Editor's Note: Atul Gawande, MD, MA, MPH, Associate Professor of Surgery at Harvard Medical School and the Harvard School of Public Health, is an accomplished surgeon and writer and is the recipient of a 2006 MacArthur Fellowship. He is an active clinician at Brigham and Women's Hospital and the Dana Farber Cancer Institute. Dr. Gawande has written two acclaimed and best-selling books: Complications: A Surgeon's Notes on an Imperfect Science and Better: A Surgeon's Notes on Performance. A staff writer for the New Yorker, he also recently completed a stint as a guest columnist for the New York Times. Dr. Gawande is leading the World Health Organization's Second Global Patient Safety Challenge: "Safe Surgery Saves Lives." We asked him to speak with us about professionalism, training, patient safety, and the writing process.

Dr. Robert Wachter, Editor, AHRQ WebM&M: You started Better with a story about the "eyeball test"—a resident correctly perceives that something is amiss in a patient whose objective signs were fine. How do you reconcile the importance of "eyeball tests" in an era that's increasingly about measurement?

Dr. Atul Gawande: Well, if we're not even asking at the end of the day whether we've won or lost, then we have no shot at getting better at what we're doing. We just have a random intuition about the direction to go. Surgery is a good case in point. I just finished an operation today, and I'm going to go out and talk to the family and the family is going to ask me how it went. And what we ordinarily say is, "It went fine," because I just had a gut feeling that it went fine. Yet some patients crash and have major complications within the next 30 days and some don't. If we don't have any way to measure whether eyeball tests worked out a month later, then we're not going to do well. Very few of us in surgery actually take a look to see what happened at 30 days to monitor whether we're getting better or getting worse. So I think there has to be more than just the eyeball test. In the story you mentioned from the book, the resident thought there was something wrong with this elderly lady who wasn't breathing well. All the numbers were right. We were all going off to our morning conferences, and he circled back to see her again, instead of waiting all day to see what happened. Well, part of what intrigued me about that case was not only did he try to think whether something more was going on than the numbers implied, but he recognized that the people around him weren't good enough to do that for themselves. Then he made the decision to act on it. Part of what that reflects is also the fact that we now talk endlessly about systems and the need to make them work right, but around us the systems break down constantly. You must have good, conscientious people who are willing to say that sometimes the system doesn't work and then take responsibility for it.

RW: Do you think that the whole safety movement, which emphasizes systems over all else, has gone too far by distracting us from the fundamental importance of professionalism and the integrity and commitment that the resident showed in that circumstance?

AG: I don't think so. There's definitely a push by people who care about patient safety to make sure that we are thinking in terms of systems, because we just don't ordinarily think that way. But when you ask, is the result that the majority of us think that systems are all that matter and individual conscientiousness doesn't? I don't see a backlash resulting in which people are just checking out and no longer caring for the patient. It's still the bottom line that "The secret in caring for the patient is caring for the patient," as the old saying goes. I don't think we've abandoned that. We just have to realize that caring sometimes requires caring about both the system and what you yourself are doing.

RW: The concern has been raised, and I've felt it myself, that as we push toward more systems thinking, inevitably we feel a little bit like cogs in the machine. And the passion and commitment that our forebears had might diminish. Do you worry about that?

AG: I do worry about it. But in surgery, our dominant way of thinking still emphasizes individual conscientiousness. We are skeptical that the system can ever be made to work well enough that we can sit back and just let it go. We are either hypervigilant or we fail. There are a few surgical residents who take the 80-hour work week too much to heart and enter into a shift mentality. But they are not in the majority. I have a fear that if that [the shift mentality] becomes a dominant ethic in any part of medicine, then patient care will, well, it will become an inhumane profession.

RW: One of the themes that goes through your work is that providers are fundamentally good people and that once we measure and recognize that we are not as good as we would like to be, our inherent professionalism will motivate us to change. Many outside observers of medicine are skeptical about that. They think that something more is needed to kick-start providers and hospitals into improvement—transparency, pay-for-performance, something more. Where do you come down on that?

