Atul Gawande on Priorities, Big and Small

Conversations with Tyler is produced by the Mercatus Center at George Mason University, bridging the gap between academic ideas and real-world problems. Learn more at mercatus.org. And for more conversations, including videos, transcripts, and upcoming dates, visit conversationswithtyler.com. I'm here up in Boston with Atul Gawande, and we're going to talk about health, healthcare, healthcare policy, and Atul Gawande himself. So let me start with a question about artificial intelligence. How far are we from having an AI that is capable of actually doing diagnosis to people? That is, they might speak into a Skype connection, something like Watson would hear what they say, and they would then diagnose the person well enough to that this would be a usable form of healthcare? Is that far, close?

Massively far. I think it is one of the hardest things. You want me to tell why?

Tell us why, yes.

Okay, so the diagnosis process, people imagine what it is, is that people come to you with a crisply defined problem. I have symptom 1, 2, 3. I have data to add to it, and now, give me the answer. And the reality is, first of all, people come to you often unable to explain what their problem is. I have pain. Where? Hmm, well, it's sort of here. And they'll point with a hand. Well, do you mean there under your rib cage, or is it, you mean in your chest? Or, you know, so you have this probing process that is part of it, and how they tell the story. And then there's also how the, their story had evolved over time, and they often have to put it in their words. So it's more of a narrative than it is a straight set of data. So that's problem one. When I think of IBM Watson put their AI on this problem, and it would never be the problem I would have put them on. The second part of it is that it changes over time, and you're adding data along the way. And you're integrating it with a little bit about your view of the understanding of the person and their likelihood to even say that something is a major symptom or not. There is no question that you can augment the human capability. But the idea that, you know, that pull out your phone and it will be, it would give you the diagnosis, it is still One of the hardest problems in reducing error in medicine is the fact that we still have a high rate of error, and the sources of the error have to do with the human being rather than the calculation.

But say you only get 15 minutes with your doctor, which is pretty common. And as you know, those conversations don't always run so well. People are intimidated. They forget the right question to ask. You could have 3 hours talking to something like Watson, Maybe 80% of the dialogue is nonsense, but at the end you apply machine learning. And keep in mind, the alternative now is that people use Google, which is in a sense the world's number one doctor, right? Mm-hmm.

Mm-hmm.

So AI only needs to be better than Google, which is already a form of AI. So in that sense, isn't it just around the corner that it would be a marginal improvement on what we have today?

Yeah, so one is the replacement question, right? So can I simply have something that will make the diagnosis? And lots of reasons why that's difficult. But now to augment the human capability, absolutely. There already is a pro— there are programs. One example is called Isabel, where if the clinician having elicited all of this information can simply put the observations into a list, it will allow them to recognize, okay, fine, you think it's problem— that what they have is diagnosis 1, but here are 8 others in rank order of consideration compared to the one that you think. And there have been plenty of studies, and it's been around for more than a decade without the need for AI. This is just crunching some basic data to begin with, that that can add real value. I think the puzzle of it is that you need that capability to integrate information coming from the person, interpret it, and be able to get it into these kinds of systems. And in many cases, people may be able to do some of that over time for themselves.

How worried are you about CRISPR? Imagine parents sitting down, trying to shape the children they will have using advanced technology of some kind or another. Will those parents on average make the right choices, improving choices? Will this be a disaster? What's your view?

So CRISPR, the great thing about this podcast is I'm going to presume we don't need to explain what gene editing is for this audience. You're right. But so there are plenty of reasons to be worried about CRISPR in my mind. The one you just named is not one of them. Just like in the days of, you know, the first in vitro fertilization and the, and, and baby Louise, who was the first test tube baby, and now what's going to happen and we're going to, we're gonna destroy humanity and all of those kinds of things. I don't think that's the case. Yes, there will be some forms of selection. We already have massive amounts of sex selection in the world, which is a problem.

And that is a problem.

Is a serious problem. That's very harmful, but has nothing to do with advanced technologies. It has to do with the basic capabilities of ultrasound and selective abortion going on all over the world. The issues with CRISPR have to do with the capacity to add other— first of all, there's unpredictable things that people will discover that you can try to do with gene editing. And then you couple it with capabilities, for example, CRISPR enables gene editing that is basically, is fairly fixed. But if you now can propagate that edited gene in a mosquito or in another organism, now you can have rapid cycle diffusion of those kinds of capabilities. And I don't think we've thought through that in the least.

But won't parents discriminate against having strange or weird or unusual children? That is, parents will be too risk-averse. So maybe a stranger person has a higher chance of being a genius. A lot of parents, they just want kids who are sort of 15% better than they are and more or less along the same dimensions.

