NIH CLINICAL CENTER GRAND ROUNDS
Episode 2009-005
Time: 01:03:13
Recorded Feb. 11, 2009
THE ART OF FAILURE IN MEDICINE
Dr. Atul Gawande
Surgeon, Brigham and Women's Hospital and Dana Farber Cancer Institute
Associate Professor, Department of Surgery, Harvard Medical School and
Department of Health Policy and Management, Harvard School of Medicine
ANNOUNCER: Discussing Outstanding Science of the Past, Present and Future – this is NIH Clinical Center Grand Rounds.
(Music establishes, goes under VO)
ANNOUNCER: Greetings and welcome to NIH Clinical Center Grand Rounds. Today's podcast is part of the NIH's Contemporary Clinical Medicine: Great Teachers series. On this edition, we'll hear from Dr. Atul Gawande, a
Surgeon at Brigham and Women's Hospital and Dana Farber Cancer Institute, as well as an Associate Professor, Department of Surgery, Harvard Medical School and Department of Health Policy and Management, Harvard School of Medicine. The topic is "The Art of Failure in Medicine." If you would like to see a close-captioned videocast of today's subject, log on to http://videocast.nih.gov and click the "Past Events" link. We take you to the Lipsett Ampitheater in the NIH Clinical Center in Bethesda, Maryland, where Dr. Paul Plotz, chief of the Arthritis and Rheumatism Branch at the National Institute of Arthritis and Musculoskelatal and Skin Diseases at the NIH, will introduce today's speaker.
(Music fades)
PLOTZ: It's a great pleasure to introduce today's great teacher, Dr. Atul Gawande, a man of many parts. He had the good fortune to begin his life in Brooklyn and to be raised in rural Ohio. I think certainly that might be considered a good fortune by some, too. I don't know. His mother was a pediatrician, and his father a urologist. His higher education was at Stanford, at Oxford, at Harvard Medical School, and then an internship and residency at the Brigham and Womens. Overlapping with getting an MPH at the Harvard School of Public Health, he got his boards in surgery, a scant 5 years ago. He is now associate professor of surgery at Harvard Medical School and at the Brigham. Sounds pretty straightforward, so far, doesn't it?
But, at the same time, he had a parallel career in political and health policy working for Al Gore and Jim Cooper in Tennessee and then serving as chief health and social policy advisor in the Clinton/Gore campaign. And then as deputy director for health policy in the Clinton transition team in 92/93, following which he was a senior advisor in the DHHS for a couple of years. Sometime along the way, he was a junior high school science tutor at Boston Latin School, a Rhodes Scholar, and he is currently a MacArthur Fellow.
Since 2000, he has been a best or finalist in non-fiction essays, crime writing, science writing, six times, and numerous other writing awards and important speaking occasions, lectures, grand rounds, beyond count and then an astonishing variety of topics. I began to worry about his sleep as I was writing at this point and he is currently a medical sage. He has talked about poverty, medical error, medical failure, about surgical error and failure and reducing it by a check list of very important papers. The character of a doctor, the future of human life, sounding very much like William Osler to me. The imperfect science of medicine, the doctor's writer, curiosity or on questions cannot be answered with a randomized clinical trial. And “Excellent Execution”, why physicians participate in the lethal injection of prisoners. I’m not done yet.
He has of course won numerous teaching awards. He has written widely for the general public for over a decade, which is how many of us here came to know you, Dr. Gawande. Here, he began like Chekov in a small way with small articles on, the Persian Gulf Syndrome, castration, weight loss, food eradiation, zinc for colds, and then a wonderful one -- when a full moon and a lunar eclipse collide with Friday the 13th, do more accidents really happen? Many of these articles on Slate. His New Yorker career began in 1999 and has led to many interesting pieces and he has written two celebrated books, “Complications, a Surgeon's Notes on an Imperfect Science,” and “Better: A Surgeons Notes on Performance.” I have been reliably informed that these are hotties among premed students.
You know that every time Michael Jordan went in for a lay up, it was new and somehow different and usually dazzling. And I think that Atul Gawande may be a medical Michael Jordan. Dr. Gawande.
[applause]
GAWANDE: I shouldn't even give this talk. [laughter] Thank you. And I’m embarrassed to be here even under the label, “Great Teacher”. I do feel that I’m a young doctor and I think we should flash up our conflict of interests here on talking about our -- it would be nice to have a conflict in interest in talking about this. But the chance to talk to you about a perspective that has come from being what I hope is still a relatively young doctor, I can only claim it for a couple of more years, I think, trying to work my way through medicine and thinking about some of the larger issues that we all work on. I want to have a chance then to just start by telling you a story.
