Melissa Davis: Can a Radiologist Trust AI?

Harlan Krumholz: Welcome to Health & Veritas. I’m Harlan Krumholz.

Howard Forman: I’m Howie Forman. We’re physicians and professors at Yale University. We’re trying to get closer to the truth about health and healthcare. This week, we’ll be speaking with Dr. Melissa Davis. But first, we like to check in on hot topics in health and healthcare. And Harlan, I know there’s a topic that got you really interested, and I’m looking forward to hearing your thoughts on it.

Harlan Krumholz: Well, there are two quick topics I wanted to hit on, Howie, and I’m curious what you think about them. The first is this reported thing with RFK Jr., famously anti-vaccine individual. He was quoted as saying that the SARS-CoV-2 virus, the COVID-19 virus, was designed to spare Chinese and Ashkenazi Jews. And this was reported in The New York Post. I didn’t want to get into any sort of discussion particularly about this. Of course, it has a lot of issues that we could be just talking about. But the one thing that was on my mind, Howie, was how do I know when I read something like that that it’s true? Because it’s so outrageous. Literally, you can’t believe that someone like RFK Jr. would be capable in a group of saying something like that. So someone says, “I’ve got a recording of it.”

But as you know now, in an AI world, anybody can simulate a voice. In fact, there’s a movie about Wilt Chamberlain that they’ve said in the movie, “What we did was we did an AI replication of his voice, so when we say things that he was reported to have said, you will hear it as if it’s coming from Wilt.” So Wilt Chamberlain, a famous basketball player. We’re capable of doing this now. I was thinking, Howie, when I first read it, I didn’t even want to retweet about it because I thought, “Maybe I’m being trapped. This has to be crazy.” But then, New York Times comes out and affirms it. The Wall Street Journal comes out and affirms it. So we could talk for a long time about that quote.

But I think we’re in a world, Howie, where I’m realizing for myself that I’m nervous about what’s information, what’s misinformation? And then of course, there are people who will say crazy things, but how do I know it’s true? What am I looking for? How are you thinking about this?

Howard Forman: But I do think this emphasizes why the journalism—what do they call it, the fourth estate—is so critical because if you have credible journalists in the room, if you have people reporting real time, no longer do you have to rely on just recordings or videos. You actually have reporters who have journalistic integrity reporting. Honestly, even that has been undermined a lot by the AI revolution because a lot of companies are laying off journalists and replacing them with AI-generated content. And we’re going to have to, as a society, make a decision, much as our guest this afternoon will talk about. You got to commit money to something if you want it to really work.

Harlan Krumholz: We’re going to need these guardrails, especially as we come up to the election. In all aspects of public health as well, you’ll have to wonder, did somebody really say that or not? Is that really true? It’s hard.

Let me get to the second one too, because I know this is something that interests you, and I’m being greedy this week by taking two topics. There was an article that came out from our friend, Zeke Emanuel, who we’ve had on this program, and Robert Berenson, really, a deep expert in health policy and healthcare, about the Medicare physician fee schedule. And I just thought it was a really nice piece about illuminating some of these issues that we’re talking about with regard to incentivizing the right behaviors within medicine. And they really, I think, bring into bright relief this idea that the current system is just paying for volume and it even rewards inefficient, poor-quality care that may not even be benefiting health.

It’s sort of like a restaurant that they’re just being rewarded for the volume of food they put out, without regard to the quality of it. Now, for food, people can actually tell, do I like this or not? But for healthcare, most patients are in a very poor position to be able to judge this. And then moreover, they’re bringing in this issue that there is this thing within Medicare where doctors get together and say, “How long does it take to do something, and how intensive is it?” So doctors are judging themselves in this respect. And they bring up this example of cardiologists estimating that it takes six minutes to interpret an electrocardiogram. When all of us in cardiology know that the vast majority of electrocardiograms are read by a machine. We’re just signing on top of the read, and maybe it takes us 10 or 15 seconds to do.

Now, there’s an occasional one that’ll take a long time, but the average of six minutes is definitely not true. But that’s coming from a board of doctors who’s trying to assign a certain amount to them. All my cardiology colleagues are going to be mad at me for exposing this. It’s everywhere. You have it in radiology. We have it everywhere.

