 On behalf of the McLean Center and the Center for Health and Social Sciences, David Meltzer and I, where's David? David, welcome you to this week's lecture in our 2018-19 series on improving value in the U.S. healthcare system. It's now my absolute pleasure to introduce our speaker today, Dr. Neal Shaw. Dr. Shaw is in the Department of Obstetrics, Gynecology and Reproductive Biology at the Harvard Medical School. In this role, Dr. Shaw serves as an obstetrician gynecologist at Beth Israel Deaconess Medical Center in Boston, where he cares for patients at critical life moments, moments that range from childbirth to primary care to surgery. Dr. Shaw also is the director of the Delivery Decisions Initiative at Ariadne Labs, which is a health system innovation center sponsored by Brigham and Women's Hospital and the Harvard School of Public Health. The Delivery Decisions Initiative is dedicated to seeking new approaches and solutions to childbirth care that produce quality, dignity and equity. Through his work with Ariadne and Harvard, Dr. Shaw is a globally recognized expert in designing, testing and applying solutions that improve healthcare. Dr. Shaw is listed among, quote, the 40 smartest people in healthcare, end quote, by the Becker Hospital Review and has been profiled in the New York Times, CNN and other media outlets. He's senior author of the book, Understanding Value-Based Healthcare. Prior to joining the Harvard faculty, Dr. Shaw founded Costs of Care, a global NGO that collects insights from clinicians that helps systems provide better care. In 2017, Dr. Shaw co-founded the March for Moms Association, a coalition of 20 organizations that increase public and private investment in the well-being of mothers. Today, Dr. Shaw's talk is entitled, I have a slightly different title, but growing a family with dignity. And so it gives me a great pleasure and to give a warm welcome to Dr. Neal Shaw. Thank you so much. Thank you, Dr. Siegler. Honestly, it's an honor to meet you, let alone be introduced by you. So you should probably start by just lowering expectations considerably from that introduction. Yeah, I'm sorry for switching up the title on you too. You know what it was? I read up a little bit on your work on the way in and noticed that, thank you, that a lot of the modern thinking on the doctor-patient relationship came from seminal work that you did. And I just thought, no, I'll still talk about designing healthcare systems at scale, but I thought I'd anchor it in the purpose, which for me right now is that. And if you don't mind, I'm not used to having a full hour to talk. So I'd like to just acknowledge how it came to be that I'm here right now. So I'm at the University of Chicago right now because 10 years ago to the day, a nonprofit incorporated costs of care. And at the time, it was just a website and a manifesto about the role of clinicians and thinking about the affordability of care for our patients. The sole reason that it is still here over 10 years is because Vinny Arora found me on Twitter almost a full decade ago. And then even though she was a dean and a professor and I was a resident, took me seriously and helped us build it into a serious organization. And that book, we wrote it together with our colleague Chris Morades. So today is exactly the 10th year anniversary. And then obviously through Vinny, I came to know David personally and here I am. So yeah, yeah, I guess the common thread between the costs of care work and what I want to talk about now is, if you don't mind, I'll just tell you a quick story. So I mean, basically I went to college and med school at Brown. It's a place where a lot of people sign up for med school when they're 18. They have one of those programs. So my whole notion of what being a doctor was was based on TV. Like, I rolled right from college into med school. We had a pretty traditional curriculum where you spend the first couple years in a classroom. And then all of a sudden you're a third year medical student. You're in 25th grade and they let you touch real patients. And for me, like, and I think probably for all, how many people here are clinical trainees still sometimes? I mean, like, basically, you avail gets lifted. And there are aspects of what you see that are inspiring. And then there are aspects of what you see that are like deeply disillusioning. But for me, you know, Providence is a pretty impoverished city. The kinds of people who come to see trainees for care are not the ones with the deep pockets. And I just had this frustration. I mean, I didn't have a soapbox going in. I was the most straight-laced pre-med Indian kid from New Jersey. But the fact that we're making decisions on behalf of patients with very little insight into what they paid for it bothered me. The fact that nobody else setting an insight into it bothered me. So it sort of began this journey of thinking about, you know, the goals of improving care and whether or not we had the right ones. Generally speaking, many efforts to improve safety in health care are essentially that. Like, they're almost synonymous, improving health care, improving safety. And I had a sense that that was important, but maybe too narrow, which, and then, you know, ultimately, for a variety of other reasons, I accidentally became an OB and my window into the American way of life ended up becoming my clinic where, you know, you see the same families about a dozen times in nine months. So let me reintroduce myself. I'm an obstetrician. I lead a research program at Harvard that's ostensibly about maternal health. And for the last couple of years, the focus of my work has been largely on trying to, you know, improve safety and childbirth. But more recently, I've realized that to really address the challenges that we're seeing, we need a much bigger and bolder vision for the world that we want to see. And it's a vision that, you know, doesn't just treat childbirth as a transient episode in the lives of some people, but it treats childbirth as the foundational episode in the lives of all people. It's a vision that's not just about the four walls of a hospital. It's equally about the communities where people live their lives. And it's a vision that's not just about the president. It has to be about the future, too. And basically the pact that every current generation ought to have with the next one to leave things at least as well as we found it. And then, you know, just the fact that when it comes to the way that we're born, we're failing. Turns out for an American mother today, she is 50% more likely to die in childbirth than her own mother was. It's a real statistic. Really. The maternal mortality has been going down with a trend line in sub-Saharan Africa for 25 years. It's been going up with a trend line steadily for 25 years in the United States of America. Now, if you're black, you're three to four times more likely to die in childbirth than if you're white, irrespective of your education or your income level. And, you know, these deaths are just the canary in the coal mine of a much deeper problem. Thankfully, it's a relatively rare occurrence. It's still 700 moms who are young, healthy women for the most part who die in this sort of universally vulnerable period from being pregnant through parenting and infant, which is how the statistics are captured. But for every death, there are about 100 severe permanent injuries. For every one of those injuries, there's a broader problem of just chronic illness, social isolation, economic disempowerment. There was a paper that I saw that came out a couple weeks ago that was picked up in the press. I think it was a Princeton economist that I'm sure David knows or may know of who was studying family leave policies in Denmark. But it's not Janet Curry. It was one of her colleagues. Although she doesn't do a lot of the work on moms and families, but essentially, you know, it turns out in