior was healthy from a mental health perspective.
Manoush Zomorodi: The love-hate relationship we have with our devices. Dhruv, you're listening to KJ. You're nodding along. What do you think of how she's using her new Garmin watch?
Dhruv Khullar: Well, there's so much that's interesting in there, both positive and less positive, I think. Look, wearables are a technology that have taken off over the past decade, and with good reason. People want to know what's going on inside their bodies, and this is one way to access this information in a much more continuous and seamless way. I think when people adopt a wearable, they should ask themselves a couple questions. The first question is why? What exactly do I want to get out of this experience by having this thing strapped to my body? Whether it's a continuous glucose monitor, or a Garmin watch, or a sleep tracker like an Oura Ring, be intentional about why it is that you have this wearable.
The second thing to think about is, is it going to change my behavior for the better or the worse? For some people, it's really helpful to be able to track their steps per day or to get a sense of how their blood sugar changes after they eat something that they thought was healthy but it turns out was not that healthy. I think information is useful insofar as it changes your behavior in some way that improves your health. Just having the information itself without doing anything about it may not be very useful.
The third thing to be aware of is more information is not always better. There is this condition now called orthosomnia, where people are tracking their sleep and getting concerned about what they see during their sleep, and that makes it more difficult for them to fall asleep. There's this vicious circle that occurs because of that. All these sleep trackers aren't necessarily always giving you reliable information. They may be able to track whether you're asleep or not asleep, but the different stages of sleep. Are you in REM sleep? Are you in deep sleep? Those things are harder to measure.
In fact, there's a current lawsuit against one wearable device right now claiming that it is not very accurate in giving that type of information. That's all to say that I think wearables have a place, particularly if you know why you're using the wearable, particularly if it's going to be linked to some behavior change. There's also these potentially downstream consequences that are less healthy when you become obsessed over the numbers as opposed to taking a more holistic view.
Manoush Zomorodi: Listeners, we want to hear from you. Are you competing with your friends and family about who gets a better sleep score? Do you find continuous glucose monitors have changed your life? Maybe not. We're taking your questions and your advice at 212-433-9692. We definitely want to get into how you're getting your advice online as well. Are you looking to influencers, or maybe you've been talking to ChatGPT or an AI therapist? We'd love to hear your thoughts, your questions at 212-433-9692. I'm Manoush Zomorodi. We'll be back in just a minute. We're going to take a quick break. This is Please Advise on WNYC. Stay with us.
[music]
This is Please Advise, the pop-up advice show from WNYC. I'm Manoush Zomorodi, host of NPR's TED Radio Hour and author of the book Body Electric. My guest today is Dr. Dhruv Khullar. He is a physician at Weill Cornell Medicine, a health policy researcher, and a New Yorker contributor who writes about healthcare and policies and politics, all those things. We are here to take your questions to get your advice. Are you protein pounding? Are you peptide stacking? Are you Oura wearing? Or are you against all of these things? What are your questions? What are your thoughts? Let us know. We're taking it all at 212-433-WNYC. That is 212-433-9692. We've got some more calls coming in.
First, Dhruv, I want to ask you there was a study published this year based on research from Harvard Medical School and Beth Israel, where they found that AI outperformed doctors in the ability to accurately diagnose patients. As a physician yourself, where are you seeing that AI is helpful in treating patients, and where are you seeing that it's just not what you want to be doing right now?
I think a lot of us find that AI is creeping into our doctor visits. I mentioned I went to go get a mammogram, and they said, "Do you want to pay the extra $50 to have AI screen your boobs?" I was like, "Sure, why not?" Nobody really explained to me how accurate it was and whether it was a good idea. It was just offered. What are your thoughts on this right now?
Dhruv Khullar: Well, AI is one of the most rapidly adopted technologies in healthcare, which is not a industry that usually adopts new technologies rapidly. That is saying something about at least the perceived benefits of this technology. There are a lot of studies that now that show that AI can perform as well or better than doctors on certain types of cases. I think this is the key point in that we know that if you curate a case just right and you give that information to the AI in the way that we might present a case to a colleague, it is extremely good at coming up with the right diagnosis. Superhuman in that case.
