Office Hours with Bapu

At the end of each episode, I encourage you, the listener, to write in and let me know what you think of the show. And many, many of you do. In those emails though, you also ask a lot of questions. And the questions are good. So, good that sometimes I wish I had thought of them myself!

We’ve gotten lots of thought-provoking messages, and we thought it’d be fun to actually try to answer some of them on the air. Or at least to talk through how I would try to answer these questions in an actual study. I’ll also get into what other research your questions make me think about.

So, in this episode, I’m going to dissect your interesting questions at the sweet spot between health and economics, along with some assistance from my producer Julie.

We’d like to try to do more of this, so, please, keep sending us those voice memos, or even an email, with any questions you have. We’re at bapu@freakonomics.com. Remember, try to record somewhere quiet, and keep the voice memo to under a minute.

Thanks for listening and we hope you like getting your questions answered.

Julie KANFER: Okay. So, Bapu, our first question today is actually about bird watching.

Bapu JENA: Okay, Julie, this is from a listener named George.

GEORGE: Hi, Dr. Jena, love the pod. After the episode that mentioned the effect a full moon has on driver’s safety, I started thinking more about my driving habits and what distractions I need to be aware of. I love birdwatching. And similarly for other friends in this hobby, birds soaring above the highway can be a huge distraction when on the road. Birdwatching requires becoming hypervigilant of bird sounds and movement that most people tune out and I’ve been wondering if there’s an uptick in car crashes during peak migration periods in the spring and fall when birdwatching is at its highest.

JENA: All right. So, Julie, I’m going to, I’m going to admit I’m—  I — not —

KANFER: Not a bird expert?

JENA: I — yeah, I’m not a bird expert. I know a lot about a lot of things, but I, I don’t know much about birds. So, I, I did have to familiarize myself with just one piece of data. I was just curious when I heard this question, like how big a deal is bird watching? So, uh, guess — just take a wild guess how many bird watchers there are in the U.S.? Just like throw out a number.

KANFER: Are there 45 million bird watchers?

JENA: There are 45 — how did you know that? I don’t know how you — Yeah, there are 45 million bird watchers in the U.S. So, I looked in this question, I didn’t find any research on the relationship between bird watching and car crashes, but that doesn’t mean that birds aren’t causing a lot of car crashes across the United States. But I did think about this. Um, I thought about like, how would I answer this question? Given that there’s no answer that I know of, um, how would I look at it? So, the way I would study is the following. I would try to figure out migration patterns of birds. So, I’d figure out actual days that birds tend to be migrating or periods of the year when they’re most likely to be migrating. Then I’d identify the most common migration routes. And after that, I’d look at the areas that are affected by those routes and look at those same routes in surrounding weeks. So, let’s say birds are migrating through Virginia in, you know, the second week of May every year. Uh, I would look at the second week of May and I’d compare car accidents in that week in Virginia to the surrounding weeks before and the surrounding weeks after in that state. And perhaps what will happen is that birds migrate across different areas, across different times of the year. And so, you could rely on that variation in when birds are migrating and when they’re not to sort of identify that causal effect. So, that’s how I would get to it.

But, you know, whenever I hear a question, whether like this, or in a research seminar, or I’m reading a paper, I’m always thinking about other questions that just sort of pop into my mind because of that. And so, I thought maybe I’d share a couple of immediate reactions to this particular question. So, the other study that this reminded me of was really cool. And this was a paper that was published in the Journal of Science earlier this year. And what it did is it looked at the effects on car crashes when municipalities post messages about road fatalities, uh, let’s say, on a highway. So, the idea is that a town or a county would post messages about how many people have died on this road or die in general on roads. And that’s supposed to be alarming to people and get them to want to drive more safely. And so, if anything, you would think that the posting of those sorts of messages would reduce road fatalities. But what they found actually was that the number of crashes actually increases by a little bit when motorists are sort of confronted with these messages. And the idea I think is that the messages are distracting. So, if you’re driving, you see this message. You’re like, “Oh my, my goodness. Like, you know, there’s this many people that are dying on roads in this area.” That distraction by itself can actually lead to, to more accidents. So, it’s sort of the mantra that good intentions don’t always imply good outcomes, applies here. When I heard about bird watching distractions, that’s where my mind went.