the United States, compared to every other developed country, the number of women who stay in the workforce after having a baby drops more precipitously. That, you know, if you're one of the 80 million Americans who earns minimum wage, the costs of childcare are about 60% of your total income, the same income that you would deploy for rent and healthcare and everything else. That's a good question. A remarkable statistic you mentioned. If you break it down by race, is it true that the rates are going up across both whites and blacks? Or is it minorities? Or is it that the rates are dramatically up over minorities compared to a generation? Both are true. So the, it's almost entirely driven by mortality rates among black women, but it is true that it's also going up across the board. And it's actually also true that this hasn't been widely reported, but it's on the Vital Statistics CDC website, not as a sentence, but if you were to crunch the numbers, what you'd see is that mortality rates among girls under 15 and women over 45 have been going down, but mortality rates for reproductive age women in general are going up steadily over time. Not to that degree, but significantly. I'm sorry to start with a labor at the point, but among women who get prenatal care, are they going up as well? It's really, really hard to divide further than that. And actually one of the few things that have been signed into law, like actually the day ahead of the furlough, there was a maternal mortality bill signed into law and the entire purpose of that bill was just to get the CDC to track more closely. The reason that this snuck up on us is that actually we weren't closely monitoring. We didn't have a good surveillance system for maternal mortality. We're even a standard definition. It's rare enough that it's really hard to get a signal, except for the fact that there's a 25-year trend and it's been... Yeah, so those are the facts. This is my friend Charles Johnson. He sort of like put a point on what's happening in our country. So Charles and his wife Kira had their second child exactly three years ago. In fact, I think Langston's birthday was last week. And he and his wife went into a tertiary medical center that actually where the labor floor is run by one of the gurus in maternal safety to have their baby. Charles is the son of a celebrity. He's wealthy. His wife Kira also wealthy. Spoke five languages, highly educated, very healthy, ran marathons. And as Charles tells it, they ended up getting a routine C-section. And after the C-section in recovery, she started to look a little off. And one of the things that's hard about OB is that young healthy women look fine right up until when they don't. And when they don't, you've basically already lost. But he felt that there was something off. And, you know, he knows a little bit of blood in the catheter and just like she just didn't seem quite as... Didn't seem quite herself. And as he tells it, he spent hours trying to get somebody to respond to his concern. And there was no one person. You know, it was the nurse. It was the residents. It was the attending on call. Ultimately, it took something like eight hours for them to do a full evaluation, realize she had an acute abdomen, get a CT scan, realize she was bleeding. And then even when they brought her back to the operating room, they still had the hubris, right, that we all have where they said, you know, we're going to be back in 15 minutes. And Kira didn't make it. And as Charles tells it, there's no statistic that can convey what it was like for him to then go back home and tell his two-year-old that his mom wasn't going to come home. There are a lot of things that are powerful about Charles' story. I mean, one is that Charles himself is an incredibly inspiring person. He spent the subsequent three years relentlessly advocating for just drawing attention to the issue. He kept testifying on the Senate floor and getting this bill signed into law by Donald Trump the day before the government shut down, almost single-handedly. You know, it's remarkable because Charles and Kira are like the last people you'd expect this to happen to. Although, you know, for those of you who are tracking, his story is very similar to Serena Williams' story. Beyonce actually has a similar one as well, but there seems to be a thing where, you know, the racial disparity here is obviously complicated. But one of the challenges at the point of service, and I think this comes down to the doctor-patient relationship stuff, when black women express concerns particularly around pain, somehow we believe them less. And if we have time to talk, I would love to unpack that. Because there seems to be a thing where, you know, as doctors we're basically taught to profile people, right? Like, that's our job. If you're a brand new intern and you walk into the hospital, the only thing you're really expected to know is whether or not someone looks sick. Like, how do you decide that without objective information? Like, somebody looks pale, but if they're black they won't look pale. Every case study in medical school tells you the race of the person so that you can make a set of assumptions about their risk. And then underneath it, there's often a deeper set of assumptions about what's driving those risks that, you know, every table one of every epidemiologic study has a row. This is race. And we all know that's a proxy, right, for racism. And there's a very, very thin line between what we celebrate as clinical intuition and what is effectively racism. If we're right, then we're a master clinician. So that was a total sidebar rant. But the other thing that was really striking about Charles is that his son Langston and my son are the same age. And meeting Charles actually made me reflect a lot on... I think the combination of meeting Charles and then experiencing being a parent was a big reset for me. I've cared for thousands of pregnant women at this point. I've delivered thousands of babies. I've studied the whole detailed pathway from which like an embryo becomes a blastocyst and then a fetus and then a human. But the whole notion of parenthood was an abstraction to me until I actually held our son for the first time. It's actually made one of my favorite parts of my job is, you know, moms feel the baby move and hiccup and kick for months. But for almost every partner, it gets very, very real the minute that the baby exists in the physical world on the outside. The look on like a father's face or a partner's face. And so like Julie, my wife, her water broke in the middle of the night. It was a couple days before she was full term, but definitely caught us off guard. But you know, I'm an obstetrician. I had it handled. You know, we made sure that we got into the hospital right when she was an active labor. We didn't come in too early. It was my own labor floor. Julie was very clear that my job was not to be the obstetrician. That I was supposed to be on the other end of the bed. And, you know, my colleagues helped take care of us. My residents helped take care of us. Obviously, I had tremendous trust in them. And Julie's labor went fine right up until the second stage. So the second stage of labor is, oh man, it's so annoying as an obstetrician. Because, you know, if it's the middle of the night and a woman's been pushing for two hours and then three hours and then four hours, there's no technology that can tell you how long it's going to be. There's no technology that can even tell you how big a baby's head is until it's out. What determines how long it's going to be is how big the baby's head is, how wide the pelvis is, and probably more importantly than both how the two are lined up. And the state of the art to figure that out is a trained pair of hands in 2019. So if you're like in the trenches as the OB, you have to just, it's an artisanal craft. You just have to decide what you believe. The