There's a great article this week about George Church's narcolepsy in Stat News. And the danger to me isn't CRISPR, the danger is the larger culture that it goes into, of which CRISPR is just one of many capabilities. So George Church, Tyler Cowen, famous genomics pioneer, biotech researcher, who's had tremendous influence and has been one of the most creative forces in the field, has narcolepsy. And he attributes a lot of his insights and capabilities to the fact that he falls into a deep REM sleep at the drop of a hat and then wakes up— During seminars. During seminars.

The best time. Right, right.

He had to stop driving because it would happen during driving, but he would drop into these dream states and then wake up with unbelievable ideas. And there's lots of ways in which he is a unique and a parent might think aberrant person. And you wouldn't select for what he has brought to the world. And there are lots of neuroatypical phenotypes. The CRISPR capability is just another bullet in the holster that can be fired, but we are narrowing that neurotypical range in lots and lots of different directions, whether it's how we employ people, what kinds of options we put out there for people to have their aberrant thinking recognized and taken advantage of, to the medic— the ways we medicate and control people. And I think CRISPR, it actually will be exceedingly difficult to be able to pick many of these capabilities out, in part because they're multi-gene. These are not conditions that have to do, that are very often have to do with point mutations that you just adjust. They are interactions among many genes.

But you run big data and every generation the chance of outliers falls by 1%, and then over time it's going to make a big difference, right?

—That's right.

—It's not that you control the kid who's going to pop out of the womb in 9 months, but over time, it seems it will compound. It seems more and more patients are awake during surgery, right? —Yes. —Awake procedures. How is this changing the culture of the operating room?

—Oh, yeah. So it's been very interesting. One of the things I brought into surgery with our team has been running a checklist, just like a pilot's checklist in the operating room. And one of the things you realize is that when you have an awake patient in the operating room, they can be part of the team. And not just someone sitting there who is annoyingly awake, and so they are actually piping up to tell you they'd like to change the music you're playing. But also— That's what I would do. Right, but also that you're, so in neurosurgery, they're actively part of the operation because they're often, Speaking and telling you things so that you know if you're causing any difficulties. But what we've found, even in the kinds of operations I do, which are not brain operations, is that being able to ask them from the very beginning of the operation to be part of the team by saying, you know, are we on the same page about what your medical issues are, what your goals are for the operation, what we have to look out for, in addition to the usual safety things, but also, you know, what are the, what are the key things? Oh, you know, I tend to have a bad back and, and I don't want to lie in this position for too long. In many ways, having people awake can be far safer. The other thing though is that there are people who, when they're awake, don't handle it very well, and you have to be able to adjust on the fly, and people have to understand that you may need to.

Why do surgeons sometimes leave sponges behind in the bodies of patients who are being operated on?

You zeroed in on the first, one of my very first projects of creating intervention. Great, great, great. So we had done a case control study of this problem of surgeons leaving sponges inside people and got it published in the New England Journal, partly because of our whole method of going about solving this problem, which was, um, we studied 60 people who had sponges left inside them compared to 240 people at the same institution at the same time with the same operation who didn't have sponges left inside them. And it was an interesting problem to me because the, the rate of sponges being left inside people had been steady for 30 years. We had not made it go down at all. It was right around 1 in 1,000 to 1 in 3,000 or so operations a sponge would be left inside, and it would be a disaster. You would have to go back about two-thirds of the time people had become infected. One of the 60 was a patient who died because a, a small sponge had been left in their brain. And, uh, the usual way that we track these things is by counting everything at the beginning of the operation twice and everything at the end of an operation. And what we found was that, uh, It wasn't because people were skipping the steps along the way. The nurses were consistently counting everything at the beginning and counting everything at the end. So it wasn't a broken system. It was a fallible system. So it's like counting whether you have 52 cards in the deck and you have 51, or you have 53, at a predictable rate of getting the count wrong. And sure enough, about 1 in 1,500 to 1 in— 3,000 times, they would just get it wrong and not know it. So we, in that case, identified that we would need to create a technological solution. So it was, in the end, we came up with a barcoding approach. It's an interesting economic problem. It's $140, at the time we were doing this, it was $140 million sponge industry for which there was half a billion dollars spent per year on lawsuits for sponges being left inside. So from a pricing point of view, it worked out to about a penny a sponge. So you had to have a solution that could double the price to 2 cents or 3 cents, but you couldn't stick a— you had to increase the costs a lot if you're going to stick a microchip or an RFID system on there. So we just put barcoding on there, kept the costs pretty minimal, and it completely eliminated the problem because now you have teams scanning them in as they to do the operation at the beginning of the operation, and then scanning all of the sponges out. And the only fallibility is that people will ignore the machine. And those are— so there's the only times there have been sponges left after tens of millions of these kinds of cases.