KANFER: One of the conclusions that the researchers made in that study was that the information was too salient. It was too distracting to people. So, is that common? Has — does that happen, like, often with this kind of work, like, where something could be too salient?

JENA: Yeah, it’s a good question. Like, you know, we’ve looked at the issue of salience in healthcare in a variety of different ways. And, and we’ve actually found opposite problem, that lots of things that you would think would be salient to people are not that salient. So, you know, we had this episode actually on a paper that we did that looked at moms who have cervical cancer. And we looked to see whether or not their children are more likely to be vaccinated against H.P.V., because that’s the, the virus that causes cervical cancer. And we thought that this would be extraordinarily salient to these moms. And it wasn’t clearly salient enough because the moms were not more likely to vaccinate their kids against H.P.V. So, I don’t know if we have sort of a uniform understanding of when things are too salient, but I agree with you. What I liked about this study was that it highlights the idea that when you try to make things salient to people, you have to be careful because they may not respond in the way that you would expect them to respond.

KANFER: Yeah. There’s, like, a fine line that we need to walk. You were telling me about your research assistant, who had an idea, and you guys looked into something else that has to do with distractions.

JENA: So, when I’m trying to come up with ideas — you know, most of the time, ideas that I have just kind of come to my mind. It’s like as I’m hearing a question like this I think of something. But we do try to actively come up with ideas and I don’t know how good we are at it, but we’ll sit down a few times a week, myself, others that work with me, and we’ll literally just throw out ideas. And most of them, including my own, are bad, but you know, it doesn’t really matter cause all that matters if you have one good idea. And a few months ago, my research assistant, uh, whose name is Charlie Bray. Charlie, um, proposed an idea to me and the others which was quite strange, I got to admit. When I heard it, I was like, “What is he talking about?” So, what Charlie thought was there are these things called cicadas. I don’t know a lot about cicadas, but cicadas are extraordinarily loud insects. I think they’re loudest noises the insect world. So, they, they pop up in various parts of the country, every like 13 to 17 years, something like that. They live underground and then they come to service and they, they are in these broods. And, and they come infrequently, but they, in theory, could be extraordinarily disruptive because they’re really loud. And so, what Charlie said was, “Well, look, could we rely on the fact that these cicada broods just happen to come up every now and then in different parts of the country, at different points in time? And under the assumption that they are really, really loud, could we then use that to study the effect of loud noise on health outcomes?” Now, this isn’t about distractions, but I think — this is how my mind thinks — birds fly, insects fly —

KANFER: It’s a thing in the air, and it’s something that’s like this outside variable.

JENA: Yeah, exactly. Cicadas could be distracting. So, what Charlie did was he actually contacted a few entomologists. I’d never worked with an entomologist before. And these entomologists had for years, I think, collected data on broods — when and where they occurred. And they were kind enough to share with Charlie that data. So, what Charlie did was he linked the cicada brood data — when it was happening, where it was happening — and he linked it to information on health outcomes of people who live in those areas. And he looked to see whether or not when cicada broods emerged, whether there were, you know — worsening of insomnia, whether there were more driving accidents, more heart attacks cause the stress of all this noise. We looked at a bunch of different health outcomes and we didn’t find anything. So, it’s kind of common for us to, to investigate things like this and not find much, and I wish we had seen something, but we didn’t. So, the cicadas are absolved, but the jury is still out on bird watching.

KANFER: Okay, Bapu, so, our next question is from Jackson.

JACKSON: Hi there. My name’s Jackson, and I’m a really big fan of the show. I’m currently a medical student down in Florida, and I’ve had a question for years ever since I went on a flight on which there was a medical emergency. I’ve been wondering how number of medical emergencies on flights would differ by flight destination.

JENA: Okay. I’ll — I’ll say this, Julie. Um, I actually know something about in-flight medical emergencies. I know more about that than I know about bird watching. I don’t know if that’s reassuring to you.