longer you wait, the higher the odds you'll get a normal delivery. The longer you wait, the higher the odds that the mom or the baby will get injured. So it took, you know, Julie, I was doing my job, I thought. I was getting water bottles and spraying her down Rocky style, like whatever she needed. But it was driving me crazy, right? Because I was basically sitting on my hands. She had a five and a half hour second stage. She was about twice as long as statistically expected. And I was imagining, those are the worst C-sections, by the way, when the head gets really impacted and lodged down there. Those are the worst. Is anybody here an OB? No one to relate to on this. We have future OBs. Let me tell you. Where's the future OB? It's just true. I mean, I feel like I've lost years off the end of my life over these second stage arrests where you put your head in and you just hope to God the baby is going to come out. And you use brute force and you wrap your fingers. I was imagining this for our son. Around the head, you try to create a wedge between the head and the pubic bone and try your best not to compress the brain directly. And then you pull with everything that you've got, sometimes with two hands. So this cleaved my identity in half. Because five and a half hours in, I was craving the operating room, the familiar blue drapes, like the finality of the scalpel, the sense of control. But I was also very, very conscious of my very stubborn wife who had the same steely look in her eye that she had when we hiked the Inca Trail together, where she was clearly exhausted but too far down the path to turn back. Like she was just ready to do it. And the notion of removing her agency bothered me a lot. And then of course, like our son stopped goofing off, lined up, and came out. When a baby is not lined up right, it's almost like a drawer that's off of its tracks. You can pull as hard as you want. And then as soon as it lines up, it's like done. That's what happened. Here's a half century story on C-sections and just what's been going on. I mean, for a lot of people who are practicing at U of C right now, what I'm about to show you has taken place during their entire time in practice. Like I have many colleagues who have watched this whole thing happen. In the late 60s and early 70s, the national C-section rate was about 5%. In 1971, something happens and it rapidly doubles. In 1976, actually, that's the first time there are papers in our journals that point out a skyrocketing national C-section rate. Five years after that, it more than triples. Five years after that, it quadruples. Guess what happens five years after that? Yeah, it quintriples. And then like the 90s happen. I feel like there's a few people in the room that probably remember the 90s. That was like when AOL CDs were being mailed to you. Yeah, that wasn't funny enough to enough people. So in the 90s, we set a bunch of population health goals, like about our target BMI as a country. We had a target C-section rate, too. This is based on a WHO recommendation of bringing it down to 10% to 15%. And in the 90s, we'd bend the curve. And if you were to continue it, we would have hit the target. But in the mid-90s, we'd change our mind and completely reverse course. And then we do this. This is a 500% increase in the use of the surgery over the last generation or two of moms. 500%. And the thing is, we don't know why. There is a lot of conventional wisdom about what's been driving this. And it's all wrong. All of it. All of it. So when this first starts taking off, that's when we introduce the technology. Fetal monitoring, the ability to do telemetry on a fetus. And within three years, it fully saturates the global market. Everywhere where there's electricity, they're using this thing. And what it's designed to do is reduce neonatal morbidity, specifically cerebral palsy from inoxia. It does not do that on average. The only thing it does reliably is increase C-section rates, but we haven't been able to get rid of it. Now, the thing is, because it fully saturates the market in three years, you can kind of blame it for starting this, but you can't blame it for driving it to 500%. Because basically beyond 1975, the way that we used fetal monitoring in the 70s, 80s, 90s, 2000s and 2010s hasn't changed at all. And meanwhile, this has kept going up. Do you buy that? Now, you might also say, well, you know, this is not a very long period of time. Most moms have more than one baby. In 2019, if you have a C-section the first time, you have a 90% chance of getting a C-section the second time, whether you need one or not. And so you might say some of this is cumulative. And I don't think I have it graphed on this particular slide, but if you were to graph the primary C-section rate, just the rate among first time people for whom it's their first C-section, they're like railroad tracks. They go down at the same time and they go up at the same time, which means it's not just a cumulative effect. And then you're like, okay, well, moms look different today than they did in the 60s and 70s. There's more obesity and hypertension. Moms are older. There's more IVF. There's more octomoms. There's only one octomom, just to be clear. And it turns out that demographic shifts don't explain any of it. Any of it. And the reason is that rates have gone up in healthy 18-year-olds by the same amount that they've gone up in 35-year-olds and there's more young people having babies than there are older people. In fact, if you're 18 today, your odds of getting a C-section, even if you're thin and healthy, have roughly doubled over the course of your life. So you really can't blame women. I mean, the thing is, there's also a narrative about people demanding C-sections. One of my favorite, there's an article in the UK Daily Mirror where the headline is too posh to push and there's a picture of Victoria Beckham. None of you were here in the 90s. Less than half a percent of moms request a C-section. It's not even well explained in medical malpractice or reimbursement. During periods where those policies have been the same or stable, this has continued to skyrocket. So 500 percent, we don't really know why. Now, one way to try to understand it is to try to look across time. The other way is to freeze time and just look across geography. And it turns out, this is probably really annoying for the people who are webcasting. I was going to go acapella a lot, sorry. But if you were to freeze time and just look across our country, it turns out C-section rates at the hospital level go from 7 percent to 70 percent. And David, I haven't found another health care service that varies by an order of magnitude in utilization at the hospital level. It's a big one. Like even like stents, joints, I don't know. And the thing is, you know, U of C sees sicker patients, right? Then some other place, I don't know. And so one way of accounting for that, I mean, this is even more awkward in my city of Boston. Like in Boston, there are so many hospitals for such a small city that if you are standing in the Boston Common and have a heart attack, if you fall one way, they'll take you to one hospital. If you fall the other way, they'll take you to the other hospital. But if you were to account for risk, which you can either do by risk adjusting or you can do by excluding high-risk women. It turns out that both methods are equivalent and it's easier to just look at low-risk women because most women are low-risk. So if you were to do that, you don't see 10-fold variation anymore. You see 15-fold variation, which makes it the only health care service that I'm aware of where after you account for risk, you see more variation, not less. It also means that in 2019, your biggest risk, by the way, this 500% thing, this is the most common surgery we perform on human beings. By far, nothing even comes close. About 120 million babies born per year in the world