Do your awake patients ever come in with a checklist and at the end say, hey, have you got all your sponges?

So I definitely have had I've not had people come in with a checklist, but invariably, and we actually talk to—

But they should, right?

Well, it's not that you want people to come in with their checklist, right? Because each place has a flow and a way that they do it, but what you want to know is, do you have, does the place you go to, do they use a checklist? Now they will always say that they do, because everybody does at some level, but the key is, does the checklist have them stop before the incision, not to check that you have the right person and the right side of the body, which is which is something everybody standardly does. But to actually make sure everybody in the room is introduced to one another and they've reviewed the goals of the operation and the key— your key medical issues as well as the key points of the operation like what equipment are needed. Those are the big parts that the teams skip but make a really big difference.

What are the biggest mistakes in how checklists are used? Keeping in mind most of the people who do or could use them, they're not people like you who've written books on checklists and who know many, many different things, but they're highly imperfect, fallible people who are making mistakes anyway. And if allowed to codify their mistakes in the form of a checklist, can it make things worse rather than better? What's your main worry about checklists?

Yeah, so basically we rolled it out in our initial trial in 8 cities around the world and found a 47% reduction in death when the teams took up the checklist. The main barrier since then has been that it either becomes a tick box effort or no one uses it at all. So it was designed to turn people's brain on by enabling a conversation about the critical events and concerns of the operation. And so what we see is a couple of things. We gathered, for example, we had people send in the current— the version of the checklist that they're using, and about 150 places sent them in. And part of what was interesting is almost 100% changed the checklist. You have to change it to fit into your environment. For example, in England, in Ireland, you put the patient to sleep outside the operating room and then you roll them in. So the checklist has to— the order of it has to change. So one concern is that if you just use it out of the box, you're not really using. You have to adjust it to your environment. The second concern— so looking at these 150, we'd see places that had turned our 19-item checklist the, the most extreme example, they turned it into an 81-item checklist. It was impossible to use. We'd specifically designed it to be something you could run through in 60 seconds or less at each, each pause point, so you weren't distracted from the main operation. And you could see that the administration in the hospital had got a hold of it, and they were using it to try to impose their ideas. And if essentially the clinical team was not the team that were designing and controlling the checklist. And so invariably, you look at that and you know that everybody is completely ignoring it and it's become just a tick box effort instead of an enabler of greater capability.

Let's say one of your kids further on is thinking of getting married and says, Dad, should I use a checklist? Yes or no? Yeah.

Well, I think every wedding planner—

this is the— Not for the ceremony, for the choice of spouse.

Oh, for the choice of spouse. So that's a very good question. I suppose the first question is, is it an arranged marriage or not an arranged marriage? Because if it's an arranged marriage, that's practically a checklist process, right? But, you know, love is— if it's a love marriage, love does not follow the checklist.

There are now so many papers out.

Although there are probably some mistakes I could put on a checklist that would be really smart to consider along the way. So you're leaning back toward the checklist.

I am. There's so many papers that you probably know all or most of them that appear to show low marginal values for health treatment on actual healthcare outcomes. So Amish people who don't get that much healthcare, they have good life expectancy. Christian Scientists. There's the old 1970s RAND randomized control trial study where free healthcare only made the poorest of people healthier. There's the Oregon Medicaid paper. There's a few papers when the doctors go on strike, more people don't die when they're sent to the hospital. What do you make of all of this strand of literature, and why does it seem to be such a weak connection between healthcare and health outcomes?

Yeah, so I think it depends on what you measure and over what period of time. So now there's been an accumulation of enough data about, for example, coverage expansions like the ACA, like the Massachusetts reforms, et cetera, that I think there's actually some good understanding of what there is. First of all, they consistently show that people get a range of services that they get increased access to: primary care, specialty care, medications, chronic illness care, emergency care. There are elements of it that are much more powerful than others, and so chronic illness care, primary care and mental healthcare seem to produce really substantial improvements in people's self-reported health, their physical health, and their mental health.

But self-reported health is tricky. You look at life expectancy. You look at the Amish. It's hard.

Well, I'm gonna say two things about that. Number one is, it's not obvious to me that that may not be a primary goal. So first of all, that your, your reported wellbeing about your physical and mental health is in fact when people come in to see clinicians what they're trying to optimize. The second part of it though is in terms of survival, then what you see is it takes time. So the Oregon Study, for example, was on average—

That's only 4 years, I think.