KANFER: Have you ever responded to one?

JENA: They don’t want physician-economists to respond. “Is there a doctor on the plane?” Yes, but when they say is there a doctor economist on the plane, I raise my hand. And then nowadays if they ask, “Is there a podcast host on the plane?” everybody raises their hand.

KANFER: That’s good.

JENA: All right. So, I actually know a little bit about this, not about this particular question, but before I answer it — so what Jackson was asking was, you know, whether or not there’s differences in how often in-flight medical events or emergencies happen depending on where a plane is going. That if you’re flying to a place, say Florida, where there are older people in Florida, that maybe you would see more in-flight medical emergencies. And I think that’s possible. It’s true and, and I’m not aware of any research that’s looked at that but the other thing that’s the key driver though of how many medical emergencies you’re going to see on a flight is not sort of the demographics of where the flight is coming from or where it’s going, but just how large the plane is. So, I just went to my parents’ wedding anniversary in Richmond, Virginia. It’s a small plane because Richmond’s a small city, even though Boston’s a big place. Now if I’m going from Boston to San Francisco, or Boston to Shanghai, or wherever, those are going to be big planes. So, you’re going to see more medical emergencies when planes are bigger, which is going to happen when planes are going from one big city to another big city.

Now, it turns out that these in-flight medical emergencies are actually more common than you might think. There’s a couple of studies that have looked at this. So, there was a review in JAMA Network Open that said that they occur in about one of 604 flights. So, that’s not that uncommon. Or 24 to about 130 in-flight medical emergencies per 1 million passengers. The most common sources of in-flight medical emergencies are probably not that surprising to people, particularly people who’ve responded to one of these. So, the, the first one, which is the most common, is something that we call syncope, which is basically when you pass out. Now, that can happen for a lot of reasons. The most common reason that people pass out is pretty benign, but there can be instances where someone has a cardiac arrest on a flight that’s obviously very serious. The other things that are pretty common are just G.I. symptoms, like abdominal pain, respiratory symptoms like shortness of breath, difficulty breathing. And then what I’d call cardiovascular symptoms, which is basically going to be chest pain.

Now, you might wonder then, all right, well, what happens when there’s an in-flight medical emergency? So, typically, what happens is that flight attendants respond. They will ask if there are volunteers on the plane, uh, who can help. And I actually know something about that. I’ll — I’ll mention that in a moment. And they’ll also contact ground medical personnel, who help manage the situation from the ground. And every so often the plane actually has to be diverted. And I’ve never been on a plane that was medically diverted, but I’ve had friends who were on planes where they made a recommendation to actually divert the plane and the ground medical crew agreed. Four and a half percent of all in-flight medical emergencies roughly lead to the plane actually being diverted from where it was scheduled to actually land.

KANFER: It’s got to be stressful to make that call by the doctor in the air.

JENA: Yeah. No, it’s extraordinarily stressful. Obviously, if you have a medical emergency on a plane that could be bad, but are there things that are different about planes that lead to problems occurring? So, the two things that come to mind, one is there’s pressure differences in the plane. And there may be differences in the partial pressure of oxygen that’s available. And so, that might pose problems for some people. The other is that if you’re in a long plane ride, there’s prolonged sitting and when you’re sitting for a long time, um, the venous flow — the return of blood from the legs through the veins, back to the heart — is going to be less. So, there could just be more kind of stasis or stagnation of the blood in the legs. There is this association between air travel and blood clots in the legs or blood clots that they can propagate to the lungs. We call that a pulmonary embolus. And the blood clot that is in the legs, we call that a deep vein thrombosis.

So, there is an association between airplane travel and those things that’s thought to be mediated through at least this prolonged sitting. So, that’s sort of a background that I think about when, when someone’s talking about in-flight medical emergencies. Now back to Jackson’s question, so I’m not aware that anybody’s actually looked at it, but I would say that the data does exist. So, there’s a couple of places, and I’ve worked with data from an organization called Med Air, which is one of these organizations that manages flight emergencies or flight events from the ground. And so, they have a database of, uh, events that I’ve that, I’ve used before but I’m not aware of anybody actually looking at this specific question that Jackson has asked.