Sometime ago, I saw an arthritic 86-year-old woman who our emergency room was complaining of a searing back pain, began in the front of her belly and went straight through into her back. And I learned in talking to her about her history that she found out she had a abdominal aortic aneurysm that had been diagnosed a month or so before and she can't quite gotten attention to. And when someone tells you they have a severe abdominal pain going into their back and they just had a diagnosis of abominable aortic aneurysm, your alarm bells shriek. So I could feel the aneurysm throbbing under my fingertips and under her abdominal muscles. Even more alarmingly, it was tender. Now she was stable. But this adds up to a classic clinical syndrome. You have a throbbing aorta. It is tender, it felt larger now than six centimeters and she has a known diagnosis of abdominal aortic aneurysm. I knew we had a situation where she was on the verge of rupturing -- an impending aneurysm rupture. I called in our vascular surgeon as a specialist and he agreed with the assessment. So we told the woman that she was going to need immediate surgery and that was going to be the only option to save her. We warned her, however, that this was going to be a big operation with a long recovery in intensive care, and probably a nursing home for her afterwards. This is a woman who told me she still lived independently.
Looking at the scans, that we pulled up from her images she had before, her aneurysm extended above the take off of the renal arteries, which meant that she would have to have a repair that would unplug the renal arteries and then plug it back in and under conditions of impending rupture, that had a high risk to go into renal failure. And knowing the statistics of people who come before, she had somewhere around minimum 20% chance of death going into this. And then we asked her, so what would you like to do? And she didn't know what to do. We left her with her family to think on the decision. I gave her 15 minutes. And then I came back in. She was with her son. They had talked it over. And when I walked in the door, he was in tears because she decided she would not go ahead with the surgery. She just wanted to go home, she told me. She lived a long life, she said. Her health had long been failing. She had drawn up her will and she had already counting her days in coffee spoons. Her family was devastated but she was remarkably steady voiced, constant about this decision she had come to.
And then I did something I had never done before. And I have not done since. I saw this woman with a terrible life-threatening problem and I wrote out a prescription for some Percocet, handed her the prescription and watched her walk out the emergency room door on the arm of her son, understanding full well that she would die.
I kept her son's number and when a couple of weeks passed, I gave him a call just to see how he weathered the aftermath. And his mother answered the phone. I stammered, hello? And I asked her how she was doing. She said she was doing very well, thank you, how are you? [laughter] A year later I learned she was still alive and well and living on her own.
The work I do, I think has arisen out of the intensity of this kind of experience in training and surgery and then in practice in surgery. And also now trying to do work in public health as well. And I am constantly reminded over and over again in the details of what the experiences actually are like when you get far away from what the studies say it is like. That medicine is a strange and remarkable and still sometimes disturbing business. And so writing and research became a way of trying to think through what happens in that connection between the world of our studies and the world of our reality. And what is striking is still how fundamentally human our endeavor is in medicine. What we think about today, and what you think can't help but think about walking on to this campus for the first time as I did today, is the massive amount of knowledge and powerful ability and technology that we have developed in modern 21st century medicine. But what you still find when you get in close, is the messier reality of an individual doctor and an individual patient trying to sort through it all. You have a terrible belly pain sitting there waiting for lunch. Or you're like my daughter this weekend, who is 12 years old and had a fever that started on Thursday and then on Friday it was 103.7 and then very alarming on Saturday morning, it was 105.2. So we took her into the emergency room. And what you want when this happens is that amazing know-how and technology and medicines and tests.
But what do you get? Just a doctor. Someone with bad hair and a weak chin and an annoying little laugh and some things they know and some things they don't. We all want medicine to be an ordinarily field of knowledge and procedure. The going words these days that I hear are, “Precision medicine” or “Personalized Medicine”. Around this idea that when you will go so a doctor, not only will your diagnosis and treatment be careful, but that it will be tuned to your specific genetic type, your specific genetic code. And that conjures an image of this being a world of orderliness. But it's not. And some of that I think can be helped. And also think some of it can't.
There is part of it that leads me to then think that when I got this wonderful opportunity to come talk here at the NIH, that I would then separate these two parts. And the part that applies to the art of failure in medicine, that art around the lack of ability to solve all the problems that have become before us, I think is one set of those problems. This is the part where we think about what can't be helped. And then there is the part that is the science, which is the part that looks at what can be helped. And so I’ll do, as you can see here, the art -- I don't think that's just -- maybe it is. I thought I had turned it off. But I hasn't. I think we have arrived and I think we arrived at a new and critical social juncture in medicine. On the one hand, medicine now has become extraordinary powerful and complex, extending and improving lives for people across the spectrum of afflictions we face from heart attacks to cancer to even depression. We are in a world where the average longevity used to be in the 40s as recently as half a century ago, to a world now, the number 1 killer is no longer malnutrition and no longer respiratory illness. It's instead cardiac disease. Survival not just in the United States where it's reached 78, but all around the world heads been extended so that throughout Asia the number 1 killer is cardiac disease and in China longevity reached 70 years. India, the longevity has gone from the 40s at the time of independence in 1947 to over 65 today. Vietnam survival is to 72. Throughout the Middle Eeast been transforming where survival is around 76 years, very much like our world. And that has meant dramatic transformation of many aspects of the kinds of care and public health systems that we have and one of the amazing things is just how deeply the technology of medicine reaches nowadays.