and one in four of them are born through a C-section, one in three in the US, one in two in Miami-Dade County. So the biggest risk factor for the most common surgery we perform is not a person's personal preferences or medical risks. It's literally which door they walk through, which in urban environments where there are options can be awkward. Like in my environment, it's like on one side of Brookline Avenue versus the other. But is the statistic that's stuck to me for hour of the floor, which was really surprising to me, is that over a third of women who come here are unregistered and have no behavioral care. And obviously people are much more concerned about them and they're already in deep and high risk, you know? So is there... when we compare ourselves to other countries which hopefully have better prenatal care and more investment in prevention, do you see a correlation between our lack of prenatal care and this trend versus other countries? That's a really good question. So the first thing is that we know very little about childbirth. And the best data that we have on quality of prenatal care is utilization of prenatal care and it's like the number of visits. And the whole care delivery model, by the way, doesn't make sense. Like, we've sort of arbitrarily decided that every woman over nine months needs 10 to 14 visits. Some need more, some need less, but our only metric of quality is just that utilization. And there have been some attempts to try to tie that to outcomes. Generally speaking, what makes our country stand out generally is access. And I think probably the last thing I'll say is that our field at the highest levels, at the New England Journal of Medicine editorial table, conflates high-risk pregnancy and high-risk labor. Turns out they're not the same thing. So if you have hypertension that's under-treated, you have obesity, whatever the case might be, you know, you can and should labor normally. It's the best thing for you. But in our epidemiologic studies, we often risk adjust for pregnancy risk factors when we're trying to, when the outcome, the right-hand variable is C-sections, and it actually makes literally no sense. I've had to gut-check that with several people just to make sure I'm not the crazy one. But high-risk pregnancy and high-risk labor are not the same thing. So that isn't to say that better prenatal care isn't a good idea or that it might help with some of this. But I think that the pathway by which it helps isn't necessarily, most of prenatal care isn't really about surveilling the fetus. So a couple of years ago, that stood out to me. In fact, there was the paper that pointed out the variation in C-section rates in health affairs that had no statistics in it. All they did was use the H-cup data to point out the level of variation, and it was the most red paper of 2013 in health affairs. But yeah, I just really wanted to know why. And when I met Vinny, I was still a resident of Brigham and Mass General, and these are two hospitals that are three miles apart that take care of the same population of people. And then I went over to Beth Israel, which again is like across the street, because it's Boston, to deliver babies. And basically I delivered babies at three different hospitals, but to me it could feel like a lot more work to take care of a very similar patient, depending on which hospital I was in. And I wanted to know if that insight could be relevant at all, and try to understand why the hospital you go to matters so much. Let me just say something before we even talk about that, about why it matters. So there are many examples of over-utilization throughout health care. This is by far the most stark in my opinion. And the thing is people can be hurt in two ways, when we do too little, too late, or when we do too much too soon. So most of my work actually over the last decade has been just pointing that out. The reason why we care about too many C-sections is that one, you can't get surgical complications without surgery. So the odds of things like a hemorrhage like what happened to Kira are three times more likely to happen with a C-section than without one. We also care about it because taking care of an infant, so hard. You're getting POW-level sleep deprivation, I'm still in it, with our infant, you're trying to earn a living wage at the same time. And doing that while you're healing from a 10-centimeter, 11-centimeter incision you're abduing is harder than doing it without one. But, you know, to connect it back to maternal mortality, obstetricians are the only surgeons that cut on the same scar over and over again. Like no other surgeon does that. Like if you're a neurosurgeon or a vascular surgeon or a trauma surgeon, and you have to go back and operate in a place that you operated before, it's like a week, but for me that's like a Wednesday. And the reason that matters is like you can train a brand new intern on how to do a C-section in like a week. But the second time there's a lot of scar tissue and it's a little more technical. And the third time, sometimes even the second time, it's like operating on a melted box of crayons. And sometimes the placenta, which is an organ that only exists in pregnancy, and is a big bag of blood vessels that gets 25% of cardiac output, gets caught up in a lot of scar tissue and doesn't detach normally, and women bleed sometimes to death. In fact, that condition, which is called placenta acreta, has become 1200% more common as a result of the increase that I showed you. We don't know causally what's going on with maternal mortality. We've got a bunch of theories, not the fact that we're not doing enough, but it's actually that we're doing too much, at least in terms of the hospital-based problem. A third of it is hospital-based, two-thirds of it are probably not hospital-based. All right, so back to this idea of like why hospitals are so frickin' different. I didn't really know how to explain why it was more work for me to take care of people in one hospital compared to the other. Does that sound plausible to you guys, by the way? Did you take care of people in more than one hospital? So I went over to Harvard Business School. Business schools exist for a few reasons. One is that they're a great cash cow for the university. The other is that they have people who are academics there who study why different places perform differently. So I thought that would be a good place and they helped me make a cartoon, which is to think of our labor floor as a pressure tank. Every pressure tank has a certain capacity. And on our labor floor, we think of capacity as, I don't know, like the number of beds that you have for women to labor in. It turns out there are no rules for how many labor and delivery beds you need based on the volume of people you take care of. There are no rules. So I found two hospitals on the West Coast that have the same number of labor and delivery rooms. One place does twice as many deliveries per year as the other, and they both think that they're at their capacity limit. The only way that's possible, according to Isaac Newton, is if one place is moving people through faster. In fact, we actually did a study, and that was the reason I came. That's the last time I was at UFC with the architects. What we did is we figured out that every public use facility has a fire escape map on the wall, which is a floor plan, which tells you how many labor and delivery rooms you have, because labor and delivery rooms are always on the periphery with the windows, and triage is always not. It's a little bit noisy, but it was good enough. We could have people take pictures of their fire escape map and send it to us, and we could estimate how many rooms they had for labor, and we were able to predict their c-section rates, which was cute and terrifying. So that's the capacity. Very different from place to place and arbitrarily so. Then, you know, every