Howard Forman: Radiology, it’s the same thing, absolutely. Absolutely. Now—and look, I remember about 15 years ago, you and I went to the White House and we sat around the table and talking about healthcare reform with people who are really leaders in the field about what we can do in terms of Obamacare. I guess it was 2009, so 14 years ago. And everybody talked about this movement of volume to value, volume to value. It was the talking point of the decade, and several years later when I realized in our own practice where I play a role in the financial management of our practice, and realized we were doing almost nothing based on value. It was still all on volume.

So I started asking people, “Are we moving to value? Because I’m not seeing it.” Everybody would wave their hands and say, “Of course. Of course.” But what they really meant is one percent or half a percent of their revenue was truly at risk of value, everything else was still about volume. And here we are now, 14 years since you and I sat a in the Eisenhower Building or the White House Complex for that meeting, and I’m not sure we’ve made any meaningful movement. We’ve had some areas where things have been proven to work, but overall, we’re all about volume. And until we change the incentives, we’re not going to change that pattern.

Harlan Krumholz: It’s going to be a setup for discussion for all the new innovations that are going to come down the pike. If it’s just volume without regard to selection people to benefit, it’s going to be an issue. Okay, so let’s get onto our segment, Melissa Davis today. Take it away, Howie.

Howard Forman: Dr. Melissa Davis is the vice chair for Medical Informatics and associate professor in the Department of Radiology and Biomedical Imaging at the Yale University School of Medicine, and the medical director of System Radiology Imaging Informatics at Yale New Haven Health. Her research focuses on patient service, process optimization in ambulatory and inpatient environments, to improve medical outcomes and innovative ideas to target social determinants of health. Before taking up her current positions, Dr. Davis was an assistant professor of radiology and physician lead of the Yale Clinical Optimization Services. She was chief of emergency radiology and the clinical lead for the Center for Outcomes Research and Evaluation at the Yale Medical School, which Harlan Krumholz, our partner in crime here, directs.

She also briefly worked for Emory, and supported the company Nines Radiology as an advisor to their clinical operations. Dr. Davis graduated from Wellesley as a psychology and chemistry major and later obtained an MD from the Medical University of South Carolina. She was a resident in diagnostic radiology as well as a fellow in neuroradiology at UNC Chapel Hill, and joined Yale in 2017 to pursue and complete her MBA, which is when I had the first great privilege to meet her. So first of all, I want to welcome you to the Health & Veritas podcast, Melissa Davis, my friend, colleague, and vice chair.

And I want to start off, because about a year and a half ago, Harlan and I had a conversation about assistive technologies in radiology, specifically how AI is influencing practice. This is what you do; you manage the implementation and practice of AI within radiology. I want you to just give our listeners a little taste of what are the AI applications that are currently being used and what do you see coming down the pike in the near future?

Melissa Davis: I want to start by saying thank you so much for allowing me to spend some time with you all on this podcast today. I’ve been listening to you all since the inception, and you have some amazing people on here, so I’m glad to be a part of that.

Harlan Krumholz: Who do you think is better, me or Howie?

Melissa Davis: I can’t answer that, Harlan. That’s a trick question.

Harlan Krumholz: You can just whisper into my ear who you think is best, that’s okay.

Melissa Davis: Sure. But you’re right, assistive technologies within medicine as well as radiology have expanded significantly, even over the past year and a half since you all did that initial podcast on these types of technologies. There are a variety of them. When you talk about radiology specifically, a lot of times we talk about augmentation and segmentation type of tools, and those are tools that allow the AI to actually look at an image and determine something off of there. So what we use at Yale are several of these types of tools. We can do things like detect head bleeds. We can detect emboli, which are clots in your lungs, based off of that. We could also look for things like a pneumothorax, or error where it’s not supposed to be in your lungs. That’s just a small sampling of the types of tools that we’ve started to use within our space, within the radiology department.