It was not even 4 years. It was under a year and a half for a small group of people and the death reduction was not significant, non-significant, but the amount of death reduction you'd have to achieve would have had to have been greater than 50% reduction in death, which is a massive level. In fact, it measured a 16% reduction in death, but it just wasn't significant. The subsequent studies that have looked over longer periods of time, they see two things. First of all, there are significant reductions in death. The estimates are from 1 in 300 to 1 in 800 people who gain coverage will per year have their life saved. Which is a high level of value translates into tens of thousands of lives. But second is that you wouldn't be surprised that people getting treatment for their HIV, for their cancer, for their chronic diseases like high blood pressure and diabetes and so on, it takes time to accumulate and it appears to be some years. So like the Massachusetts data where basically what they show is that the strength of the value gets higher as time goes on. There's a further underlying point, which is though, that we are not actually maximizing by any reasonable measure those outcomes. So the biggest killer in the country is high blood pressure. We have a third of adults have high blood pressure. 60% are receiving incomplete and inappropriate care within the medical system. And so they're not actually getting the biggest killer under control. And they don't follow instructions. Well, that's a mix of not necessarily getting the right treatment in the first place, not getting the right follow-up, and yes, it might not be not following instructions, but when we poke into why, well, there's copays and deductibles that can block out being on those medications, and then people are making choices. Some people are having side effects that lead them to stop it. Some people are just way disorganized need, you know, there's actually, CVS has been able to track why people stop medication. They have 6 different kinds of patterns with which people are not adherent to their medications, each of which actually have effective solutions to target. So, you know, the big point is high blood pressure is one of the simplest things that we can take on. We're terribly organized for being able to deliver on it. It's the biggest killer in the country. Kaiser, followed through on being able to take it instead of from only 40% appropriate care to greater than 80% and saw their entire population flip from the number one killer being cardiovascular disease. They also worked on improving statin use and flipped from it being number one to no longer being the number one killer. Cancer now is the number one killer in their population. Minnesota followed through on a similar program, got adherence up to 70% and made the same flip as well. And now we have several states where cardiovascular disease is no longer the number one killer by getting some of these basic measures under control. And that's just about healthcare being good, uh, but, uh, having really poor reliability given the amount of money we spend in healthcare.

What's the number one thing missing in medical education today for doctors?

I think the number one thing is an education around the fact that we are no longer a craft. It's no longer an individual craft of being the smartest, most experienced, and capable individual. It's a profession that has exceeded the capabilities of any individual to have, have to manage the volume of knowledge and skill required. So we are now delivering as groups of people, and knowing how to be an effective group how to solve problems when your group is not being effective and to enable that capability. That I think is not being taught. It's not being researched. It is the biggest opportunity to advance human health and we're not delivering on it.

The famous late 19th century, early 20th century Dr. William Osler, he once said, here's a quotation, look wise, say nothing, and grunt. Speech was given to conceal thought. True or false?

False, but there's a grain of truth. So one of the things that I learned in writing my last book, Being Mortal, was talking to palliative care clinicians and geriatricians and others who really have a different kind of approach to practice. They, examining the way I practice, what they would say I do is talk too much. And as one put it, "Your problem, Atul, is that you're an explainaholic." Our model of what a clinician does is that they are doctor informative. They give you data about what condition you have, data about the options. Here are the option A, option B, option C, risks, benefits, pros, cons, and now what would you like to do? We prov— it's a conversation where we do 95% of the talking and then the patient does 5%. And what they point out is that they, the clinicians of the future, really need to be oriented in a counselor mode where they are not just telling you what the options are but also eliciting from you very clearly what your goals are and then making a recommendation about what much most matches your goals. So what are your priorities? For your quality of life as well as quantity of life. So, you know, people have priorities besides mere survival. When we don't ask and don't know how to ask what those priorities are, the treatment is often mismatched with those priorities, and that's where you get suffering, and that's where you get lots of hot air from doctors, and you have total misalignment. When you are able to elicit those goals, and then align the care with it, you have massively better outcomes, both for quantity and quality of life.

Do you ever worry that doctors become corrupted by their own authority? Not corrupted in the sense of taking bribes, but they have such authority, people look up to them, they say, well, you know, take off your clothes, sit on the table, people do it. People are in the mode of obeying. And a lot of patients, frankly, are probably a pain. So doctors, in a sense, can manipulate patients to make each interaction easier because they're such high-status figures, and then doctors fall into this mode of being used to doing that with people in general. All right, what do you think are the main psychological dangers of corruption that can stem from being a doctor and having such high status? Control over life or death. You can cure people the gods cannot. That's what Sir Thomas Browne said, right?