pressure tank has a certain amount of work load in it, a certain amount of pressure, gas, whatever it is, and it turns out like this is also arbitrarily different. So physicians usually don't measure our work load formally, we just suck it up. Although I guess in hospital medicine services, you can cap. That was like a mind-blowing concept to me when I rotated through medicine. Yeah. Not here. You can't cap, okay. But nurses very closely track work load, because it turns out 80% of the cost of running any unit is nursing staff. And it turns out the primary way that you measure nursing work load is midnight census, which is probably fine on a med-surge word and makes literally no sense in a labor and delivery unit. Because at midnight, your census can be zero, and at 1 a.m., your census could be like five. Inactive labor. So it turns out there's not good ways of measuring work load. And then all this sort of filters down to this dichotomous choice that we have in healthcare all the time between doing the right thing and the easy thing. And so often they don't line up. Right? Like if you're trying to admit somebody in the ED or not, you can either try to coordinate their care and send them back out into the wild. Or you can just admit them to a Vinny. And that's really the design challenge that all of us have, is like how do we do things that don't just bludgeon people at the front line and try to fix healthcare by trying to get them to do more stuff, but actually make it easier to do the right thing. Because David was here, and I felt the need to have some kind of data on a slide, this is unpublished, but this is from a paper that we did with David Cutler and to basically try to prove this pressure tank hypothesis. And we had a bunch of ways of trying to get at the idea, like from the architecture to actually studying the management systems, but it took a really clever economics graduate student, I think, to... So what we did is on the X axis is a number of women that are simultaneously laboring as any given index woman. And it says overlapping women, which I know is not a great term, but that came from basically the literature on overlapping surgeries. So a couple of years ago, there was a Boston Globe spotlight article on the fact that surgeons at MGH were operating on two people at the same time. And, you know, it wasn't a good luck. And then there were a bunch of papers that followed in JAMA about overlapping surgeries and the safety concerns, so that's why, the terminology. But she basically figured out a really, really clever way based on the time of delivery and the time of admission to count how many women were simultaneously laboring as any index woman. And then we just looked at the relationship between that and almost every other outcome that you might care about, adjusted for all the things you might want to adjust for to really isolate the effect of busyness. And what we found is that more women that are laboring at the same time as you, the higher the odds that you're going to get an intervention that's designed to accelerate labor or truncate it. And that it goes up and then it asymptotes and then kind of falls away, which does kind of indicates that maybe at some point you recruit more resources, you call in additional staff that you have to pay time and a half, but you're definitely doing it too late. So that's just the intervention rates. And then... This is in one institution. This is, oh, I'm sorry. This was in Philadelphia over a decade. And the reason that we used Philadelphia originally, although it didn't end up being part of this paper, is because a bunch of hospitals closed in Philadelphia in the 90s, and the population didn't change. So the remaining hospitals got way busier in absorbing all that volume. We should check up this. There's a very similar literature about hospitalist workload. One of the key things that's tricky in this sort of analysis is that the number of overlapping women at a point in time is actually a choice variable. Not randomly generated. It's generated because people have a certain part of the cycle. So the more conservative someone is, the longer they wait, the more overlapping women they're likely to be. And that actually goes against so that, in fact, if you adjusted for that, your hypothesis would be even stronger. I think that's absolutely true. There's also just something endogenous about the fact that the longer your labor is, the higher the odds. So we also created types of overlaps and different ways of constructing the bins to account for that. But that's also a really good point. It's not that spontaneous. One hundred percent. I've got a picture that illustrates what it looks like on our labor floors. 25-year-old literature in physician variation in C-section rates. They're okay. But yeah, I mean, essentially what we found is it's not just procedure rates. It's actually meaningful quality measures, too. And that the difference between being on a quiet labor floor, basically, and a busy one is the same as aging five years, in terms of your risk of hemorrhage and infection and other things that we care about. Here's my picture of what the cognitive problem looks like that Vinny was pointing out. So a couple years ago, I visited McGee Hospital in Pittsburgh. Most of you were born in a hospital. It does about 500 or 1,000 deliveries a year. McGee does over 10,000. Most labor and delivery units, except for the one that was designed in the beautiful new hospital here, are in some retrofitted space, like where the old labor floor was. And some have forgotten about corner of the labor floor. And McGee is a stand-alone free hospital for women, and it's got an endowment, and there are statues of pregnant women in the lobby. It's just like a different place. There's also not a lot around Pittsburgh, so they take care of pretty sick women. They have a wide catchment area, and they have a relatively low C-section rate. So a few years ago, I just showed up there and called first, and then I showed up. And I walked by their nursing station and took a picture of Vivian, who I'm 100% convinced has the hardest job in healthcare by far because of the cognitive load. So Vivian is a nurse in charge of the unit for that shift that day. She's got two computer screens in front of her that are helping her figure out her nurse staffing assignments, so which nurse goes to which patient. And she's got four more giant computer screens that are about that size that are her bed management system to figure out which patient goes to which bed. Then she's got eight more screens that each have four fetal heart tracings on them. And other people have to watch those, too, but they're like, you know, Vivian, while you're hanging out there, you may as well pay attention. And then, the best part is that dry erase board. Do you guys see that? That has 15 columns and 30 rows. And that's where she manually reproduces most of the information on the computer because she doesn't like the way it's displayed. And we had an MIT student like formally calculate that this is an NP hard problem, which what that means is Vivian's job would crash the world's fastest supercomputer based on the number of future states she has to anticipate and the number of degrees of freedom that she has, which made Vivian happy but she'd terrify us. And there are two things that I noticed in watching Vivian do her job. The first is that she's an air traffic controller. Without an air traffic controller's supports or tools, we actually went to the air traffic control tower at Logan airport to get a head on hand comparison. And the first thing that you'll notice in an air traffic control tower is that you would never see one person managing this much complexity. The other thing I realize is that it turns out labor and delivery units are ICUs, which you know, might be obvious or not, but the definition of an ICU is nothing to do with the presence