But even beyond that, there are other tools out there that are not like that, that we typically see in the media. There are tools that actually help us with communication. For instance, there’s tools that if I dictate a report and I say, “There is a right-sided pneumothorax on this chest x-ray,” we can get an auto-generated impression based off of that. And that helps the radiologist speed up their work, but it also helps decrease that cognitive load of having to rehash something that they’ve already said. So there’s these nondiagnostic tools that are also coming into this space that we’re starting to look more and more at as well.

Harlan Krumholz: I’m so happy to be on the program with two extraordinary radiologists and an expert in AI. And actually, Howie is really growing to be an expert in AI as well.

Howard Forman: I’m just growing.

Harlan Krumholz: But not like you, Melissa. Not like you at all.

Melissa Davis: I was like, you guys know more about this than me!

Harlan Krumholz: No, come on. And so for our listeners, they may know that Vinod Khosla, a very well-known investor, said, I don’t know, maybe five or 10 years ago, that he thought AI was going to replace doctors. And the place where people have thought about this maybe being most likely applied would be in a place like radiology or pathology, where there’s a lot of pattern recognition. And not counting the interventional radiologists, who we’ve had the pleasure of talking with Dr. Shapiro recently and talking about these issues, but for the pattern recognition. And I just wonder if you could tell me how do you think about this?

Because in some ways I think that it’s at least possible that we could have each radiologist be able to do more with some sort of performance augmentation with AI. But in many cases, in places in the world where there just aren’t enough specialists available that we could have access to these kind of expert systems that could help people, on the battlefield even, for example, where it may be hard for radiologists to be present. Now you could have command centers, but also with the AI. Do you see this happening? I mean, what do you think it’s going to look like in 20 or 30 years as this technology continues to advance?

Melissa Davis: Yeah, I think that there is a short-term look, and there is a long-term look at this. And I think that these types of technologies are so transformative, not only within healthcare but within society, that we’re going to have to rethink the way that many of our jobs look. And I don’t think that it’s just radiology or pathology. I think we’re going to have to re-look at what does general medicine look like as well and other types of specialties going forward. I think that there is a space to help increase access to radiology by leveraging these types of technologies. And we’ve actually started to see some of that across the world, where maybe we implement a technology that prioritizes acuteness, essentially.

And so for a radiologist who is sitting in a country where maybe they’re the only radiologists—I’ve been to a couple of countries where there’s one or two radiologists for the entire country—we need to prioritize what actually needs to be seen by that radiologist, as opposed to what doesn’t need to be seen as quickly. So a mass that needs to be seen, or some sort of acute pathology that needs to be seen and needs to be transferred over more quickly.

Harlan Krumholz: And just one quick follow-up to this, just to this point in time just talking about today, have you seen anything in your field with AI that has blown your mind? Have you seen anything that you said like, “Whoa, that is incredible”?

Melissa Davis: The first time that I saw AI detect any pathology, I was like, “Whoa, that’s incredible,” because that is not something that I had seen ever before. It’s becoming more commonplace now because it’s been a couple of years since we’ve seen it.

Harlan Krumholz: And what did you see?

Melissa Davis: One of the biggest... so a head bleed, that’s the first platform that we had at Yale. And the story behind this is actually what gave the “whoa” is that we had implemented it; we were running it on all of our patients who got head CT scans. We had a patient who came in at 5 p.m. to get an outpatient head scan for pain. Those are images that we would never read after five; our outpatient radiologists go home at that point. It was flagged, our radiologist saw it, made sure the patient didn’t leave, got that patient treated ASAP. That patient could have gone home and had a very bad outcome. So that, even though I saw the technology and I was like, “Whoa,” when I had that story, I was like, “That is actually amazing and what we’re trying to do.”

Howard Forman: And to Melissa’s point, just want to add, there’s two elements to it. One is that it’s an assistance in terms of an extra set of eyes, so to speak. But it also, as you mentioned, prioritizes cases so that when you’re in the emergency room and you’re 30 minutes behind, in cases, which frequently happens during high-volume states, if you have a pulmonary embolus patient moved to the top of the list instantly because the AI technology decided that it’s positive, that can have an impact. As you know, Harlan, in so much of your work is informed the timeliness of treatment in patients, the fact that we can advance diagnosis by even 30 minutes can be important.