You go to the village around the area that my father grew up, which is 400 miles east of Mumbai, a village where he is one of 13 brothers and sisters and in a farming family. To get there you have to fly and then drive by jeep for 8 hours and then you have to take a rutted road for a mile by bullock cart. And you still find in this village, the reach of people who understand everything from ultrasound to the importance of getting certain kinds of tests and wanting to know what to do when they have chest pain.
We have, in this world of vast global reach now, extraordinarily powerful and complex care. On the other hand, even there, they understand in a way that we feel very palpably here at home, they understand the notion that doctors are gods is long gone. Which is ironic, right? We were treated as gods when we knew nothing. And now we know a great deal and people are all too aware that although we have got tremendous discoveries, we make, we are hazard with grave mistakes. And no one, I think, has really prepared or few have prepared any of us for this, not doctors nor patients. I think few of us have really tried to examine how medicine works, what decisions really matter, how they are made, how much uncertainty there is and how that uncertainty is dealt with in the real world. And understanding that is not something is that can be answered at the bench or in a randomized trial.
Science is about trying to look at the general example, what is a randomized trial but trying to remove the specifics of the differences among all the people in a cohort? And contrast that with art, which is all about the specifics and understanding the details and understanding what you can make of them and what the differences are among those that exist in the world.
And so, I have pride in the writing that I have done to set out to try to understand this layer, how things go right and how things go wrong in medicine and why. So how often do things go wrong? The brief science here is, there are now a number of studies that have tried to capture this world but there aren't all that many. But one of them I was lucky enough to get to participate in, it was a study directed by Troy Brennen at Harvard Medical School. We went into 28 hospitals in Utah and Colorado and we examined 15,000 charts from patients admitted during that year we were studying things. And what we wanted to know was, how often did things go wrong? How often did the patients come in for care and experience adverse event? A complication that was due to their treatment rather than an ill consequence of their disease. What we found was that 3% of the time it was something that went wrong that resulted in genuine harm to the patient, a death, a disability or prolonged hospital stay. What was also striking was that when we then had a team of reviewers go through these cases and try to grade how many of these were avoidable under the existing standard of care that was available, and how many were not avoidable, that this broke down into being roughly half and half.
So, you enter a world where we have on the range of well over 100 million people admitted to hospitals each year in the United States. 3% means that upwards of 3 million experience significant harm. And half of those involve avoidable errors in care.
The puzzle is, why then do things go wrong? And in the end, there are, I think, 3 reasons. What is embarrassing is that we never really spelled this out for ourselves. The answer to how to understand our mistakes and failures, we ended up having to look to people entirely outside of medicine, business professors and even more embarrassingly, philosophers to help us understand this.
So number 1, were mistakes. That was that 1/2 where there was knowledge that existed but an individual or an institution failed to apply it correctly. And that meant things like a patient given an incorrect drug dose and putting them into renal failure. A patient undergoing an operation and having the main duct to their liver, their common bile duct cut and then having permanent liver damage resulting from that. A patient who died from a complication of uncontrolled bleeding in the operating room. I’m not interested at this hour in talking about that. I will talk in some detail later in the day about what we know about that half. Instead, I’m interested with you today in talking about the other part. What about the half where we really didn't seem to know what we might have done to avert failure? Where we didn't know how to understand things or how to understand why we didn't do better? We tend to forget the other half in the media debates and malpractice courts and so on.. But I don't think we can really understand one without trying to understand the other.
And in that realm of the other half, I think breaks down into two different kinds of problems. One is just mystery. There are problems that we simply don't understand and these are not rare. Obesity is one of our most common health problems and although we understand a good deal about some of the components of obesity, our understanding is woefully grim. We have very little advice we can offer patients that would help them understand how they could as a group, successfully avoid obesity or if they are obese, reduce their obesity, in a sustained way for 5-10 years of their life, which is how long you need it to happen in order to avoid the ill consequence. We have lots of diets that can successfully lead them to lose weight for a year or even two years. But upwards of 93% will be back at their baseline weight among the obese within 5 years. So the only thing we offer nowadays radical surgery. We are going to staple your stomach, redirect your intestine and make you so nauseated you don't want to eat again. And even that, we are still struggling with how long those affects will last. We are at a 60-70% success rate in that group of patients.