So I think this, there is, I think that the health system of every country is a demonstration of our corruptibility. Kenneth Arrow's 1960s essay on asymmetry of information used healthcare as its prominent example that sellers are more powerful than buyers when we not only control the, the decision set, we control the option set. And the opportunity and ability of the individual to understand all of those, to be able to make choices among them, is hindered, especially when you are sick. So the most powerful, tool that a clinician has is their pen and is the power to order medications to tests to doing an operation. So yes, the power of the fact that, look, it's an unusual profession in that, yes, we can ask you to sit on the table and make yourself the most vulnerable in the most vulnerable way possible, not only to ask you to to take off your clothes, but then to actually have permission to cut you open and do what I choose to do inside you. That is a tremendous power, and the ways in which we see that just the payment incentives alone dramatically affect whether my tendency is to give you overtreatment in certain situations and undertreatment in others is a reflection of our failure to follow through on being able to deal with the variability in how we use that asymmetry of information.

A lot of critics have charged that to get a new drug through the FDA, it takes too many years and too much money, and that somehow the process should be liberalized. Do you agree or disagree?

So I generally disagree.— It's a trade-off in values at some basic level. So in the 1950s, we had no real FDA, and you had the opportunity to put out, to innovate in all kinds of ways, and that innovation capability gave us modern cardiac surgery and gave us steroids and antibiotics, but it also gave us frontal lobotomies, and it gave us the Tuskegee Experiment and a variety of other things. The process of asking that our— that we have regulation around both the ethics of what we're doing and that we have some safety process along the way is totally appropriate. Now, I think a lot of lessons about when the HIV community became involved in the FDA process to drive approaches that smoothed and sped up the decision-making process, and also got the public enough involved to be able to say, that community said, look, there are places where we're willing to take greater risks for the sake of speed. That is, that to me is, people are trying to treat the FDA process as a technical issue, when what it is, is it's an issue about what are the risks we are genuinely willing to take, and what are the risks that we're not? So—

But the big risks seem pretty invisible. So some drug you've never heard of, it doesn't even exist yet, you never get it and you die, that's not very salient to people. So it seems insofar as the public judges the risk, maybe they're getting it wrong.

No, that's right. The way to think about it though is that we have a certain speed with which we have accelerated passage through the FDA. We are doing surveillance afterwards and then finding that the approval process leads to drugs being withdrawn from the market. That is a decision that we have decided to accept acceleration, especially of drugs that have to be tested on small numbers of people. And then we'll monitor afterwards to be able to catch problems after they happen. And some of the disasters that have occurred, like Vioxx and others where there were substantial cardiac, effects happening to many, many thousands of people. We have, you know, there's been a backlash against that. On the whole, I think we have decided to trade that we're going to lead, cause some passage, allow, that we've decided to trade letting lots more drugs through in a little bit more quick process, but then try to have surveillance afterwards. Now, of course, one of the results is we're not investing very much in that surveillance. The worst of all worlds is speed drugs through and then close our eyes to what harm is being done after the drug has hit the market. And that, of course, is where all of the incentives are from the drug maker point of view. And then when the public's not paying attention, that's where we start dialing back. And that's one of my fears in the current regime is that Okay, fine, we speed up the approval process, but then gut the ability of the FDA to hold the medical community to having surveillance of what happens and then pull the drugs on the market that are actually causing harm.

New surgical procedures, should the FDA have a greater ability to regulate them than it has right now?

Yeah, so one of the really interesting things is that basically surgical procedures are not easily protocolized. Correct. And so we're still in that sort of craft world in surgery. And so as a surgeon, I can kind of do whatever operation I deem appropriate. And the only thing that's regulating me is the malpractice, the threat of being sued for malpractice and colleagues saying that what I did was way outside the standard of care and made absolutely no sense. The challenge is that that in a certain sense, operations are crafted for each person. And so the, the devices you use and the drugs you use are highly regulated, but the procedures you use are not. I don't think it makes sense for the FDA to regulate the procedure. However, I do think it makes sense for the organizational organization to be held accountable for tracking what kinds of outcomes people have for procedures. So the part that gets to me is that for a given procedure, proven procedures, we have at least 250% variability in the outcomes depending on the institution you go to. It is primarily a function of the institution, not the individual surgeon. And we have no tracking, virtually no tracking, certainly no transparency about what what those outcomes are and what the factors are going into them. And I think that's where we need to have much more information.

Now, there's a middle segment of all of these conversations, overrated versus underrated. You're free to pass on any of these, of course, but I'll toss out a few things and you tell me what you think. Underrated or overrated? Stevie Wonder. Hmm.

I think he's now become underrated. He might have been overrated at one point, but now he is significantly underrated. And what's his best album? Oh gosh, I'm gonna have to pass on that because I lived in the world of singles, like Sir Duke was my favorite song. Songs in the Key of Life, yeah. Yeah, well, so what I might go to is going back to that era, so Songs in the Key of Life, I would say bring that back.