In the real world, a lot of the job of a doctor is actually doing the work of curation. It's figuring out how to ask the right questions, where to probe, which statements might be relevant and which might be irrelevant to actually cracking the diagnosis. I'm much more skeptical that AI, at least right now, is able to perform those things in the real world as well as doctors can. One of the things that I think people should take home from this is that the way in which you engage the AI can make all the difference. I'll give you one example from last year. A friend of mine who's a doctor, he started having explosive diarrhea.
Manoush Zomorodi: Oh, boy.
Dhruv Khullar: He ended up getting some really bad GI upset. He had a waxing and waning illness over the course of the month, ended up losing like eight to 10 pounds. He's a primary care doctor, but he put his symptoms into ChatGPT as he would a gastroenterologist. Now it comes up with the right diagnosis. It says it's Cyclospora, which people have probably heard about this year, but was less prevalent last year. He goes to the doctor, the doctor does the test, it is Cyclospora. He gets Bactrim, an antibiotic, and gets better. Great story.
I did the same thing. I said, let me not describe it in the way that a doctor might. Let me give it more a general set of symptoms that I thought some of my patients come to me with, "I'm not feeling great, I've had some diarrhea," but maybe I leave out the exact character of my stools. I leave out exactly how much weight I've lost. I would leave out the fact that my symptoms got better and got worse at a certain frequency over the past month. It tells me you have viral gastroenteritis, you don't need to see a doctor, and that I'm here for you.
Manoush Zomorodi: [laughs] I'm here for you.
Dhruv Khullar: I got a totally different answer from the AI when I prompted it in that way. I think that's one of the key challenges, is that people should be aware of that potential limitation. There are better and worse ways to use AI chatbots. A good way is to maybe prepare for a doctor's visit, get some questions that you want to be asking your doctor, maybe unpack that visit with the AI afterwards.
If you're having symptoms and you can't see a doctor, ask it about red flag symptoms. What are the things that I really should be on the lookout for if I have, I should make sure to get medical attention? There are ways to use these bots, of course, but I think just plopping in your symptoms and asking it for the diagnosis may not be the right way at this time.
Manoush Zomorodi: Does it come down to saying, like, with access to all this information, we need to be more cautious about how we present our questions, how we follow up on the information? There's so much there that it could come back to you with that you need to take it all with a grain of salt.
Dhruv Khullar: I think so. I think part of what you're hinting at is the back-and-forth, I think, is important and going a level deeper than the initial answer that you get. That probably applies not just for medicine, but for all sorts of things that people are using AI for. The first pass might be helpful as a gloss, but you really often want to get deeper, and you want to understand the blind spots that the model has or that you have. I think engaging that conversation, then cross-checking it with non-AI sources, all those things are going to be really important.
Manoush Zomorodi: We're talking to Dr. Dhruv Khullar. He is a physician at Cornell Weill. He also writes for the New Yorker magazine. We'd love to take your questions, or if have found that something works great for you when it comes to using tech for your health, give us a call. 212-433-WNYC. That is 212-433-9692. We're going to now talk to Jackie in Orange, New Jersey. Jackie, you had a question about what you're eating, I think.
Jackie: It's more about the proteins that you mentioned. It's on every shelf. It's everywhere. I don't eat red meat. I'm 70 years old. I'm pretty active, and I don't know if I'm getting enough protein in my diet. Should I be considering buying more products that have more protein in it since I don't eat red meat, I don't eat a lot of chicken, once in a while fish?
Manoush Zomorodi: I'm with you because everywhere I go, it's like it has extra protein. I was walking to the studio this morning, and a coffee place had a sign out front that said you could get extra foam on your latte, protein foam, which would give you an extra 17 grams of protein. I was like, "Why? Do I need that?" Dhruv, what is the deal with protein?