It's the sign of something we simply don't understand. Another realm is pain. Probably the most common problem that patients come to us with. I wrote about an architect, a man who designed the University of Massachusetts Medical School and countless buildings around Massachusetts who developed in the course of recovery from a fall, an unremitting back pain that led him to have to leave his work in the law firm, sell the firm, and wind up on chronic disability collecting checks from the government because he can barely get out of bed in the morning. He had x-rays, MRIs and they show that his back is completely normal. So what is going on? We have some theories. We have some way of understanding how a cycle of pain is begun in the brain and cause to produce a perception that there is some tissue damage that may not be there, that is not there. And we understand that this can be an affliction that is deeply disabling and feels exactly the same way as being stabbed in the back feels. But we don't understand why some people get it and other people don't. And we don't understand what to do about it in any convincing and long-lasting way.
The problem of mystery really falls in your realm here at the NIH. It is the realm that scientists tackle. It is the world of misunderstanding. We have come to understand that pain, for example, is not pain. There are distinctions. There is some kinds of pain and other kinds of pains. You need to be able to parse the world into the proper categories so then you can begin to understand what you do about it and so we reached a world where we had the ICD9 code book where we had 13,600 diagnosis and we found in the 10-12 years between the ICD9 code book and the ICD10 code book, that our complexity of understanding of what diagnosis really are has deep end to the point that in ICD10, we have now more than 100,000 diagnosis codes. We went from a world where leukemia was a disease to one where we now understand there are some, I’m told, 50 different kinds of leukemias. We understand the world as c kit. Not a morphus disease entity. That is the world that could only be understood by the investigations of science. It's the possibility of discovering what is in this realm of mystery that we need to categorize better, draw the proper distinctions, and then act upon in different ways from the ones we now perceive.
My realm in the writing world and one of the things that I try to do is really in what happens when while we're in this world, where those distinctions have not been made, or understanding doesn't yet exist. This architect, when I met him, struck me as one of those patients that we will see not infrequently. He comes into the clinic, says they have a pain that we know we're not going to find an answer to describing, nor a treatment that is likely to help them and they make us want to run away. And it's both the nature of our ignorance there and how we deal with it that seems interesting. We run away or we decide, I’ll operate. Because you know, maybe something will work. Or we send the patient to go see a psychiatrist. It's all in your head. We like clear problems that we know how to fix like, a tumor that is removable or a pneumonia or an appendicitis. But that's not what we get. Instead, we get patients with pain or nausea or obesity or fatigue.
I had a television newscaster who I wrote about who was such a severe blusher on television that she ended up losing her job. She went to doctors who sent her to psychiatrists -- or doctors. I didn't mean that. And none of them could explain why she blushed so uncontrollably. They tried medications and none of them worked. So she went to a surgeon and asked him to take away her ability to blush. And actually, this was a possible avenue for doing that. The question is whether he should have tried. And what I’ll tell you, he did try. He did a thoracic sympathectomy. And it did succeed in taking away her ability to blush. And I see no insurmountable reason he shouldn't have given that try.
That is precisely where we start finding that medicine is interesting and also vexing. It's in the nature of medicine that we often must ask despite unclear evidence about what to do, and it's not surprising that these situations are such a struggle for doctors and for patients and it's not surprising that the results can be less than we expect them to be.
I said there were two reasons why we end up having failures where we didn't find avoidable reasons that were obvious. We didn't find they were due to errors. One is, mystery. But the second, and I think a powerful critical reason for imperfection in medicine is uncertainty. Even if our errors were eliminated and the science of human afflictions were to be completely understood, there is still an avoidable uncertainty in what we do and I think that uncertainty is what makes perfection unachievable and also makes medicine so hard.
One day I was seeing patients in clinic with one of the surgical attendings while I was a resident, I was really struck as I was nearing the end of my residency by how often he had to say the words “I do not know” to his patients. I happened to start listening at the beginning of clinic and counting the number of times he would say “I do not know”. And those are 4 little words we are reluctant to utter. We are supposed to have the answers. We want to have the answers. But there was not in the end a single person that day he didn't have to say those 4 little words to. One was a patient who had come in two weeks after hernia repair who wanted to know why he still felt a pain next to the wound instead of right in the wound. One was a patient who had come in one month after a gastric bypass operation and wondered why she hasn't lost weight yet. One was a patient with a large pancreatic cancer and she asked the surgeon if he could get it out. And to all of them, he said, “I do not know.” You still have to have a plan, though, as a doctor. So he told the hernia patient to come back in a week for a checkup on how the pain was doing. He told the gastric bypass patient that everything would be all right and she could come back in a month. And he told the cancer patient that he could try to get the tumor out and although, a colleague who saw the scan thought he shouldn't try, because it would be futile and risky, and although the attending himself thought the odds of success were slim, he and the patient who was only in her 40s, and still had young children at home, decided to go ahead.