AG: I told the story in the book about the 117 cystic fibrosis centers across the country, which work the way we want all of medicine to work. They take care of high volumes of kids with cystic fibrosis, they have well-trained specialists, they follow clinical guidelines, and they participate in innovation and clinical trials. That said, there is a 14-year difference in survival—to age 33 versus age 47—between the median centers and the top centers. When they shared data with one another but did not make it transparent, it was not clear that the gap was closing in any way. Now they're engaged in an experiment to answer this question. They have gone public this past November with the scores and records for cystic fibrosis treatment in all those centers by name. Now you can look on the Web and know where the great centers, and the mediocre centers, are. I am nearly positive that miniscule numbers of patients will shift their care. But it's very clear that a great number of centers are very upset about this having gone public. The physicians and leaders are suddenly sharing information, traveling to other places, seeing what the top people do that's different from what the other folks do. It has caused a great deal of ferment that I think is tremendously positive. So my short answer is, just having the data isn't enough. I don't believe that the power of transparency comes from patients voting with their feet. I think it comes from some combination of professional responsibility, plus embarrassment, and then the sudden recognition, ah, here is where I can find out where the best people are and learn how they do it.

RW: The concerns that people typically have about measurement are that the measures aren't very good and that, as we move toward outcome measurement, people will take easy patients and shun hard ones. Do you share these concerns?

AG: I partly feel that some of that is not a bad thing. That if a place is concerned that they're going to drop into the bottom tier because they take one or two really hard cases, maybe they aren't the right place to take on those hard cases. Furthermore, it bothers me tremendously when people say that they are going to potentially hurt people because of being measured imperfectly. For example, bariatric surgeons do gastric bypass surgeries that in the last decade have exploded in number and have caused unacceptable rates of death and major complications in certain places. They have opposed making the data public. But the single most important piece of information you could have about whether that surgery is worthwhile for you is the mortality risk. It's a worthwhile operation if it can be done with very low mortality, and it is not at all worthwhile if the mortality rate is above 1%. That information ought to be public. It is imperfect—no data are ever going to be perfect. But I think the public is able to sift through that information in a more sensible way than we give them credit for. They do it with car information and all kinds of other potential major decisions: housing, schools, and so on. Yes, some people will not use that information correctly or not use it at all, but the dominant effect is for the good.

RW: As we move into an environment with more transparency, people are paying more attention to quality and safety. Obviously, this poses a fundamental tension in teaching environments, often pitting the need for oversight of trainees against the importance of trainee autonomy in their development into competent professionals. How do we balance that?

AG: That was the central part of the discussion in Complications, which I wrote from the perspective of being a resident, trying to learn how to do surgery and perhaps to learn on people. Now I'm writing from the perspective of a teacher, having to teach on people so that there will be people of the next generation to take care of me [in the future]. My thought about it is that if you gave people a choice about whether they'd like to be learned upon, very few sensible people would ever say yes. That they might say, "Okay, you get one stick of a needle to try to get the IV in, that's fine"—but not for something like learning to take out half your liver. As a teaching institution, we have to end up teaching surgeons and doctors, and we have to be able to say we're going to make the learning process as safe as possible. That means doing some simulations and supervising everything. But at some point you have to let the learner try. The one responsibility we have then is to be willing to say that every patient is subject to that process. We cannot go to the public and say we will have to learn on patients if we don't make that happen to fellow doctors—if the Chairman of Surgery gets a pass, or his wife comes in and the intern is told, no, you don't put a nasogastric tube into her nose because only the attending will do that for her—then we're hypocrites. We're going to fail to convince the public. If this is a process that is truly necessary for everyone, then number one, we'll make it good enough that it's acceptable for every patient, and number two, it will be a burden shared.

RW: Much has been made of the "surgical personality" and the perception that surgeons are arrogant and their relationships are particularly difficult because of that. Is there a surgical personality, and is it changing?