Dhruv Khullar: Well, protein is everywhere these days, as you mentioned. Candy has protein in it now, and coffee has protein in it now. To say nothing of the foods that have always had protein. A couple things to know. One thing is that there are recommendations for how much protein people should get. The traditional recommendation has been something like 0.8 grams per kilogram. Let's say you're around a 70-kilogram person. That's around 150 pounds. That's about 55 grams a day.
Now, I think a lot of people in the scientific community think that maybe that's too low. Maybe we should be aiming a little bit higher than that, particularly if you're an athlete, you're an active person. Particularly if you're older and you want to prevent muscle loss, maybe we should bump that up to something like 1.2, 1.5 grams per kilogram. Again, if you're 70 kilograms, maybe that's around 100, 120 grams of protein.
The thing about protein is that at least according to the traditional guidelines, something like 90% to 95% of Americans already get enough protein. Most people are not protein deficient in our country. Now contrast that with fiber. Something like 5% of Americans get enough fiber. If you want to focus on getting something else new and important in your diet that you're not currently doing, fiber is probably more important than protein.
Let's say you do want to get enough protein. I think there's certainly no reason not to aim for a little bit higher than the traditional guidance has been. Many people don't eat red meat. There are other ways to get protein. If you're vegetarian or if you're plant-based, beans tend to have a lot of protein. Lentils, nuts, seeds, soy. If you're able to have eggs and dairy, milk, fish, poultry. There are ways to get protein that don't require even getting a protein bar or getting protein foam on your coffee.
Manoush Zomorodi: [laughs] Can I just run my breakfast past you? All right. I had Greek yogurt with berries and flaxseed and some granola.
Dhruv Khullar: That's fantastic. That's something that I also have quite often. I think Greek yogurt, in and of itself, is high in protein, low in fat. It's a wonderful thing that everyone who has dairy should consider adding to their diet.
Manoush Zomorodi: How much is the whole protein-maxing fad? Because there are so many people on GLP-1s, and one of the side effects of GLP-1s is that you have trouble building muscle. In fact, some people lose muscle.
Dhruv Khullar: I think that could be part of it. I think there's an independent stream here which is people-- it's a good branding exercise because people feel like if it says protein on the label, then it's going to be good for you. Even if they're also sneaking in a bunch of fat and sugar and other things like protein bars. Many of them are good, but some of them actually sneak in a lot of these other ingredients that you may not want.
It reminds me of things like vitamin water. What's wrong with vitamin water? Well, some vitamin waters have half a can of coke's worth of sugar in them. These types of things, if you slap a sciency sounding positive label on them with a macronutrient, it doesn't necessarily mean that it's healthy, but you can get a lot to slide behind them.
Manoush Zomorodi: Let's take another call. Let's go to Gabriel in Astoria. Hi, Gabriel.
Gabriel: Hi.
Manoush Zomorodi: What's going on?
Gabriel: Today I'm celebrating three years of total foot reconstruction surgery that I had done at the hospital for special surgery, and also had genetic testing done at Weill Cornell. I'm just calling, one, as just to say thank you to all the doctors that helped me through that. I'm walking, and it's amazing. I just want to say that I'm a 40-year-old male, and sometimes it's not easy for guys to, one, admit that they need to go see a doctor and actually get there. That's one part.
The second part is with all the AI and research, when I was diagnosed a few years ago, I just went down the rabbit hole with my wife. It's a lot, and it's intense, and it's scary. At the end of the day, you are your own best advocate, and you need to go see multiple doctors, you need multiple opinions, and you need to take your time, and eventually you'll get there. I think, like what you and your guest are saying, that the information is good, but it can be overload.
At the end of the day, you are your own best advocate. I would attach a question to that as well with my public thank you. Also, do you know, I do have a nerve condition, and I was diagnosed, and then I was undiagnosed, just to the point of we don't know a lot what's going on. I'm still trying to figure out what's going on. The surgery did help. I do suffer from nerve stuff. My question would be, what can I be doing to help the nerve damage that I have in my diet or in other aspects as well? Thank you.