The core predicament of medicine, the thing that makes being a patient so wrenching and being a doctor so difficult, and being a part of a society that pays the bills they run up so vexing, is uncertainty. With all that we know nowadays about people and diseases and how to diagnosis and treat them, it can be hard to see this, hard to grasp how deep the uncertainty runs. As a doctor, you find out that the struggle in caring for people is more often with what you do not know than what you do. Medicine's ground state remains uncertainty and wisdom for both patients and doctors is defined by how one copes with it.
My daughter arrived in the emergency room and they confirmed that her temperature was up above 105. They did a bunch of tests. Strep was negative. Flu was negative. She had a cough. She had a pain when she took deep breaths and especially over on the right side. They noticed her skin had a mild erythema. I said, she's just red when she gets a fever. They did blood work. They got the cultures cooking. And the question was, with strep negative and flu negative, they ultimately got an x-ray that seemed negative, do they sent her home or do they not send her home?
I was working in the emergency room and I had been asked to see a 23-year-old woman that I call Eleanor. And she had a red infected leg. She had a blister on the top of her foot from some cruddy sand ales and then went to a wedding where she danced through the whole night and by the morning after she woke up with her left foot feeling sore. She didn't think too much about that at first. That night, she got a fever. And by the next morning, her foot was too swollen to fit into a sneaker. The redness had climbed up her calf. She saw her doctor who diagnosed the cellulitis and gave her a tetanus shot and a prescription for antibiotics. The doctor also outlined the redness with a black felt marker and told her to come back the next day. The next morning, the redness had moved beyond the black line. And so the doctor told her to go to the emergency room to be admitted to the hospital for intravenous antibiotics. And that's when I saw her.
Now this happens. We have all seen cellulitis that requires some iv antibiotics to get better. But I looked at her leg with the redness going up her knee and the pain in her ankle and I couldn't get out of my head another patient I had seen a few weeks before. He was 58 years old and he had 3-4 days of increasing pain in the left side of his chest under his arm where he had abrasion from a fall. He got antibiotics but that night, the rash spread. The next morning he spiked a fever and by the time he got to the emergency room, the skin of his chest had become numb and widely blistered and he began to go into shock. They transferred him to my hospital and we took him to the operating room. Upon opening the skin, what we found was necrotizing fascitis. All the mush else of the left side of his chest going up around his back up to his shoulder and down to his abdomen had turned gray and soft and foul with invading bacteria and had to be removed. The first day in the OR, we had to take even the muscle between his ribs doing something that I learned was called, a bird cage thoracotomy. The next day, we had to amputate his arm.
For a while, we thought we had actually saved him. His fevers went away and we reconstructed his chest and abdominal walls. But one by one, his kidneys, lungs, liver and heart went into failure and then he died. It was among the most awful cases I have ever been involved with.
So, that was what was in my mind. Now you probably know this, necrotizing fasciitis is incredibly rare. It kills up to 70% who get it. We have no known antibiotic to stop it. The most common bacteria is streptococcus, an organism that usually causes little more than a strep throat but in certain people it seems to do far worse. I don't think at least in the last year since I have known we have discovered exactly why. Maybe one of you will tell me. And as for the cellulitis, the organism enters through breaks in the skin. It can be as large as a surgical incision or an abrasion. I found reports of people getting the disease from a rug burn, bug bite, a friendly punch in the arm, paper cut, blood draw, toothpick injury and chickenpox lesion. In many, the entry point is never found at all. Unlike with the cellulitis, the bacteria invade the skin and deeper advancing rapidly along the fascia and consuming whatever soft tissue it encounters. Survival is possible only with early and radical surgery often requiring amputation. To succeed, it has to be done early. By the time the signs of deep invasion are obvious like shock, loss of sensation, widespread blistering, the person is usually unsalvageable.
So standing at Eleanor's bedside, bent over examining her leg, I felt a bit foolish considering the diagnosis. A bit like thinking Ebola virus walked in to the ER. In the early stages of this, it can look just like a cellulitis with the same redness, swelling, fever, high white count. But there are only about 1,000 cases a year of necrotizing fascitis in the United States. Mainly among the elderly and chronically ill, and there are over 3 million cases of ordinary cellulitis. What’s more, her fever had gone away on oral antibiotics and didn't look ill and I knew I would be swayed by that one single case. If there were a simple test to tell the two diagnosis apart that would be one thing but there wasn't one. You have to do an open biopsy. Not something you propose arbitrarily on everybody who comes in with a cellulitis. There I was. I had thought of it. I couldn't help it. I was thinking it. So what should I have done?
