Episode 364: Anupam B. Jena
Listen to Episode on:
Watch the Unabridged Interview:
Order Book
Innovation at the Crossroads of Medicine and Economics
How can economics shape our understanding of healthcare? How can cognitive biases lead even seasoned doctors into harmful errors, and what potential does machine learning have to mitigate these mistakes?
Anupam B. (Bapu) Jena is a professor of Health Care Policy and Professor of Medicine at Harvard Medical School, Associate Physician in the Department of Medicine at Massachusetts General Hospital, host of the Freakonomics M.D. podcast, and the author of the book Random Acts of Medicine: The Hidden Forces That Sway Doctors, Impact Patients, and Shape Our Health. He has an MD in medicine and a PhD in economics.
Bapu and Greg discuss the impact of timing on healthcare. From the intriguing effect of birthdate on ADHD diagnosis and how patient outcomes in medical care correlate to how recently a physician was in residency to the puzzling improvement of cardiology patient outcomes when there are fewer cardiologists present. They discuss the lessons learned from Covid through both medical and economic lenses and why the effectiveness trajectory of surgeons differs from that of other physicians. This is a fascinating dive into the data side of medicine, with several surprising takeaways for any listener.
*unSILOed Podcast is produced by University FM.*
Episode Quotes:
On the intersection between medicine and behavioral economics
25:30: One of the things that I've thought about the intersection of behavioral economics and medicine is that it may not be surprising that the types of things that behavioral economists will often study show what they do, right? Because when the stakes are not high, it's not hard to sort of rely on that autonomous part of your brain that tells you that someone who's 40 is different than someone who's 39. Why does it matter? You go to the grocery store, something $7.99 versus $8, and yes, you're more likely to buy it when it's $7.99 versus $8. But you wouldn't think that if that applied to life-saving chemotherapy, you'd be more likely to buy the chemotherapy that's $7,999 versus $8,000. You would think, wow, life-saving chemotherapy. I'm going to figure out which therapy I need. I'm not just going to let a mind trick or heuristic move me in one way or the other.
What do we need to do in order to do medicine better?
47:20: What you really want to train doctors to do is be able to create differential diagnoses, like clinically problem solve, and understand that when things aren't lining up the way they should, to know that they need to search further to figure out the answer. In that aspect of problem solving, I think economists do very well because the nature is, the work is different in that respect in medicine; I think we rely too much on pattern recognition to sort of help us understand answers to questions. And there's like a reasoning that is sometimes not taught.
Do we demand too much from our doctors?
28:48: It's easy to miss things, right? And it's easy to not realize what's the big thing that you need to be looking for versus the small thing. And you get that with experience. I think with experience, you do better there, but it's certainly the case that even with experienced doctors, they still miss things, and I think that's where computers can be really helpful. They can be in the background, as they see. The data that is generated on a daily basis for a person, seeing what the past medical history of that person is, seeing the trajectory of all of those things, like how are the labs looking over time, what is the imaging looking like over time, what are the nurses' notes saying about delirium or agitation, are there more mentions of that as we're going on? A computer could see all that information and put it together in a way that a human being might not be able to. And at the very minimum, I think it could offer us some good insights that can help us consider things that we weren't considering.
On the origins of Anupam's study
18:21: It was another data point to tell people that sometimes less is more, but what I think it did more of, and probably what most of my studies do more of, instead of moving the needle in terms of one specific clinical practice, is just getting the medical world more in tune with these ideas of natural experiments and trying to just be a little bit more curious and innovative when we come up with approaches to studying questions because a lot of what we study in medicine is critically important. I mean, it matters for our health; sometimes it matters for life and death. We have the ability to do randomized trials, and those are great, but sometimes we can't do them, or we don't do them. And we can't phone it in for those other types of analyses. We've got to be as curious, as intellectual, and as creative as we can be to try to figure out the right answer.