Manoush Zomorodi: I'll pass it to Dhruv. I just want to say, Gabriel, congratulations. Three years, that's awesome. Dhruv, what do you think? Nerves or pain?
Dhruv Khullar: Well, I'm so glad, Gabriel, that you had a good experience, and wishing you a continued recovery. One thing that your comment raises for me is how AI can potentially be a useful adjunct and not necessarily a replacement, but an adjunct to what happens traditionally in the medical system. Particularly for conditions that are less frequent or more difficult to immediately diagnose. You mentioned that you were diagnosed with one thing and then the diagnosis was reversed potentially.
I think generally medicine does a very good job with conditions that are straightforward, that we know what to do. There's a algorithm to go down. If you have pneumonia, if you have heart failure, if you have particular types of cancer that are well characterized. When people have other types of conditions that are not as clear. You might have a GI problem that is really hard to diagnose, and they don't see anything on the CT scan or the endoscopy. Or you might just have really significant fatigue and joint pain that has not been diagnosed for many years.
Those are the types of areas that medicine, I think, traditionally struggles with. Many people bounce around from one doctor to another or one system to another aren't able to get the answers that they want. Sometimes AI can actually be helpful, at least anecdotally, I've heard for people to bring all that information together to maybe go down avenues that hadn't been unturned just yet. I hope that going forward, people who are in that type of situation are able to get some assistance, both from the clinicians that they have, but also clinicians who are armed with artificial intelligence.
Manoush Zomorodi: You're listening to me, Manoush Zomorodi, host of NPR's Ted Radio Hour, and this is Please Advise on WNYC. My guest today is Dr. Dhruv Khullar. We want to take more of your questions, your thoughts, your experiences about your health, technology, all the information that is out there, it is confusing, it is frustrating, sometimes it is enlightening. You can call us at 212-433-9692. That's 212-433-WNYC. We'll be right back after a quick break.
[music]
You're listening to Please Advise, the pop-up advice show from WNYC. I'm Manoush Zomorodi, host of NPR's TED Radio Hour and author of the book Body Electric. My guest is Dr. Dhruv Khullar. He's a physician at Weill Cornell Medicine, a health policy researcher, and a New York magazine writer. Dhruv, much of what we've discussed this hour, it seems like we got to talk about who has access and who doesn't, right? The mammogram examples, getting an AI screening, I had to pay extra. My insurance wouldn't cover it.
At the grocery store, all that added protein. You're paying a premium for these fancy products, buying an Oura Ring or wearing a special watch that costs money. I guess I have two questions. One is, are we seeing the further bifurcation of who has access to health information because of technology? Two, are we paying with our personal data and information?
Dhruv Khullar: Well, we've had a bifurcated healthcare system for a very long time. The hope is that some of these technologies democratize access to medical information and care delivery in a way that we've never been able to do before. I'm actually an optimist that this technology, if applied in the right way, could move us in a direction that is more equitable. Now, there are all sorts of concerns that arise with greater use of technologies like AI in healthcare. You mentioned data privacy, data security.
I think AI has the potential to improve diagnosis, but also to encourage misdiagnosis and potentially create also an avalanche of patients coming to the hospital or not coming to the hospital, depending on what the AI is telling them. There's a lot of challenges that need to be worked through. I think as a general matter, we have an opportunity, and we have to shape this technology and the incorporation of the technology in a way that allows more people to access the type of care that they deserve.
Manoush Zomorodi: I want to talk to Peter, actually, who-- we've got another foot. Somebody who had another foot issue, but a different sort of experience. Peter in Forest Hills. Peter, are you there?
Peter: I am.
Manoush Zomorodi: Tell us what's going on with you.
Peter: Hello.
Manoush Zomorodi: Yes, hi.