Now, I was a resident then. I had an easy out. I called the attending. He came by and took a look and said it looks like a bad cellulitis. But that didn't really solve the problem because it's the reality of medicine that choosing to not do something, to not order a test, to not give an antibiotic, to not take a patient to the operating room is harder than choosing to do it. Once a possibility is put in your mind, especially something as horrible as a necrotizing fascitis, the possibility that doesn't easily go away. We have reached the contradictory ideas about how these kinds of decisions should be made. We say that evidence from research and statistics are what should decide. Evidence-based medicine. But then we say that experience matters that. A physician's judgment is what should decide. And then we say, forget all that. Sometimes the patient should decide.
I tried thinking through how we could have handled the situation from all those perspectives and I had a chance to run a decision analysis with folks on this problem. It took us two full days to get through all the literature and map out the decision tree and reach the right calculations and I got the advice of two decision analysts. But that's how we decided that women over 40 and not under 40 should get mammograms and that it is how we decided we should have bailed out Mexico in the last financial crisis. So why not individual medical decisions?
And when it was done, the analysis had a clear answer. Do not do the biopsy. The likelihood my hunch was right was too low and the likelihood of catching the disease early was too high. But the attending did decide to recommend the biopsy. I had a hunch and although I was uncertain, he found it hard to ignore. We want to believe in the physician's judgment but the data can make it hard to. Studies show over and over again, a consistent tendency to over estimate vivid danger to weigh certain factors too heavily and to over compensate for previous mistakes. We’re enormously inconsistent. We map all the ways in which we are inconsistent over and over again. The likelihood a doctor will decide your gallstones require a gallbladder removal require a -- there is 270% depending on what city you live in. The likelihood they give you a back operation for back pain, varies 880%. Over the past 4 decades, you have cognitive psychologists who shown that a blind algorithmic approach will trump our human judgment in making predictions and diagnosis. The classic is the 1954 book, “Clinical Verses Statistical Prediction” where he described a study of Illinois parolees which compared estimates given by prison psychiatrists that a convict would violate parole with estimates derived from a rudimentary formula that weighed age and previous offenses and type of crime. Despite the formula's crudeness, it predicted the parole violation far more accurately than the psychiatrists did. And they have gone on with other scientists to review more than 100 studies comparing computer or statistical judgment with human judgment in predicting everything from the likelihood a company would go bankrupt to the likelihood of death for liver disease patients. And virtually all cases, statistical thinking equals or surpassed human judgment.
You might think, let's put the human and the computer together then. But, as the researchers went on to point out, that claim doesn't make any sense. If opinions agree it doesn't matter, if they disagree, the studies show you're better off sticking with the computer. We want to trust in judgment but human physician judgment and medicine is all over the map. And so what else is there to do but to trust in the statistics over experience? Over what any particular doctor tells you? And yet, all of us when we are ill, balk at this.
So what should we recommended to Eleanor and with her red infected leg? A biopsy or not? We did recommend the biopsy. And it still didn't answer the more important question, which is what should she decide? There has been a century of writing about the art of doctoring. But nowadays there is an art to being a patient too. There are times when the right thing is to push to ask hard questions, to disagree with doctors even and then there are times when you have to be willing to put yourself in another person's hands even knowing that they are imperfect. She didn't want the biopsy. All she’d come in for was some antibiotics and now we are propose something crazy disease and a operation. Her father was there and this is what he saw. I looked like a kid. And my attending looked hardly any more experienced. He was only a couple years older than I am. And this might have been Harvard but he knew we were just those who happened to be on call.
So what did they do? They asked for a second opinion. So we got another surgeon down, a plastic surgeon who had seen his share of these cases before and he also doubted that it was the disease but he too couldn't feel safe ruling it out without the biopsy. So 40 minutes later she was in the operating room asleep. We took an inch long piece of skin and tissue from the top of her foot down to her tendon and also again another from her calf. The pathologist looked at a frozen section under the microscope and this is what he said. “I do not know.”
We called in another pathologist. He had to drive in from out of town. He hates coming in late at night and he hedged. He said he couldn't rule necrotizing fascitis out. It had pollies in there but there were no organisms to clinch it. So we filleted open her calf and that was when we realized that she did have it. The fascia of her foot and calf were grey, A brownish dishwater fluid seeped out with a faint scent of decay. The bacteria crept up to her leg almost to her knee.