Harlan Krumholz: So one of the questions I’ve been asking people is, how do you know what to trust? Because vendors are coming out and saying, “Hey, we’ve got this. It’s this amazing new thing.” You’ve observed something great, but how do you know you can start to rely on it?

Melissa Davis: Yeah, and that’s something I actually think about a lot. It’s not even trust as you bring it in, but trusting it 5 years later or 10 years later when we know that these algorithms degrade in some capacity. Is it still going to work going forward? So what we do is we do a lot of validation before we onboard a platform to make sure that there’s some clinical trust within it. This is also why I don’t think that radiologists are going to be replaced soon. I think radiologists who leverage this technology will be better radiologists. So you’ll want that, but it’s really going to come down to that clinical determination if you trust it or not. Because there are several times throughout the day where it’ll flag something as a false positive.

So meaning, that it says that there’s something there. I look at it and it’s not. It’s something else. It’s a calcification rather than a bleed, or something on the periphery that it thinks it caught, but it’s not there. So these technologies are not perfect, and a lot of us are aware of that, which is why we’re very cautious about the pre-implementation piece of that. But as we go forward, we’re going to have to think about how well do they remain at their job going forward? Do they get worse at their job as our populations change based on the training dataset that was there before, or as the algorithm itself changes going forward?

Howard Forman: I want to come back to AI if we have time at the end, but I also don’t want to forget to mention that you are a graduate of our EMBA program. And this is a podcast that we’ve been doing from the School of Management from the beginning. And I think the EMBA played a significant role in your career. But I’d love to hear what specifically. What are the things that happened between the time I met you when you arrived on campus, and I think, if I’m not mistaken, July of 2015?

Melissa Davis: 2015, yeah.

Howard Forman: ’15. Yeah, 2015, when you arrived on campus, and when you graduated and you were working in radiology the whole time, but a lot of things changed for you during that time. What courses, what content actually were most important to you, and how did they inform your current vision of healthcare delivery?

Melissa Davis: Yeah. Well, I’m just going to take a step back because I know you asked about specific courses and content. But Howie, you were a major impact on me—

Howard Forman: I appreciate—

Melissa Davis: ...coming in. If I say the number one person who has impact on my career, it’s Howie Forman.

Howard Forman: I appreciate that, but Harlan already—

Harlan Krumholz: Super Howie.

Howard Forman: Harlan already praises me enough, but I appreciate that very much. But tell us more.

Harlan Krumholz: He is extraordinary. He’s extraordinary.

Melissa Davis: Yeah. Well, I think it leads into my next point, which is that mentorship and sponsorship are very important. And especially for somebody like me, who was new at Yale. I didn’t know anybody when I landed. I was really just trying to figure out why this campus was so much larger than the one I had been on before and how I could get from the business school to the hospital to my home. Always getting lost trying to get home the first couple of months. But having somebody who was willing to mentor me and then later sponsor me into roles actually accelerated the path that I went on. So when I started the MBA program, I really thought that I wanted to do some sort of organizational management type of change process, leadership process, things like that. But I didn’t really have an idea of where it would be.

Going through the MBA program really expanded my view of healthcare. When you’re in medical school, you’re just learning anatomy, you’re learning the basics. You go into residency, and it’s very much tunnel vision on understanding your specialty. I didn’t understand how we get paid. I didn’t understand how hospitals worked. I didn’t understand quality measurement, and that’s Harlan’s space. I joined CORE while I was a fellow as well. I didn’t understand startup culture and how a lot of these technologies are actually being built outside of academia and in these spaces like Silicon Valley, and how that’s going to impact healthcare. And I didn’t understand finance. And so taking courses that aligned with those things really opened my mind when it comes to healthcare in general. And so I feel like I got a bigger overview, and I could see the bigger picture going forward.

Harlan Krumholz: I just wonder if you think all medical students should get more exposure to these things?

Melissa Davis: Absolutely. I think that medical students are at a disadvantage with that tunnel vision that they’re given, and I wish that we could incorporate that into their education a bit more, for sure.