Michael Crichton, doctor, Harvard, writer, what do you think?

Where is he now? I would say he might be I think he's been overrated. I say this as someone who was totally inspired by Michael Crichton, the fact that he graduated from medical school and then a year into his residency left to do everything from directing movies like Coma to writing The Great Train Robbery and then onward to Jurassic Park. But as a as a kind of influence on medical writing and on writing writ large. I think that the power— his power is waning. His power is now— well, now see, I'm switching from Overeit down to Underreit.

Now you're sounding like Underreit. Sphere is my favorite book of his. He was also a brilliant art collector. He sold a lot of Jasper Johns before other people.

And I just realized in the critical way he's underrated was in creating programs like ER, which showed how it really has become now that the medical procedural, just like the criminal procedural, is now a mainstay of television. And I think ER created that, and that was a brainchild of Michael Crichton. So yeah, I've reverted.

The idea of nudge.

I think overrated. Why? I think that there are important insights in nudge units and in that research capacity, but when you step back and say, what are the biggest problems in clinical behavior and delivery of healthcare, the nudges are focused on small solutions that have not demonstrated capacity for major scale. Nudge is something that— the kind of nudge capability is something we've built into the stuff we've done, whether it's checklists or coaching, but it's been only one. We've had to add other tools. You could not get to massive reductions in deaths in surgery or childbirth or massive improvements in end-of-life outcomes based on just those behavioral science insights alone. We've had to move to organizational insights and to We've had to piece together multiple kinds of layers of understanding in order to drive high volume change in healthcare delivery. Wearables. Wearables, I think underrated. Why? So wearables right now don't do terribly much, and I think there's a real downside that, you know, what we see right now are wearables that might, say, track your heart rate in a, and catch variations in your EKG and so on. What the problem is is that the wearables have not been able to be integrated into the practice of medicine in a really critical way. So right now, the way a wearable is used, whether it's for tracking cardiac events or your mental state or other things like that, is that then it says, notify your doctor, or it's a dump of a ton of data that a clinician is supposed to use and know how to integrate into practice. It hasn't been able to be used in such a way that you're actually demonstrating major improvements in people's outcomes. However, we are entering this phase where we are now starting to be able to track, take your genomic data, take your laboratory data, take your imaging scans you've done, couple it with information from wearables, like how you're doing over time, whether you, you're getting the medications you should be getting. Our team here is actually just deploying a version of a wearable, which is just turning on the sensors in your phone to track patients after surgery to see whether we can define what their— how long their recovery pathway is, how long before they are back in their normal sleep habits, are back mobilized and moving around much more. I think that ability to have knowledge of the wellbeing of people that goes beyond just whether you had a complication or a death, um, you know, success of what we do for people, uh, can be enormously, um, uh, improved with, uh, followed by wearables. And then that creates these incredible learning loops where we can, we can maximize those right now invisible indications of the outcomes of care. So I think the wearables are being used on the diagnostic side, where it's semi-powerful, but not, not critical, but as part of how you're doing in your care and whether we're helping you achieve the goals. Very powerful. Carl Knausgaard. Knausgaard. Overrated. I, I've read about half of the trilogy. I'm totally absorbed and loved, loved them. But, you know, I feel like Elena that Ferrante is an example of someone who was doing much the same, deeply mining the territory of somebody over a history of time, but managed to do it by fixing them in a place and a time and a history and a dramatic change in social conditions and everything else that took it to this whole nother level in my mind.

What makes you weep?

What makes me weep? You know what invariably makes me weep is when it might be a movie or might be just seeing someone tell the story of devoting years of time to a problem or to something they're trying to do at great pain, but sticking with it and then actually getting to the end. I think we undervalue— I'm going to give two examples. The movie Bliss, this Australian film based on Peter— oh my gosh, I'm blanking on the Australian writer's name. Peter Weier? No, Peter— did Oscar in this movie. Peter Carey. Peter Carey, yes. Has this moment where the man is in love with a woman and plants the seeds of a forest, and it takes the next decade or more before she realized this thing he built out of love, and it grew, and that's when she realized that in fact he loved her. And that was just extraordinary, and of course made me bawl like a baby. And then this coming week will be our surgery graduation, and it's 8 people who have given the last decade of their lives to trying to craft and become who they are. And they will tell the story. They each get 15, 20 minutes of time to tell their story of what they've gone through. And they have had families and had family tragedies, and they've had things that they've missed along the way. And it always makes me cry because those— that ability we have to commit ourselves to something larger than ourselves over a long period of time, not knowing whether it'll turn out the way you think it will, uh, that is, I think, the most incredible thing that human beings do.