of a ventilator. For like a health services researcher, it's just the ability to staff one nurse to one patient. So the cardiac ICU does that, so does the labor floor. The cardiac ICU can titrate medicine on a minute-to-minute basis. We do that for nearly everybody with oxytocin. The cardiac ICU can monitor vital signs in real time with telemetry. That's what all those monitors are. The only difference between the ICU and the labor floor is that our operating rooms are actually attached. Which means we've got the most intense treatment environment of the entire hospital for what are fundamentally the healthiest people. So then the question is, like, how on earth do you unwind that kind of complexity? So, I don't know. But we're trying something. Which is to aim to design a solution that has simplicity as a design principle. And it's not simple. Simplicity is, you know, what it is is taking the wide option set and trying to just put it into categories. And it starts with just having basic goal clarity. Which I don't think that we have right now. If you were to ask most people who are taking care of women in labor what the goals of care are, they would probably say safe baby, safe mom. And then they would drop the mic. And these are good goals, but they are insufficient goals. The reason is if they're your only goals they create a false choice between the mom and the baby where it often doesn't exist. In most cases it's a false choice. There are some cases around peri-viability and other things that we get into in our field, but in most cases it's a false choice. The other reason is that most moms have goals in labor other than emerging unscathed from the process. And honestly, like, survival is the floor of what women deserve. If we're trying to design a better system we should be aiming for the ceiling. Which is care that's not just safe, but also supportive and empowering. It's like a moment of identity formation for people. Like, if I'm in the dog park and I'm wearing my OBGYN fleece, people tell me their birth stories. Like, simultaneously, you know. And then, of course, like, this is one of the only windows in the delivery of healthcare services where you can influence the long-term while being a two human beings. Right? The melted box of crayons is six years away from the index surgery that we do. And it's often out of our line of sight. Because the average insurance carrier doesn't hold you that long. Like, they don't... We don't see it in our data. And then, you know, the other sort of principle is that, you know, as Dr. Siegler and I were talking about, childbirth is a team sport for homo sapiens. Always has been. Gorillas deliver their own babies. This is your cocktail party fact. So, gorillas are born OP, which is oxypit posterior. They're born facing their moms and they deliver their own babies. Human beings walk upright, which means we have a narrower pelvis. It also means that we have dexterity, though, because we have our hands back. And compared to gorillas, we have larger frontal lobes, which gives us our social intelligence. Maybe that helps. But that's... It's a team sport. And the thing is, at 3 a.m., when you have calls in the second stage of labor, you have to be able to trust experienced clinicians to use their judgment, because there's no technology that's going to tell you what's going to happen. But ideally, they'd be doing it with all the information that should be available to them in the room. And it turns out that information lives in the brain of multiple people. It lives in the brain of the mom, who can tell you things that nobody else can, not just her symptoms and preferences, but things like her energy to push. Like, Julie, my wife, was very clear about that. And then the nurse, who spends more time at the bedside than anybody else. You know, and... And the thing is, the team that comes together to take care of women in labor comes together randomly for every person, every time, and has to become high-performing from the most important moments in our lives. Um... And essentially, like, the way that you get a high-performing team is by creating psychological safety. Right? So that everybody in the room has not just... in the opportunity to share their information. Um... And that, ideally, that information is structured in some way. So this is our, like, brilliant idea. There's no blockchain involved. There's no AI. There's no widget of any kind. It turns out, in every labor and delivery room, in every patient room of this hospital, probably, there's a whiteboard somewhere. It's usually small in the corner, mostly for the nurse to talk to herself in highly variable and quality of content. And so what we did is we made a big one. And we put it opposite the mom's head wall and we simplified it. So there's a place where you write down the name of every member of the team starting with the mom who's, like, you know, at the head of the team. There's a place where you write down only things the mom can tell you. There's a place where you write down the plan. And then there's a place where you write down the next time that the team is going to come together. And we don't prescribe when that happens. Like, the team just decides. And it can be based on time. It can be based on events. It can be based on when you think you might want an epidural. It can be a window, whatever. But what it does is it makes it so moms in labor don't feel the way that I felt this morning when my plane was held on the tarmac and the pilot never came overhead to tell us what was happening. And that's the experience of most moms and their families in labor right now. So it's super simple. That's, like, the entire thing. But we're not testing it. Right now we're testing it among tens of thousands of families in Oklahoma and Washington State in Massachusetts. But we're not doing a giant quality collaborative to do this. We're not doing an RCT. The thing with quality collaboratives is, like, you sort of put something out there and you do, like, a monthly phone call for an hour. And then at the end of the year everybody takes credit for the weather. We're not, depending on how it comes out. And, you know, the challenge with the RCT is the first step is that we have many things that we know are effective and very few things that we've proven are doable. So we're trying to take the same approach that I would expect Merck or Novartis to take if they were deploying a new drug or device. Which is to do the Phase 1 FDA trial first. And just prove doability. So we're spending the same amount of money that probably more that you'd expect for a large RCT on just a small handful of hospitals and having human-being observers on the ground and collecting lots and lots of survey data from clinicians and patients about their experience. And then we'll do an effectiveness trial. And one of the things, actually, that we're noticing in our preliminary data is that, you know, they're going up in trend, near misses are going down. But we're putting some dignity back into the process because both for clinicians and families they feel more aligned. And a lot of families feel like they have a better understanding of, like, just what's going on. When most moms in the real world tell me they had a C-section and then they tell me why, I'm usually hearing a whole separate narrative in my head. Like, they're like, oh, the baby's life was saved because the cord was wrapped around his neck and, like, I deliver babies vaginally with the cord wrapped around their neck every day. Most moms actually don't know their birth story. And that's just tragic. So we seem to be doing something there, too. So I just want to end. I know I'm a little bit over time. And to say I'm conscious of the fact that this idea may not be the end all be all. And that even if it is, like, driving it to scale requires much more than proving that it's doable or effective or efficacious, that it's a