Harlan Krumholz: One thing I wanted to ask you while we have you here is I know that you have a deep commitment to and interest in health equity and how we’re going to address this within the healthcare system. We’ve been doing a lot of research and really basically showing that over the last decades we’re not making any progress. That for all the rhetoric and for some of the investments, it’s not returning efforts. Now, I don’t know the counterfactual. I don’t know if things would’ve even gotten worse than they are, but they’re certainly not getting better. I just wonder what you’re thinking are some of the central strategies we should be thinking about that are new because the old ways of doing things just aren’t standing up to what they need to do. So what are you thinking now that we should be doing?

Melissa Davis: So yes, health equity is a passion of mine, and it’s something that, over my career, I’ve gotten more and more involved in. When I think about health equity, I think you’re completely right. There are things that we have tried to do and we’ve not really moved the needle that much. But I also think about the fact that we are actively dismantling some of that, which probably kept us at par as well when we think about it from a societal level. So for instance, the Supreme Court ruling around affirmative action. We know that if Black doctors are present in Black communities, those patients do better. But we know that with this type of a ruling, it’s going to be more difficult for Black doctors to be created, essentially. It’s going to be more difficult for them to get into college, it’s going to be more difficult for them to get into medical school, it’s going to be more difficult for them to get involved in specialties and subspecialties, especially.

Radiology does a terrible job at diversity. And so we’re not going to see as many doctors that look like me who are radiologists going forward because of actions like this. So at this point, I think that it’s imperative that we look at it from a policy angle, and say that there are things that we need to do and we need to stop some of these changes so that we don’t go backwards, to be honest, so that we can continue at the field that we’re in. And then we need to figure out how we can actually push people forward who we know will actually care about this. And that means that you need to tackle students when they are in high school, when they’re in middle school. We need to focus there, get them interested in it, get them aware of it, so that they can also always move forward. And I don’t know if that’s so novel, but it’s things that we know will work but we just don’t do well.

Howard Forman: I think I asked a similar question of Elizabeth Arleo when she was on a few months ago, but radiology does have a relatively poor record, putting it lightly, of diversity in terms of ethnicity and race and diversity in terms of gender. I mean, it’s grossly underrepresented among women and among people of color. And this has gone on for a very long time. And I may be wrong, so I’m just stating this and you can correct me, but I feel like you were almost dragged into being involved in this because there is such a shortage of people of color and women of color in radiology that you had to get involved because it was just demanded of you, basically, by the specialty.

And correct me, tell me the real story. But it is true that right now you are very actively involved in trying to advance this cause. But tell me more about what can we do and what can we do better? Because this exists at Yale, it exists nationally. It is not going away. And I’ve watched it for 30 years and spoken about it for 30 years and it’s not been improving much, as Harlan said, about health equity.

Melissa Davis: One program that I think is doing really well at this is the ACR [American College of Radiology] peer program, and that’s run by Michelle Johnson who’s on our faculty at Yale. There is a significant positive impact for students who go through that program. Their attrition rate into radiology is very, very high. And so that program targets first- and second-year med students who are interested in radiology or the radiological sciences. They tag them to a mentor, get them a project over the summer, and then they go through coursework through it. So they get really immersed in that. Those types of programs aren’t necessarily scalable, but they’re very impactful for the people that they come to.

We have to put dollars behind that. A lot of times we try to do work for free; we try to get volunteers. We don’t incentivize it, essentially. And so if you really want something like that to get bigger, you have to fund it. And we don’t actually fund diversity efforts. We talk about diversity, we talk about how we want it, we talk about the things that we’re going to do. But when it comes to the dollars and the cents that it requires in order to put those things in place, we actually don’t do it. So the first thing I would say is set aside money and put a lot of money into it in order to do it.

Howard Forman: Yeah, I mean, I do want to say for people that may be listening who have not chosen a career yet, you are the epitome of the quadruple threat because you have been steadily publishing, you are an exceptional teacher, you are a great administrator, and you are one of the best clinical radiologists I know. What—

Harlan Krumholz: And a kind and collaborative person. I think Howie—

Howard Forman: That’s true.