You've written about yourself, uh, that you are highly indecisive by nature and that you will agonize over many choices. Is there any way in which you feel this, which you describe as your weakness, is actually your strength?

Um, yes. So I, I chose surgery— and I wrote in the passage you might be reflecting on that I chose Sergio in part because I was drawn to the character of people who had that kind of decisiveness, took action when inaction was the worst thing, and then, you know, lived with the consequences, took responsibility for whatever happened, learned from it, and then brought it to next time. I brought to surgery my particular persona, which is to want to have more planning and more deliberation inserted into situations where there doesn't seem to be time. And one of the really striking things is we get this complaint, for example, that using a checklist in surgery, well, that's fine, but in emergencies we don't have time for that kind of thing. And we demonstrated that in that situation, taking the 1.5 minutes for the very beginning part of it has enormous value. It's in fact the group in whom we have the highest reduction in death are those in which the time is the most pressing. And so I think I've flipped my weakness into a strength where I can, but, but by, you know, battling with it to a certain degree.

So are you a good healthcare patient?

I'm a terrible healthcare patient. What do you do wrong? So for starters, I don't see doctors. Okay. We share this. I don't go in hardly at all. And I haven't had to, you know, so I've crossed 50 and I have never had to really be in the hospital or have a major event. But to the extent that the minor conditions I've had, had addressed, I invariably come way late in the game when the mild infection has become a much more significant one. And it's partly 'cause I know too much about the ways in which healthcare really is pretty unreliable. I think a way in which I am good at being a patient is once I'm in the door, I pick my team. I'm really careful about picking my team and then turn myself over to their process. I'm trying to pick them for their ability to run a process that, that seems appropriate. And so I think the interesting thing about doctors is that they do tend to not micromanage their doctors very much.

Do you feel you've underachieved in life?

Oh, that's a hard question. Tyler Cowen. I mean, I know objectively that it's kind of ridiculous that I would think I've underachieved and that I'm proud of all the random things that I've been able to be part of, but I bear a kind of chronic dissatisfaction and sense that I've got much more to follow through on than I've managed to. So yeah, I think underachieved is the wrong word, but— and yet I don't feel I've achieved nearly enough, and that half of what I've achieved, I wish I could go back and fix.

Yeah, I'm of the view the Clinton-era welfare reforms were mostly a good idea, but they've come under more and more criticism lately, and you had a hand in that earlier in your career. What's your 2017 view of all that?

How has it gone? Well, I think the fact that that we have the number of people who are living on under— what is it? $4 a day in income and that that number has skyrocketed. That is unconscionable and a fundamental error and failure. That we have sought to— there are a variety of things that we sought not to punish people for being married, that we sought to not punish people for working, the shift to the Earned Income Tax Credit. There's a deep and fundamental— and maybe this is the Midwestern in me— belief that we need to reward work. It's one of my sort of core concerns is that the people who are at the 30th and 40th percentile are served by none of the major political agendas— I agree. —on the left and on the right. And that even Obamacare provided its great aid to those under the 20th percentile. But the people at the 40th percentile are not getting the subsidies, still don't have a system that works for them and actually had to pay more into the system along the way. The welfare failure that is my great concern is the number of people that are subsisting on horrendously little, but the value of taking away the things that were punishing people for moving in the right direction still feels like an important agenda.

There are many smart people in medicine, in surgery, in writing books for that matter and writing medical books, right? What would you say is the Atul Gawande production function? I asked Raj Chetty this. I said, Raj, you get a lot done. What's the skill you have that other people don't have? What would your answer be to that question?

I think it is to constantly— I think part of it is that I'm pretty clear about my goal, which is I'm trying to have impact and to try to do stuff that feels cool along the way. You know, my team knows that my mantra is I want to do cool stuff that lasts and let's see if we can do that. And so I think that from early on I was always iterating against that set of goals and that I more or less, I mean, I organize my life around it. I don't watch that much TV. I get enough sleep, but I'm— So you're decisive. Well, I've learned that I say yes too much, and so I've learned to say no a lot. Like saying yes to this podcast. Yes, sometimes I guess it's serving some part of that production function. There's a cool— it's cool. Will it last? We will have to test that question.

—over time. We don't talk about Obamacare too much. We want people 2 years from now to still consider the content fresh.

That's how I write my New Yorker pieces. I want it, I mean, my ideal is that my New Yorker pieces, I'm always thinking about, am I writing them in such a way that in 5 years they could still justifiably be read? And I hope would be that way while trying to address something that feels really sharp and of concern in the moment as well.