Peter: Hi. I was saying a couple of months ago-- I'm a 61-year-old guy out walking my dog, and I twisted my foot, and I had a fracture of my metatarsal, first bone I've ever broken in my body. I went to urgent care and then referred to a podiatrist. After a few weeks, I took the X-rays and the doctor stuff, and then I fed it into AI, and it was great because you could say every day, what should I expect to feel this day? What's a normal thing? My toes are blue. Is that normal? When can I expect to start walking? When can I expect to start walking the dog?
It was very reassuring, to the point of when I realized every time I came back-- we go maybe every two or three weeks at first and get an X-ray, and then they'd see me a few days later. My foot continued to mend, thankfully. Then I realized I don't need to keep going back there. It's feeling better, it's healing, I'm on track. I have a high deductible health plan. I haven't met the deductible. I don't need to pay $200 for an X-ray and $200 or more than that. Then to see someone then. Actually, all the time, I never even saw. They never even refer--
Actually, I never actually saw the actual podiatrist whose name is on the appointment. I always saw a resident or a nurse, and all they did was rewrap my foot and look at it and just say, "You're doing great." I was like, "Okay." It would be one thing if my foot wasn't feeling better, but it was feeling better every day. I knew things were going well. AI was great and very reassuring in the middle of the night when I'd feel a twinge and go, "Oh my gosh, did I re-injure it," or what have you.
Manoush Zomorodi: Thank you, Peter. Dhruv, I think what we hear in Peter's voice is the frustration that he did not get the personal care from actual humans. That he wanted the reassurance that something was happening in his body and he was on the mend that AI can give people. When you hear Peter's story, does that reassure you, worry you, both?
Dhruv Khullar: Well, I think what Peter's story suggests is something that a lot of people have encountered in that the medical system can be inaccessible, it can be inconvenient, it can be unaffordable. I think that is one reason that there is still a good degree of optimism about AI in health care. There's been recent polling that suggests that Americans are generally pessimistic about AI on every dimension, like national security, what it's going to do to education, what it's going to do to our democracy. They're still generally positive about health care and scientific research.
I think part of that is because people see a lot of room for improvement, and they think that AI will be able to meet some of the needs that haven't traditionally been met. I think in Peter's case, it's great that things worked out well and moved in the right direction. You could envision the AI. In fact, a lot of people have also had the experience where it's not giving them the right information. We know that more and more people are using chatbots, but a significant percentage of them end up getting information that later turns out to be wrong. There's also that risk.
I think one of the things that it raises for me is this Idea that in the future healthcare might have different tracks that it operates on. There may be a track that is more transactional care where you have a foot injury, you need an X-ray. You have a urinary tract infection, you just need to pick up your ciprofloxacin from the pharmacy. Then there may be a separate track that is much more complex and requires trade-offs and values. You have a serious cancer diagnosis or advanced heart failure or dementia care. These types of things where we may still really want a human first, maybe AI-assisted set of approaches for medical care. I think we will see this bifurcation of care being delivered on one hand mostly by machines and on the other hand mostly by people.
Manoush Zomorodi: I think we have time for one more call. I want to go to Michael in Sunnyside. Michael, you have been waiting patiently, listening to this discussion. What's going through your mind right now?
Michael: Sure. They're related in the way that ChatGPT has been tested on a lot of people out through the society, but not really regulated or tested in any way before that. We're seeing how GLP-1s have made a good example that we haven't really seen positive examples from ChatGPT going through the general population. If you think about the self-harm issues and other hallucinations.
Manoush Zomorodi: I think you make a very good point. Thank you, Michael. Michael is saying, I think Dhruv and I think about this often as well, that we are in a period of experimentation in real time. We don't see medical research necessarily keeping up with the real-time experiences that we are having, whether we are tracking ourselves, or injecting ourselves, or looking up information. How do you think about this? Are we in a new place when it-- you talked about a two-track or a different track healthcare system, What about research and advice? Where are we with all this? As a physician, how do you deal with that?