KANFER: But if we did want to look at it, what kind of data would you need? What do we know about like the oldest places in the country and the youngest places? Obviously, we’re not going to, I guess, get into international flight. So, let’s say you’re keeping this the United States.

JENA: So, I think Maine and Florida are the two oldest ones. Uh, I don’t know what — do you know what the youngest ones? I don’t I actually don’t know what the youngest states are in the country.

KANFER: Okay. So, the five oldest states — and this is based on percentage of the population over 65 — are Maine, Florida, West Virginia, Vermont, and Delaware. And the five youngest states are Utah, Alaska, Texas, Georgia, and Colorado.

JENA: Okay. All right. So, I would use data from one of these in-flight medical event databases that exist. And I would look at flights going to and from cities that are younger versus older. And I would try to hold constant the size of the plane, because remember the point that I made earlier, which is that big planes are going to have more medical emergencies than small planes, all things being equal. So, you’d want to hold that constant. Jackson’s question is whether or not there’s more emergencies when there’s a flight that goes to a part of the country where people are older, like what would you do with that finding. And it occurred to me that like when there’s a situation where a plane is going to have a higher probability of having in-flight medical emergencies, you might think that the attendance on those flights could be better trained, in theory. You could do that. You could have a more robust medical kit than is standard. So, the F.A.A. could mandate that for bigger planes where the likelihood of bigger medical problems is greater, you might want to have more at the disposal of any physician who happens to be on the plane.

So, those are like practical things that you could do, if Jackson’s hypothesis turned out to be correct. But I do want to spend a moment to just talk about other things that came to my mind when I heard Jackson’s question. So, there’s a, a surgeon and health services researcher named Ben Mundell, who’s at the Mayo Clinic who, about a couple years ago, sent me this email. This really, really interesting question. So, let’s say that Chicago is hosting a big cardiology conference in a given year. Uh, presumably there’s going to be more cardiologists on those planes going to Chicago uh, during the week of that conference than in other weeks of the year. So, Ben got this data and he looked to see whether or not there’s any differences in how in-flight medical emergencies are managed during the dates of those conferences versus the surrounding weeks. And he basically found that there is no difference. So, there’s no increase in in-flight responders — medical responders. So, that’s what comes to mind when the words “in-flight medical emergency” are put into my head, as Jackson just did.

KANFER: That was a good question.

JENA: That’s a good one. I agree.

After the break, we’ll talk about how Covid lockdowns did, and didn’t, impact our health, as well as why screening for certain cancers earlier may, or may not, be worth it. I’m Bapu Jena, and this is Freakonomics, M.D.

*          *          *

KANFER: So, Bapu, you just mentioned your colleague, Ben, and you told him about this episode we were preparing. So, Ben sent us a question.

MUNDEL: Hey Bapu, Ben Mundel, a general surgery resident and health services researcher. Do you think that during the Covid lockdowns, there’s an increase in incidents in pilonidal cyst development due to more people sitting on their butts?

KANFER: Bapu, what is a pilonidal cyst?

JENA: They’re basically cysts that occur when hair punctures through the skin and then gets infected. And an abscess can form. They’re often very painful. Sometimes they have to be drained through a small incision. Sometimes they respond to warm compresses or just get better on their own. They often occur near the tailbone of the top of the — near, near the buttock most typically. A risk factor for those cysts is sitting for prolonged periods of time. So, think about, like truck drivers might be a classic example of a group of people who are at higher risk of developing a pilonidal cyst.

KANFER: So, people were sitting around a lot, uh, at home, social distancing in those early days of the pandemic. Um, and, so, what did you guys look at?