product that has no market outside of a few bleeding heart philanthropists or funding the work. And that a market requires demand. And the hard thing about demand is that moms and their families have to know what they deserve. And the thing is in every part of the world I've been in seeing how people are born, whatever the status quo is is normal. People will procreate whether you give them a safe and dignified way of doing it or not. So here in the United States, it turns out we spend 0.6% of our entire GDP on hospitalizing people during childbirth. Like, if you were to take our whole $20 trillion GDP and spread it out across the table, it's like the Great Wall of China. You can just see hospitalizations for childbirth alone. And for that spending, one in three people get cut open and one in ten of their babies goes to the NICU. But that's, like, normal. So anyway, I've been thinking about this demand problem. I just want to talk about that really briefly, which is, you know, if you have a meeting on Capitol Hill about breast cancer, you can, like, push a button and then your survivors will show up. For maternal health, that does not exist right now. I don't know why. Part of the reason is that moms are used to putting themselves last to put their families first. Moms are busy. But there's moms against drunk driving. There's moms against guns. Moms are very good at advocating for every progressive cause, except for their own well-being. So about two years ago, there was an election and a transition and nobody was watching the National Park Service. So we figured out how to get a permit to use the National Mall on Mother's Day. And it was a fake it till you make it situation. We then rented a stage for $100,000 before we had the money to pay for it. But it was 50 feet tall and it was fit for, like, a kiss reunion. In fact, it took four days to set up and come by and be like, who's playing? And we'd be like, Beyonce at 1pm. And we just used it as a platform to let families like Charles that's how we're, you know, the whole thing with Charles kind of started. Ended up landing him on the Senate floor and getting a federal bill passed. So we did it in 2017. We did it in 2018 and we're doing it again. We learned over time that moms are busy on Mother's Day, so it's probably not the best day to do it. So this year we're doing it on Saturday and we just hired our first executive director and the whole idea is that to drive demand and to create a marketplace for this kind of work. I'll just end with my favorite definition of a designer. Designer is anyone who divides his courses of action aimed at changing existing situations into preferred ones. I love that because it's how I think about my job. I don't really know what I am. I'm an armchair economist. I'm not a real one. I'm not even like a real ethicist. But I'm definitely a designer. And it's how I think of all of you too. So thanks so much. We're redesigning labor and delivery and the institutional context for it. So, you know, it's obviously a business and there's a hospital president who has colleagues who have professional fees and things like that and their own incentives and concerns. So could you talk a little bit about sort of what this looks like on the ground and how those interests play into your redesign process? Should I assume you're doing at the brand? Yeah, so we're working we're working locally but we're actually doing a lot of work with community hospitals because it's where most Americans are born and, you know, in testing the feasibility we want it to be in places like Tulsa, Oklahoma where not only are the resources really different but the docs aren't even in-house and there's a whole different set of constraints they haven't expanded Medicaid yet but it looks like they're about to actually every superlative that you can apply is in Oklahoma second highest uninsured rate second highest teen pregnancy rate only behind like Texas or Arkansas depending on which statistic you're drawing so I guess in context they met the business so for an institution both of the lines matter the revenue and the cost and generally speaking labor and delivery is a low margin service it depends on volume to get positive margin but actually most hospitals that deliver babies run in the red and that's a bigger problem about how we think about investing in this enterprise but it's the reason why especially in states that didn't expand Medicaid but even ones that did if you're low volume and you're getting Medicaid reimbursement waiting around for things to happen is really really expensive and so we've seen a 15% decrease in the number of rural hospitals that have OB services there's rural hospital closures across the board but the first service that you shutter is OB because of the way the costs are structured and so there's opportunities on the payment side I mean for the professional fee for the individual clinician that you're trying to incentivize usually for them at this point in most settings they're getting paid the same amount irrespective of mode of delivery so it's not like there's a greedy doctor that's trying to boost their margin the institution though is collecting a much higher facility fee by about 50% that's reflected in the design because it ends up setting up a bunch of throughput pressures like that pressure tank I don't know if that answers your question there's many ways to answer it but definitely the very top heavy cost structures that are very much tied to the ICU level yeah I mean it sounds like your institution wouldn't necessarily want you to do this no it doesn't make me super popular but in some ways I think what's made it easier is that because Medicare doesn't pay for childbirth nobody's paid attention honestly for a very long time that's why it's kind of like this vast wasteland and now people are starting to pay attention to Medicaid and now purchasers are starting to get much more active which is why I've been spending a lot of time with them in the commercially insured population this is your number one spend and a primary focus for you is your reproductive age workforce because you're this is a benefit so that's been really helpful like where I am in Seattle the reason I'm in the Puget Sound it's going online it's okay Amazon and Microsoft collectively pay for the plurality of all the births there and I'm working at two community hospitals Evergreen and Overlake which are six miles apart have very different C-section rates and those two entities can move their volume however they want to and will if you could please comment on how the existing situation in the U.S. is related to legal liability I just learned I just learned this weekend that in Illinois and I'm not sure that supplies to other states one's insurance for labor and delivery physicians must cover you not only the year that you bought it but for 18 years after you bought it 18 I mean it's a staggering number I'm just curious as to how the legal situation may be affecting the the U.S. model I take new interns through difficult cases in July I've been sued myself I think about this a lot I also as a researcher in OB dig out of a giant hole to be to have the privilege of thinking because my overhead for medical malpractice is so high so I fully sympathize with that that being said it's more complex on both a policy level and an individual decision making level on the policy level very little of this trend is explained by medical malpractice policy by premium rates, damage caps or any of those other things it's not even very well explained by the experience the exposure of having been sued at the individual level I think that there is a very very low risk tolerance and very high risk aversion but it's not just the fact that there's an 18 year statute of limitations it's that you just want to do your goal in your mind is survival right because I can always the thing about C-sections is if I do a