Dhruv Khullar: Well, this is what's so critical is that there are all sorts of claims being made about AI and how it will do X, Y, and Z in healthcare, but we don't have a ton of evidence about its deployment in the real world. There's some of it, but for the most part the claims and the usage are outstripping the evidence base that we have. I think one of the critical things that we need to do in the coming years is to rigorously test some of these models in healthcare systems after they've been deployed in the real world, in the way that we do for other things like drugs and devices.
We need to build up an evidence base of what these algorithms are doing when they reach people in the real world. Some of the approach that the FDA traditionally takes to this type of regulation doesn't really fit very well for AI.
Manoush Zomorodi: No.
Dhruv Khullar: Just having a regulatory apparatus where you have Phase 1, 2, 3 clinical trials and then you approve a drug based on its safety and efficacy. Well, the drug doesn't make decisions, and it doesn't evolve after you put it into practice in the way that an algorithm does. Some of this will require not just pre-market approval, but post-deployment surveillance. It'll be up to health systems and maybe other and new agencies to determine what are the effects both in terms of safety, but also effectiveness after an algorithm is deployed into clinical practice. As a general matter, I think my view is that we're so early in this process, it's like we're talking in 1995 with the advent of the Internet. We know it's going to be big-
Manoush Zomorodi: Wow. Yes.
Dhruv Khullar: -but we don't know all the things that are going to happen. This podcast wouldn't be possible without the Internet. Uber's not possible without it. Social media, all these things that are really dominant parts of our society today were not predictable in 1995. That's where we are right now with artificial intelligence.
Manoush Zomorodi: We have just one minute left, Dhruv. If you were to say to someone listening, if there's just a few words I can leave you with, when you think about your health every morning when you wake up, what would you say?
Dhruv Khullar: Well, I think there's so much excitement about health, and there's so many claims that people make on social media and elsewhere. A lot of it has to do with just trying to sell you something and get your attention. If I could say anything, it's basically, we know the basics for how people can live long and healthy lives, and we just need to create the conditions under which people can actually engage in those behaviors. There's just a handful of things. We know them. Exercise, moving your body every day, eating a healthful diet, by which I mean mostly trying to avoid things like ultra-processed foods. Those are ubiquitous in our society today.
The more that you can cook at home and not eat, what is going to be out in the world when you go out there, the better. The first is exercise. The second is cooking for yourself or avoiding ultra-processed foods. The third is sleep. People poo-poo sleep, and they think it's a waste of time. If you can get seven or eight hours of sleep, we know from studies that's better than five or six hours of sleep. The fourth big area I would say is maintaining social connection, investing in relationships, recognizing how important those are not just to our emotional well-being, but actually for our physical well-being as well.
Manoush Zomorodi: I love this because tech is amazing, but so is our biology. My guest has been Dr. Dhruv Khullar. He's a New Yorker contributor. If you haven't read his articles, you absolutely should. They're so fun and fascinating. He's also a doctor at Weill Cornell. Thank you to you and your staff for caring for us in the city. Dhruv, thank you so much for joining us on Please Advise.
Dhruv Khullar: This is a real pleasure. Thanks for having me.
Manoush Zomorodi: Okay. Coming up tomorrow, Paul Ford. I have a weird crush on him. He's a programmer, a technologist. He's been called the poet of the Internet. We are going to talk about work and AI. Yes, it's going to get spicy. Bring us your questions. Do you think this is a disaster for the economy? Maybe you're excited. You've been using AI at work, and you think maybe it's not as bad as you thought. We want to take your questions, we want to take your comments. You can call in tomorrow or submit questions online in advance at wnyc.org/advise. I'm Manoush Zomorodi. This was Please Advise, and I'll see you tomorrow. Thanks for listening.
[music]
Copyright © 2026 New York Public Radio. All rights reserved. Visit our website terms of use at www.wnyc.org for further information.
New York Public Radio transcripts are created on a rush deadline, often by contractors. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of New York Public Radio’s programming is the audio record.