The other thing you realize in medicine, no sooner you made one decision then another one is upon you. Amputate or not? I thought we should. My attending couldn't stand the idea of taking this attractive 23-year-old girl's foot. A sentimentalism I thought could get us in trouble. He that’s what he did. He debrided her anterior muscle groups. He brought her become 3 times in 3 days for further debridements. He put her in a hyperbaric oxygen chamber with no evidence for two hours twice a day. What studies there are, show no benefit. And I still think we just got lucky but she is alive and well and living in Hartford, Connecticut. She runs partial marathons. She has a new job. She's married. She had a child. She's the single greatest save I ever got to be a part of.
I don't think we in medicine have explained or understood the nature of our foul ability particularly well. You have got mystery. You have got error and what we do about it. And to some extent, both of those, the mysteries, and the error, have been measured and funded with work done. But uncertainty, this is science to be done on how best to cope with uncertainty. I think we are undergoing massive experiments on how we cope with uncertainty, with temporary nurses, 80 hour work weeks, reductions in primary care, all of that is unmeasured, untested and very, very important. We could test and derive more algorythms of care but I think we must examine the realities and the only way to do that is in the details. And the best way to do that is art.
What is the experience of doctors and patients in this world? And in writing about these matters, here is what surprised me the most. People have found hearing about these details more reassuring than they found them frightening on the whole. They know we will never achieve perfection, but they also want to know they will never cease to aim for it.
My daughter was sent home on Saturday. She slowly gradually got better and then on Tuesday, they called us to tell us that the lab tests in the ER was wrong. She had influenza A. Who knew? Thank you. [applause] I'd be happy to take questions for the remaining few minutes if there are any.
QUESTION: [low audio]
GAWANDE: So the question is, if this happens in one of the best medical centers in the world, I’m glad you might think of us that way -- what happens in a village like the one my father comes from? One of the striking things, and I’ll get into it a little bit, is that I went -- when I went to India, I’ll get into some detail in the later talk this afternoon. But when I finished my training, I took 3 months to do some surgery across in different systems in India partly as part of a study of surgical care in India. I went there thinking, I’m from Harvard, I have some things to show people. And what you discover are people of extraordinary skill who had much more to teach me than I had to teach them.
They also -- you have the same variability that you have in a lot of the world, which is that there are some good people and there are some not so good people at what they do. What they are really hampered by in a way we are not, is lack of resources, and lack of systems. The surgeon may be terrific but their ability to make sure that the operating room is clean, that the equipment doesn't get stolen, that the power isn't going out, those kinds of problems just seem beyond coping. I don't know how they survived in many ways. And it doesn't surprise me that under those conditions you had people who were desperate to get out of the public health system there either to the corporate private sector or out of the country.
And so, you know, my sense is and a lot of my work is premised around the idea that the struggles of making things go well, of the world that people exist in, at least as I understood it from the analogy and surgery, has very similar themes across all of them -- that you have people of high training, of varying capacity nonetheless, in worlds in which you're struggling constantly against the system and to make things go right. And we're not well-equipped to do some of that work but I also think there are solutions and one of them is what I’ll talk a little bit later, which can include something as simples a check list.
But I also want to the just finish by saying, I came away really struck by the innate abilities of individual human beings that we are all at the end whether it is at Harvard or rural India, ordinary, with limits to what we can do and some things we know and some things we don't. And that success in medicine really is defined around how you help ordinary people succeed in doing this extraordinary thing we call medicine
QUESTION: [low audio]
GAWANDE: thank you for that question and just for those who couldn't hear it -- I see there are microphones. Should we be asking people to come to the microphones? I’ll answer yours and maybe others could step to the microphone. First part is how to understand the line between mistakes and accountability and the relationship between them. And the second part was about how the prospect of being sued weighs into how people act and what they do in day-to-day practice. I think the heart of the problem is that as we have become successful, then success becomes expected. I mean, it's striking that one of our most successful arenas, obstetrics is one in which the costs of malpractice are among the highest. When you reached the point that only one in 1,000 children are harmed in child delivery where it used to be one-30, for a child who is healthy at the time of birth, the expectation then is that every instance would be avoidable. And when you find that still there are mistakes that can happen, that only heightens the sense of wanting to hold the doctor accountable.
You know, my feeling about this just stems from my own practice. I do about 300-350 cases a year. I have about a 3% complication rate. And that means I hurt seriously hurt, rather than help about 10-12 of my patients a year. And those folks weigh on me tremendously. And about half of them will be due to error. And I think we are at some level, accountable for that. But I also think of us like the vaccine world, where the good we do with vaccines outweighs the harm we do and our responsibility is to work to ever minimize that harm. But because we know the harm exists, I would love to see a system like the vaccine world where we have a vaccine injury compensation system that offers much swifter, much more consistent compensation d hold manufacturers accountable, publicizes the events and their studies to look at the patterns of what those vaccine injuries are, rather than putting it in the adversarial world of malpractice where most people never actually get compensation. It's incredibly costly and most of the judgments are secret so you can't see the patterns. And it of course does affect people. It leads to some defensive medicine. I don't think it is as large as people say it is, but it is a force.