C-section the baby comes out pink and is swimming around looking great I think it's a good thing to do C-section and if the baby comes out looking blue and floppy I think well it's a good thing to do C-section it's pretty good to be me I'm always right there's never a counterfactual that's really the main problem thanks for your very thoughtful presentation a lot of what you focused on it seems is around workforce change the way in which an operation works within the hospital obviously a very big hill to climb in terms of making change and you pointed out many of the reasons why are there factors that can change for after delivery since so much of the death and morbidity and mortality is therefore happening after the baby's born not before the baby's born thank you so much for asking that question so there's a public narrative about maternal mortality that maybe I'm perpetuating that it's women who come in in labor and then hemorrhage and that does happen but two thirds of the deaths happen in the months surrounding the childbirth event and if you're on the margins of wellness and illness or even just basic security and insecurity having a baby in our country has become so socially isolating that it actually has become mortally dangerous and that's actually the bigger problem I was part of the CDC working group that put out the report trying to explain this trend through case study and root cause analysis which is all we really had but we could attribute 6 out of the 10 preventable deaths to some kind of basic failing of social support one of the ideas that we have there is we're starting a new initiative that is going to use cities so if you're the mayor of a city you don't want to be the best city to retire in that list exists in travel and leisure magazine but that's not really your goal you want to retain and recruit young families it turns out for a lot of cities you're also one of the largest purchasers of healthcare like in New York you purchase healthcare on behalf of 50% of the people who live in the five boroughs and also if in New York you're not three to four times more likely to die if you're black in childbirth you're 12 times more likely to die because of redlining and a few concentrated areas in Brooklyn so the idea there is that we're going to create a wave the performance of cities and supporting the well-being of families after they leave and the idea there is to identify the social services that already exist so that central planners whether they're cities, employers, parishes, whatever can stitch them together and gap fill I don't know if it'll work but your question is right on that yes, a big part of the problem is social support in the community needs to be addressed So are there behavior changes that obstetricians family medicine providers can do themselves so that it's not just this large system change but rather are there individual behaviors that those providers can learn and apply so that they can impact the morbidity and mortality absolutely one concrete way I was really humbled to learn recently that nothing that I do to manage the blood sugar of a pregnant woman with diabetes really matters during her pregnancy for her well-being if I don't ensure that she's connected with a PCP afterwards the single most influential thing that I can do is just at her six-week post-partum visit make sure she's connected with her PCP and that's actually not a thing that we do routinely I'd say in the majority of cases even in my own practice I have rarely done that there's also a huge opportunity to re-establish just trust in the profession and in our institutions like right now the floor is falling out underneath the medical profession around trust and it partially is a product of the fact that this is the least affordable healthcare has ever been for the average American and a lot of people are feeling fleeced part of it is generational mistrust especially among the most vulnerable communities particularly around reproduction I don't know that's fine I want to ask you a question about your definition and then add another definition to it so first of all the definition of designer suggests it's the idea that you're creating new things but it says they're not just two things that they're better at at least I'll interpret preferred is better it could just mean people choose whatever reason so then another word is applicant and a designer is a little different than an applicant because a designer newly produces it can create crazy clothes and people can wear them or not but an applicant goes beyond that because they're trying to suggest that others in some way should do something and maybe we should direct tax dollars so as you think about the direction you've gone where you're designing it I wonder and in fact you're doing more than that you're advocating in some ways although it's not exactly clear to me what you're advocating in terms of action you're sort of advocating in terms of perhaps awareness of a problem although I suspect it leads into action I wonder what your perspective is you could talk a little bit about what you're doing to understand what you think needs to be done to understand whether these things that are designed are in fact preferred and how one then makes decisions about how to allocate once time in the context of the degree of information that you have now about what's really preferred and this is a decision we all struggle with as researchers because presumably we and clinicians we go into this hoping that we're going to make a difference and sometimes it's quite obvious maybe it's well established but quite often it's more subtle than that and so I'm wondering how you're thinking about it for yourself and your career and the organizations you care about in this point you are where you clearly hit on something that people care about and have the ability to articulate it really well but then the question is what's the highest use of your time is it spreading the word about the problem is it trying to produce data about the solutions like how do you think about it wow what a question so the way that I think about effort is definitely that there's a target and a portfolio of effort that's all aimed at the same thing even though it seems like I wear a lot of hats and do a lot of things in my brain it's really coherent and when it comes to this issue I think of a way of defining preferred for the product which is this care process innovation and there there's people who want to decrease C-section rates but everyone can agree that maximizing uncomplicated ventral deliveries is a good goal and that's an area where you get a ton of signal relative to noise because it's a degree of variation so that's one concrete thing and then this notion of trying to measure dignity the idea that patient experience is not customer satisfaction but press gain and age caps that's essentially what they're doing and I don't know the answer to that question yet but we're trying to figure out if there's a way of trying to get at that at the same time I sort of divide my brain into product development and market development and this is what private equity people do apparently that they do both simultaneously and the challenge of all design is that the end user doesn't know what they want even if you have to co-create because the people that you're creating on behalf of can tell you their existing pain and then they can tell you what alleviation of that pain is but they can't imagine the iPhone so there's a different metric which is trying to right now we're trying to build the button that you can push to get people to show up and talk about what their needs are and that's sort of the metric on that side is trying to create this pipeline of change agents which is the thing that Vinny taught me how to do that I'm trying to apply over here I want to thank you for an extraordinary talk and for coming to visit us today thanks so much