JENA: So, this was actually a question that we could look at. We could actually get some data to, to look at it. And the data that we used was from this company called Tru Veta, which is a E.H.R. or electronic health record company. By the way, one of the benefits of electronic health record data is that they’re close to real time. So, most of the data that I use is not even available during the pandemic because it just lags in getting the data. But E.H.R. data tend to be more recent. And so, we looked at whether or not there are actually increases in diagnoses of pilonidal cysts or procedures to treat them during the pandemic. And you know, to Ben’s point, I would’ve that we might see an increase. The only thing that would’ve made me think that we wouldn’t have is that we knew that medical care actually failed during the pandemic because people were afraid to go to the hospital and hospitals were prioritizing care. But we looked at it. And so, we did not find any evidence that these pilonidal cysts, these abscesses on your butt became any more common during the pandemic — which again, could have occurred because more people were working from home, possibly sitting on their rear ends for, for longer periods of time. So, that’s what we found. I will say that these cysts are pretty common though. More than 70,000 cases are diagnosed each year. So, they’re painful and they’re common.

KANFER: There was a lot of other research, right, that looked at other ways in which staying home and lockdowns did affect people’s health, whether because they didn’t do things or because they did.

JENA: Yeah, yeah. So, again, just sort of train of thought here, thinking about all the ways that the pandemic has affected health and healthcare behaviors outside of Covid, right? So, the prominent way that it’s affected people is through the disease, which is COVID-19. But there’s a lot of things that also happened in that pandemic that affected people’s health. I mean mental health, for example, we know has been an issue during the pandemic. The other thing that I think is quite interesting, there’s been some work looking at delays in care. We did a study which looked at delays in mammography and colonoscopy. This is screening for breast cancer and colon cancer. And in the data that we looked at colonoscopy and mammography rates fell by about 90 percent in the commercially insured adult population. So, that’s a huge decline in cancer screening and it remains to be seen what will be the impact of that on actual delays in cancer diagnoses. Perhaps worsening cancer outcomes — like all that remains to be seen, but we know from a number of different studies that there were delays in care.

The morbidity and mortality weekly report, or the M.M.W.R’s as it’s called, it’s published by the C.D.C. They found that by June of 2020, because of concerns about Covid-19, 41 percent of U.S. adults had delayed or avoided medical care, and that included even urgent or emergency care. Uh, and certainly, included routine care. So, routine care was much more likely to be delayed, but even urgent and emergency care was also delayed. I think if you talk to people who are working in healthcare settings during this time, they would also vouch for that and say that when people were coming in, they were often coming in at much more advanced stages of disease. The other thing that sort of line of questioning reminds me of, there was a study that just came out really recently, just a couple weeks ago, from the National Bureau of Economic Research and it sort of looks at the reverse issue. What they were looking at is whether or not Covid-19 vaccination affected people’s propensity to delay or skip medical care. And the basic idea is that when the vaccines come out, people can get vaccinated. They might feel much more comfortable now resuming medical care that they felt uncomfortable doing before they were vaccinated, because they were either worried about getting Covid-19 themselves or perhaps spreading it within a healthcare setting. They went even further ahead in what I think was a clever approach.

They don’t just look at people who are vaccinated and not vaccinated. What they did is they relied on the fact that there are these specific age cutoffs for vaccine eligibility and there is differential rollout, also, of vaccines across states over time. So, they use that as a natural experiment to look at what happens when some groups of people by chance, maybe because of their specific age, got access to a vaccine while others didn’t. And so, what they find is that receiving a Covid-19 vaccine reduced the likelihood that people would delay care for any medical condition by about 37 percent. So, that’s huge. That means that getting the vaccine made people less likely to delay medical care. They also found something that was really interesting, which I’d call sort of a spillover effect. So, they looked at families and they found that children were also significantly less likely to delay or skip medical care as a result of vaccine availability for their parents. So, we’re talking about sort of parents who, by chance, because of their age or where they live get access to the vaccine. When that happens, their kids are also more likely to resume normal routine medical care. They also found that the decline in this delayed or foregone care that was caused by getting the vaccine was actually quite a bit stronger among minorities and those in low socioeconomic groups. So, sort of pointed to an important way in which vaccination might help promote more routine care or might reduce delays in care across socioeconomic, or race-based lines.

KANFER: That’s really interesting and speaks to even really why the vaccines were even more important than maybe we thought.