QUESTION: Thank you for your talk. It was wonderful and it was a really good reminder of the power of the spoken word and how telling stories eventually is different from telling stories on the page. I’m interested in your 86-year-old and how we perceive error. Imagine a scenario if that 86-year-old had had a different doctor and who had pushed very, very hard for her to do the surgery and she had died during the surgery or suffered some complication. How would we perceive that -- would that have been an error of operating on the wrong patient, in effect?
GAWANDE: One of the puzzles is that if I had persuaded her to go to the operating room, we would never have known that that was an avoidable problem. That everything she would have gone through. And we would have been thrilled to get her through and off to a nursing home never to return home again. So, there is something that is just unfathomable about the layer of what you will never know. And but, had we known that, then we would absolutely regard that as an error. Just the trouble that most of the time we don't. And that's what makes it so hard in those situations. If there is one more question, let's make this the last.
QUESTION: Thank you for your talk. I am a medical student and I did read yore book when I was a premed. My question is –
GAWANDE: I’m glad it didn't scare you away.
QUESTION: My question is, what little I have learned so far and very little I have learned, what I have been taught in that is that evidence-based medicine is the finality of everything and we should trust it and it's counterintuitive as a student to treat the patient but that's what they drill into us. Like your last story you mentioned about the necrotizing fascitis, it's about the art form of where doctors and patients have been. So as a student when you're developing the scales and trying to develop the skills to treat patients in the future, how do you take the person experience and how do you put them together to be a successful physician.
GAWANDE: Thank you for your question. There is a place where the evidence and the art does meet and I think that the best way to understand it is that I think that the pinnacle of evidence based medicine is the randomized trial. And if it is proven in a randomized trial we roll it out in the world and stay that is what you do. But that is not the pinnacle of evidence. The ultimate evidence is, do people get better? And on the whole, this is where the world of the scientist connects with the world of public health.
We are responsible, all parts of the health care system that should produce better care for people next year than it did this year. So, although there is art as well as science to many of these ultimate results, I think our question that we don't measure well and we need to are things like, we do now, according to the National Center for Health Statistics in data that was just released two weeks ago, we do over 100 million operations. It's up from 60 million in 1996 to 2006. So by doing 40 million more operations on people per year, we are up for one every 3 human beings in the country. Are people actually better now than they were in 1996? A lot of them for example are outpatient arthroscopy. So are people getting back to work, exercising and those kinds of things more in this time than they did before? Or in depression, now we have come up with all of these pharmacological treatments, are people more likely to recover from the depression more swiftly in the year 2009 than they did 5 years ago or 10 years ago? And to my knowledge, we have not been able to answer those questions adequately.
We know there is a mix of things that come together. I wrote in my second book about the example of maternal mortality, one we steadily reduced over time. And that's partly in the understanding that it's art obstetrics but a tremendous amount of science that has gone in. Very few randomized trials but the ultimate measure is in a world where we have or used to have maternal mortality, the number 1 killer for women, we had fewer than 500 women die and 4 million child deliveries last year and that's extraordinary.
And so I do think that whether you're a student or any of us, the place that it comes together is being able to ask whether we as a whole system are getting better results with time. It means, changing the way we act. We simply don't measure ourselves. We have more funding for measurement of agriculture in the United States than we have for measurement of public health. And on the basic level, it means setting some markers for yourself and your own work so I picked a couple of things that are my measure of whether I’m harming people or not and I make it my goal that over time I would be harming fewer of my patients. That it won't be 3% a couple of years from now. That it might be down to two. And then maybe a few years after that down to one and keep on going. Thank you.
[applause]
(Music fades in, under VO)
ANNOUNCER: You’ve been listening to discussion on the topic, "The Art of Failure in Medicine" Our speaker was Dr. Atul Gawande, a Surgeon at Brigham and Women's Hospital and Dana Farber Cancer Institute, as well as an Associate Professor, Department of Surgery, Harvard Medical School and Department of Health Policy and Management, Harvard School of Medicine. Once again, you can see a closed-captioned videocast of this lecture by logging onto http://videocast.nih.gov -- click the "Past Events" link. The NIH CLINICAL CENTER GRAND ROUNDS podcast is a presentation of the NIH Clinical Center, Office of Communications, Patient Recruitment and Public Liaison. For more information about clinical research going on every day at the NIH Clinical Center, log on to http://clinicalcenter.nih.gov. From America’s Clinical Research Hospital, this has been NIH CLINICAL CENTER GRAND ROUNDS. In Bethesda, Maryland, I’m Bill Schmalfeldt at the National Institutes of Health, an agency of the United States Department of Health and Human Services.