But those are your goals. That's your framework. What's the talent you have that, say, your competitors don't?

I really think it's just— when I get into our hospital settings, whether we're working in surgery or medicine or childbirth or other areas, invariably, the main thing we have to work to do is identify what are your priorities. And people in the system, number one, usually can't say what that is. And then if you observe what their priorities are, it's actually far away from delivering the best possible care to patients at the lowest possible cost. That actually is a very tiny portion of the healthcare industry that's trying to deliver that. And so in my— individual capacity. I think it's that I'm able to spend a really high percentage of my time on my priority. I think that's all it is.

Now, we're in Boston. We're at a place called Ariadne, which is well known, but maybe not to all of your readers. It's become one of your major projects. Could you tell us just very briefly, what is Ariadne?

Well, so it stems out of the recognition that It is a center for health system innovation.

It's a for-profit, nonprofit, part of Harvard, or how does it fit into the universe?

Yeah, so it is an academic center that is part of Harvard Chan School of Public Health and the Brigham and Women's Hospital, but it sits as a platform we've created for people who want to do experiments in how we bring science and innovation to the delivery of healthcare in the same way that you bring science and innovation to biological sciences. And so we have people come here from all of the different hospitals in the area, from the business school, from the School of Public Health, from economics department, and then really built it as a place to enable capabilities to do large experiments like the experiment we just finished with the state of South Carolina, 5 years of working with them to see whether you could get the surgeons in state to adopt the safe surgery checklist without regulation, because it's a red state and they didn't want to mandate it. But without financial incentives because they didn't have the money to pay for pay for performance, and we got a program that successfully got 40% of the population through adoption, lowered the death rate 22%, but to get the other 60% on board, we believe you would have to bring on some mandate or other component. So that's kind of the thinking. I'm the director here. We launched, it'll be 5 years this year. And so we launched in 2012. We now are running about 20 projects in surgery, childbirth, and end-of-life care, improving how you come into the world. The surgery— you, the average American has 8 operations in their lifetime. It's the highest risk, highest cost, highest failure moment in your lifetime. And then how you leave the world, end-of-life care. And about half of our experiments are in the United States and half are abroad for demonstrating ways to get better and better.

So there's a lot of people here. It looks great. If you're director, you play some role in hiring. We all want people who are smart, have experience, dedicated, cooperative, diverse, and so on. But what's the quality you look for that you think other people doing hiring are undervaluing? What's your entrepreneurial secret to hiring well? Yeah, so I think—

Again, it's that thing like, what is your goal, right? What is your priority? Yeah, so the secret to the hiring is actually before you ever meet anybody, know what the heck you're trying to hire. So the hardest part sometimes to get people to understand is what is it you are asking them to do in 2 years? Give me the list, their scorecard. What are the 5 things that they will have accomplished? And then make sure you show it to them. When they come in the door and ask yourself as you are interviewing them and then assessing who you're going to pick, not do I like this person or are they, you know, would they be fun to be around or all of those kinds of things, though those are factors. The ultimate question is, will they be successful in this list of things that you said would define their success in 2 years? And being again faithful, that is, maybe it is my focus down. Like, you know, intend, do what you intend to do and do it with intention. So, and it is over and over, that's what people fail to do. You know, we've got up to now 90% consistency in delivering on people who, as we call them, A-players. And it's most of it is in that front end before they've even walked in the door. Being clear about that set of steps. And then it's that they don't tell you all of it, and it's how you talk to their references. That's the other real big mistake people make.

Last question. According to your chat with Ezra Klein, you're a big fan of what you might call contemporary independent music, but define it as you wish. What would be a musical group that you love or is meaningful to you that maybe the rest of us haven't heard about that you would recommend?

Oh gosh, so I think one of the most underrated bands is Frightened Rabbit. Do you know them out of Scotland?

No, tell us about Frightened Rabbit.

So Frightened Rabbit, they're sort of the bards of sorrow and nonetheless sticking it through. Their album Midnight Organ Fight is the great breakup album of all time, and you have to listen to it now. They're Scottish, so there's a whole lot of cussing going on. And they are, and you know, they're, it's painful and it's dark and all those kinds of things. But I always find like there's this glimmer of hope that runs through it despite their ability to sort of express pain and heartbreak and disappointment and being betrayed. While still nonetheless waking up and sticking it through the next day.

On that note, we close. Atul Gawande, it's been a pleasure. Thank you. Thank you. Thanks for listening to Conversations with Tyler. You can subscribe to the podcast in iTunes, Stitcher, or your favorite podcast app. And if you like this podcast, please consider rating it on iTunes iTunes and leaving a review. This helps other people find the show.