JENA: Yeah.

KANFER: We just talked a little bit about cancer screenings when we were answering the question about Covid lockdowns, and how it may have affected our health. So, we did get a question along those lines from another listener.

JENA: Okay, let’s hear what our listener named Sara has to say.

SARA: Hey, Dr. Bapu. Big fan of the Freakonomics M.D. podcast. My question for you is do you think cancer screening programs work, or does lead time bias take an effect? Would love to hear your take and yeah, thank you for the podcast.

KANFER: What is lead time bias? Let’s start by explaining what that means.

JENA: So, lead time bias is this, is this idea that when you’re screening to identify cancer, if you screen and find a cancer earlier, then that cancer would’ve been diagnosed based on, let’s say, symptoms, that lead time bias can cause a problem because the earlier diagnosis may do nothing to change the course of the disease. So, all you’ve done is identify that someone had cancer earlier without changing their ultimate prognosis or the day — literally the day at which they would die. The reason people call it lead time bias is because if you measure survival from the time of diagnosis, that can be misleading. And so, let me give you an example. Suppose you’ve got a man who’s 67 years old. And you know, he comes to you with shortness of breath and a cough. And he’s a smoker, let’s say. So, you’d be thinking about lung cancer. Let’s suppose he has lung cancer, he’s found to have lung cancer, and then he dies at age 70. So, he was 67, had symptoms, found to have lung cancer, and then died at age 70. Now, suppose you had screened that man for lung cancer earlier, say because there’s a diagnostic test like, C.T. scan that would allow you to do so. So, you screen that man for lung cancer at age, uh, say 65.

The key question is whether or not screening at age 65 would have made that man live a day longer than to age 70, which is when he died in the first scenario. If not, all you did was find his cancer earlier. It didn’t affect the survival. And the reason we, we say this is a lead time bias is if you measure his survival from the time at which he was diagnosed with cancer, you would say he survived three years from diagnosis from age 67 to age 70. If you then said how long would he have survived from the time of diagnosis if you measured his survival with the screening test? He would’ve lived five years. And so, that naive comparison you might say, “Oh, well, we should have screened him because he lived five years from the time of diagnosis,” but that’s not correct. All you really care about is: when did this individual die? Now, the particular question that was asked is, you know, what is the relationship between — or is there any evidence that cancer screening affects outcomes. And I’ll say that there is. And I actually have another thing I want to mention in a second. So, there’s a couple of cancers that are very common like breast cancer, lung cancer, uh, cervical cancer, colon cancer. And for most of these cancers, there actually are robust randomized studies looking at various cancer screening strategies that have demonstrated that screening is effective in prolonging survival. So, that’s important to keep in mind.

So, lead time bias is not an issue there because we’re looking at when people are actually dying. That’s their survival, um, and they live longer. There is a famous study that was published in the New England Journal of Medicine a few years ago called “The National Lung Screening Trial.” So, in that trial they looked at about 50,000 people who were at high risk for lung cancer and they randomized them to one of two screening approaches. One group of these individuals were screened with something called a low dose C.T. scan and they, they’d receive that annually for three years. And the other group received a chest X-ray annually for three years. So, the C.T. is, is, is going to be able to give you a much more granular image and tell you what’s happening in the chest much more so than a chest X-ray.And it was a randomized trial. So, the researchers found a relative reduction in lung cancer mortality of about 20 percent in the low dose C.T. group. And actually, they had lower overall mortality, not just focusing on lung cancer, if they received a low dose C.T. compared to chest X-ray. And so, that sort of has actually made it into clinical guidelines as practice recommendations for people who are at high risk of lung cancer, for example, from a history of smoking.

KANFER: Is lead time bias only relevant to cancer screening and cancer detection?

JENA: It’s primarily relevant to that, but it’s really true for any disease in which you’re screening for something. So, let’s say you’re talking about heart disease. Okay? And you’re, you’re measuring the degree of plaque in someone’s heart. You might identify something that would not cause a problem, but if you measure, like, the time of survival from the time that you identify “heart disease”, you’d say that this person’s living longer from the time of diagnosis, but it’s only because you diagnosed a problem that would never have actually led to an issue in the first place.