Harlan Krumholz: ... that’s the kind of role model, that kind of generosity of spirit—

Howard Forman: We need more.

Harlan Krumholz: ... in terms of people that you work with. And I really think that’s what we want to try to—

Howard Forman: That’s a good point.

Harlan Krumholz: ... emulate and spread, right? Spread that kind of spirit that kind of spirit that you have, Melissa—

Howard Forman: That’s a great point, but I—

Harlan Krumholz: ... which is just so wonderful.

Howard Forman: I want to know though, of those four things, what surprises you the most today that you might not have thought about 10 years ago when you entered radiology or 12 years ago, whenever that was?

Melissa Davis: Yeah, I mean, that’s easy. I never thought I’d be an academic. That was not my path. I was heavily going into private practice. I had mapped it all out, and I was going to get an MBA and then go become a partner in a practice somewhere and rake in the dollars, essentially. But—

Howard Forman: We are lucky we got you. Even if you’re not able to rake in the dollars, we’re lucky we have you here.

Melissa Davis: No, but I am very grateful for the path that I’ve been on and the doors that have been open from it. This is much more lucrative actually, because at the end of your life, you’re not going to think about dollars, you’re going to think about impact. And I feel like the type of career that I was able to choose will lead me with at least some sort of impact or to be surrounded by people who have great impact every day, like you, you and Harlan.

Howard Forman: Thank you very much for being here. We are so lucky to have you.

Harlan Krumholz: Yeah, it’s wonderful to have you on the show. Thank you so much.

Melissa Davis: Thank you.

Harlan Krumholz: Wow, that was a terrific, terrific session. And she’s just an amazing member of our faculty. I’m so glad we could have her on the show. Yeah.

Howard Forman: Yeah, love her.

Harlan Krumholz: So Howie, let’s get to another favorite part of the show for me, which is to hear what’s on your mind this week.

Howard Forman: Yeah, so there’s been a lot of news... and by the way, this is a topic that is more something that you would talk about because it’s based on a journal article on science, but it’s really, at heart, a big policy issue as well. There’s been a lot of news about Alzheimer’s treatment in the last few weeks. The FDA has approved a new drug called Leqembi or lecanemab, and it shows some marginal benefit for patients that have earlier mild Alzheimer’s disease. Medicare is going to cover that treatment. It’s going to be very expensive, like maybe 25,000 or more dollars per year. And Medicare is separately talking about expanding options to image patients, my specialty, both to help in the diagnosis as well as managing treatment and outcomes in these patients.

And as if that’s not enough, earlier this week, as I mentioned, an important new study came out in the Journal of the American Medical Association, we call it JAMA, showing yet more evidence that this new class of drugs, which is antibodies directed at amyloid deposits, those are tangles of proteins in the brain, that it does seem to impact outcomes favorably. It’s not just treating the finding, it’s treating outcomes. Patients do seem marginally better. Six million people in this country have Alzheimer’s, and it is a major contributing factor to death and disability. We recently had a guest, a friend of yours, a cardiologist, talk about his personal experience with Alzheimer’s. And it is a topic that touches almost every family, often closely.

Donanemab is the monoclonal antibody in this week’s JAMA, and after doing a randomized control trial of 860 treatment in 876 placebo patients over 76 weeks, not an inconsequential length of time, results indicated a slowing of cognitive decline equivalent to four and a half to seven and a half months. And what that might mean is that if somebody was going to become significantly impaired by June, instead, this might delay that decline to December or November. There were meaningful improvements in the actual amyloid deposits in the brain as measured by PET imaging and true clearance of these deposits, meaning just got rid of them in a large minority of patients. It’s not known if these findings are sustainable once you stop treatment, but ongoing trials will help answer those questions.

The impact on outcome is not without risk. Some patients died, probably due to the treatment. And many other patients had side effects that could have been worse if they didn’t stop using the drug. The weight of the evidence suggests that the greatest benefit accrues to really early treatments. So what can we say? A few quick things. One, this is adding to the evidence that we can positively impact Alzheimer’s disease measured by clinical outcomes and imaging findings, and earlier is better.