AG: There definitely was a certain surgical type of personality—surgeons who could scare the hell out of everybody. And it was their way or the highway. The best surgeons did not usually have that kind of personality. There's a saying about surgeons that I love, "Sometimes wrong, never in doubt." I actually liked that kind of personality, the person who was willing to make a choice in certain situations and just live with the consequences. I think that element of the surgical personality remains. But it's definitely true that there's a generational shift. The nurses describe the fact that in the operating room they never witness from my generation of surgeons what they still sometimes put up with in the older generation, throwing angry fits, getting testy after somebody says something—that kind of thing. Now you're seeing young surgeons who are very respectful. They understand that there is a team aspect to what happens in a hospital or in surgery, especially when lives are on the line, and that you need everybody working together. But they also understand that there still needs to be a captain of the ship, and you have to be able to exercise leadership in an effective way.

RW: What is your view of the progress in the field of patient safety over the last several years?

AG: I think it's a mixed review. There is absolutely no evidence that hospitals today are safer than they were when the Institute of Medicine report came out. In the last year, though, I would say there's been a remarkable turnaround in the way we've gone about it. We went from focusing on a few small-bore solutions like putting in computers that could prevent medication errors and trying to find the correct side of the body so you didn't chop off the wrong limb when you went to surgery. We shifted from that to looking at larger bore questions, something like the 100,000 Lives Campaign. They've essentially tied universally important, broad-based steps in care to the things that actually save the most lives. By doing that, in the next 5 years, I think you'll start to see hospitals shift toward measuring how we do on the key things and how can we make it better.

RW: So from your standpoint, the movement from fairly small process changes to more global measures is helpful because it brings a more holistic approach to the entire problem?

AG: There is this palpable feel of a difference in results-oriented thinking. We became frustrated after we realized in the last several years that we had no idea if we were making fewer mistakes or not. The concern was there, but the action and sense of effectiveness weren't. We have now seen several examples of effective things you can do that really change complication rates. For example, you have Peter Pronovost going to the state of Michigan and carrying out a project with all of the hospitals there, showing them five steps that can prevent ventilator-associated pneumonia. As a result, they've knocked down the rate of pneumonia in patients on ventilators by more than 80% in a phenomenally short time. That is real population-wide improvement in health and safety. We're just on the cusp, I think, of finding things to do that work that way. Some of what I wrote about are precisely some of those discoveries.

RW: When you start writing, do you start with a theme and then try to find a patient to illustrate the issues, or do you start with a patient and that gives you the theme?

AG: It's more of the first process. Medicine is incredibly rich with stories. The trouble is that that's almost a problem with writing about medicine. We have patients who die. We have patients in pain. We have patients suffering difficulties with great bravery. And telling these stories is not actually new. They've been told over and over again. The hard part is having the ideas that illustrate the real world of decision-making and morbidity, having ideas that cast that struggle in a new way. So what I generally find is that I read something in the paper or have a conversation with someone, and it opens up a whole new set of ideas that I hadn't thought about. For example, the idea of washing hands, it's not something we think about at all, until I realized that we had not made fundamental improvement in this mundane but critical process. The thing that opened my eyes about it was the recognition that SARS was being transmitted from one hospital in China to killing 10,000 people in more than a dozen countries mainly because health care workers weren't washing their hands. And that's when I realized that there was something interesting in our failure to wash hands. But I really didn't have it come together until I started thinking about the hospital in Pittsburgh that has managed to go close to 2 years without a wound infection from resistant bacteria. Sometimes my patients will ask, "Am I going to be a case to write about?" And I always tell them, number one, I won't write about you without your permission. Number two is that the answer is no, because what I really feel I struggle for are ideas. There are plenty of patients or stories or cases that test whether the ideas are strong or weak. That's what I hope my writing does, that it tests the ideas in the real world.

RW: How does your life as a writer make you a different surgeon?

AG: The main way in which it makes a difference is I'm a happier surgeon. I think without the chance to step back and see things from 30,000 feet in the way my writing lets me, I would burn out in the trenches. If you don't have some way to pull yourself back from the grind, you fail. For many people, it's their research or academics or family or a musical instrument. For me, it's this combination of writing and the research that makes that possible. But when I'm in the operating room or seeing a patient in the clinic or heading down to the ED, at that moment all I'm thinking about is the person and what's going on and trying to decipher it. I don't think the writing actually changes that.