KANFER: So, can we do anything about lead time bias?

JENA: Yeah, I think we can. So, I think that the way to deal with this problem of lead time bias is A, for people who do these studies to know what lead time bias is, and B, when you’re talking about survival in any condition not to focus on survival from the time of diagnosis, necessarily. Sometimes it’s okay, but you want to just look at overall survival or the age at which you die. So, any screening intervention that does not make people live longer is not effective, even if it diagnoses you with cancer earlier.

KANFER: And this just sort of makes me think of like, if someone is diagnosed earlier, then they maybe otherwise would’ve been, and maybe they had pretty good quality of life and then they start undergoing cancer treatment, so they’re simply just undergoing treatment for a longer period of time without actually doing anything to extend their life. Is that an issue?

JENA: Yeah. That’s a huge issue. And I think, like, some people may want to know that they have cancer not because identifying the cancer and treating it would affect how long they live, but simply because they might do things differently. So, that, Julie, I think your — your point is it’s a really important one. So, when I’m talking about lead time bias and what screening programs should or should not focus on, I’m being a little bit fast and loose. I mean, I’m talking about mortality, but a lot of things matter besides mortality. So, if you can tell someone that they have a medical problem that’s going to shorten their life, even if you can’t extend their life in any way, perhaps the quality of their remaining life might be different because of how they choose to live their life. Maybe they quit work. Maybe they spend more time with family. Maybe they travel more. Um, whatever it may be. But so that’s — I think that’s really important.

KANFER: Okay. Well, that was a really good question. And now we have one more question.

WIN: Hey, Bapu, my name is Crystal Win. I’m a peds resident around Chicago, Illinois. My question to you is: is there an increase in adverse patient outcomes during the July to August timeframe, when new residents are starting and current residents are transitioning into more senior roles?

JENA: Wow. Crystal, Crystal, Crystal from Chicago.

KANFER: Crystal clear.

JENA: Crystal clear. Yeah. That’s a really good question. And it’s so good that we’re going to devote an entire episode to it next week.

WACHTER: My first day. I showed up and one of the third-year residents was finishing his year and about to leave and handed me the beeper. And said, “Good luck, sucker.”

Every July, a medical training rite of passage occurs. Medical school graduates begin their residency programs as interns, and current residents and fellows ascend the training ladder. But where does all of this leave patients?

HUCKMAN: Somehow the system has to train new physicians. And anytime you have a new physician, there’s going to be some learning curve.

We’re going to talk about what’s known in medicine as the July effect. Is it real? And if it is, what should we be doing about it?

ARORA: There’s a great variability of how people can approach this in a thoughtful way so that the residents have a positive experience and that patients are not harmed.

That’s next week—on Freakonomics, M.D. Thanks for listening and thanks to my producer Julie Kanfer for joining me on this episode. Don’t forget to send us more of your questions, either as a voice memo or an e-mail. We’re at bapu@freakonomics. com. And remember to leave us a review wherever you get your podcasts! I’ll talk to you again — next week.

*          *          *

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. This episode was produced by Julie Kanfer and mixed by Eleanor Osborne. Our staff also includes Neal Carruth, Gabriel Roth, Greg Rippin, Rebecca Lee Douglas, Morgan Levey, Zack Lapinski, Ryan Kelley, Jasmin Klinger, Emma Tyrrell, Lyric Bowditch, Jacob Clemente, Alina Kulman, and Stephen Dubner. Original music composed by Luis Guerra. To find a transcript, links to research, and a newsletter sign up, go to Freakonomics.com. If you like this show, or any other show in the Freakonomics Radio Network, please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

*          *          *

KANFER: So, wait, it was the — it was the cicada data?

JENA: Oh, cica —

KANFER: Do I — should I go?

JENA: Cidata. I like that. I like that.

KANFER: Should I leave after that joke?

JENA: Yeah. Don’t make me say that six times. Yeah. It was the cicada data that Charlie put together.