Two, we will likely see greater screening for earlier disease and many billions of dollars, much more than I think is being reported, much of it coming from Medicare for these drugs and the associated imaging that will accompany it. And I think we’re talking about $10 billion or more a year at the beginning, and it’ll be much more toward the end of the decade. We are just entering this era. One drug approved, another that seems successful, and others in the pipeline. And the last big point, in addition to these drugs being expensive for Medicare and the beneficiaries’ premiums, they will also be expensive for patients with copays and deductibles being quite consequential, maybe $5,000 for a Medicare beneficiary. This is going to worsen disparities if we don’t do something separate to accommodate that.

And as a side note, there is an editorial that reminds us that this particular trial disproportionately underrepresented Black and Hispanic patients. So my take-home: this is going to be one of the most consequential areas of diagnosis and treatment in the next decade. I think it’s going to require patients to carefully consider risks and benefits, particularly for those that want to be early entrants. But it is always good news to have real hope about a horrible disease.

Harlan Krumholz: So I agree with you. It’s always good to have hope about a terrible disease. Look, there are lots of issues here, Howie, as you know, vis-à-vis discussion in the first part of this segment about volume. Now there’s going to be an amazing incentive for medical centers throughout the country to start identifying mild to moderate Alzheimer’s. Mild, Howie. You know how difficult that is to identify and how entangled that can be with depression.

Howard Forman: 100%.

Harlan Krumholz: And then Medicare just lifted the restriction on the number of PET scans that you can get for which you can be reimbursed as part of this, so that this is going to start building mills where people are going to start doing this. So that might be a good thing. And I’m going to say I’m not an expert in this field, but as I read this trial, I just wondered about it. So 1,736 people, 76% actually finished the trial. They’re looking at an Alzheimer’s disease rating score that ranged from 0 to 144—0 to 144, lower is worse—and at 76 weeks, now that’s far into treatment. In the treatment group there was a negative six points decline. Again, lower is worse. And in the placebo group it was nine points.

So range is 0 to 144. And in the one group that was treated, they went down six points and the other group went down nine points. I don’t know, I mean, I’m not, again, an expert in these scales, but I’m just wondering how perceptible that is to people, how meaningful it is to people and what it’s going to do to their resources and what it’s going to do to the healthcare system. It may all be worth it, and maybe that next-generation drugs will show much better benefit, but I’m having trouble... I need to spend more time wrapping my head around exactly what this means.

Howard Forman: Now, you and I are 100% aligned on this. I mean, yeah, you can’t stop this train, but we’re 100% aligned. I’m really worried about, what are we crowding out when we spend $10 billion a year on just this thing, which has such little benefit? But I recognize that people need hope.

Harlan Krumholz: And I’m all in on actually continuing this. It’s a big deal. But I just feel that we’re going to need a lot more research to understand. Then the question will be, Howie, what happens in the real world with this thing? Because this is within the best case scenario of a clinical trial. So there’s just going to be lots to learn here. Lots to learn. You’ve been listening to Health & Veritas with Harlan Krumholz and Howie Forman.

Howard Forman: So how did we do? To give us your feedback or to keep the conversation going, you can find us on Twitter.

Harlan Krumholz: I’m @hmkyale, that’s hmkyale.

Howard Forman: And I’m @TheHowie. That’s @T-H-E-H-O-W-I-E. You can also email us at health.veritas@yale.edu. Aside from Twitter and our podcast, I’m fortunate to be the faculty director of the healthcare track and founder of the MBA for Executives program at the Yale School of Management. Feel free to reach out via email for more information on our innovative programs or you can check out our website at som.yale.edu/emba.

Harlan Krumholz: Health & Veritas is produced with the Yale School of Management and the Yale School of Public Health. Thanks to our researchers, Ines Gilles and Sophia Stumpf, and to our producer, Miranda Shafer. They are extraordinary.

Howard Forman: We’re very fortunate.

Harlan Krumholz: Talk to you soon, Howie.

Howard Forman: Thanks very much, Harlan. Talk to you soon.