Atul Gawande
Atul Gawande; drawing by David Levine

Among the huge number of men and women practicing medicine during the past century, only a few have been able to write well about their profession for the general reader. Of the most accomplished among them, it can be said that their words are a natural extension of their healing. Any physician becoming familiar with the stories and essays—and the poems too—of such doctors soon recognizes the depth of their concern for those who come to them for help and their understanding of the difficulties and satisfactions of caring for the sick.

The best example of such dedicated physician-writers is William Carlos Williams, who was for seventeen years co-director of pediatrics at the Passaic General Hospital and physician to tens of thousands of adults and children, some three thousand of whom he delivered. In the poems and stories of Dr. Williams can be sensed what it means to be in a state of clinical watchfulness, and it serves him well. Like Dr. Williams, the best medical writers have let the story come forward unhampered, though they themselves are not only in it but of it as well. Being good clinicians, they have not hesitated to immerse themselves in another’s life or another’s narrative, but they do not take it over. The contemporary writer who best exemplifies these qualities is surely Oliver Sacks, whose immense learning is so unobtrusively woven into his accounts of nervous disorders and evocations of his own personal experiences and those of his patients.

Now we have Dr. Atul Gawande with a new and different voice, bringing to modern high-tech medicine the same clinical watchfulness that writers such as Williams and Sacks have brought to bear on the lives and emotions of often fragile patients. The reader feels that he is accompanying Dr. Gawande at the bedside or operating table and senses that the events are being described and confronted as they should be if we are to get an honest sense of the complexities of twenty-first-century healing. If there is a genre to which the chapters of his collection of essays belong, it is not that of any known form of literature, but is instead an experience familiar primarily to surgeons—the Mortality and Morbidity Conference, in which doctors meet to discuss what went wrong in particular cases or how a better outcome might have been achieved. The purpose of the conference is to reveal and examine human error, specifically the errors of members of specific surgical teams. Acknowledging fallibility is the key to improvement.

But at such conferences, that acknowledgment of fallibility is made only to other physicians. Laws in most states protect the conference discussions from public scrutiny; otherwise few if any participants would be as forthcoming about their errors as they must be if the meetings are to be useful. If the participants cannot be absolutely frank, the aims of the conference are frustrated. Nothing is learned and nothing changes.

Gawande contends that the learning and changing should not be restricted to doctors. The more knowledge that the public and individual patients can acquire about the realities of modern medical practice, the better may be the choices that any of us makes about our own care. The old patterns of a paternalistic form of medicine nowadays are rapidly giving way to what bioethicists call autonomy—the principle that the patients should make their own informed decisions about treatment. One result is an increasing demand for the information they need, whether general knowledge about the ways in which disease affects the body, or specific information about the variety of treatments available. But, with the exception of a few loudly reported cases of malfeasance, very little has been written about the ways in which doctors actually do their work, and most specifically the sources of error inherent in diagnosis, in treatment, and in the training of young physicians.

In this book and in his articles in The New Yorker, where much of its contents first appeared, Gawande has written about the art of healing as it takes place day to day in America’s leading academic medical centers. As though to make clear from the outset that he is dealing with revelations about the unsure ground on which so much of medicine stands, and has always stood, he has chosen to divide his short book into three parts, called “Fallibility,” “Mystery,” and “Uncertainty.” A glance at those headings in the Table of Contents should be enough to concentrate a reader’s mind—if the book’s title has not already done so.

Gawande begins with his own beginning, a fumbling start to his surgical training that will, I am sure, bring back disquieting memories to many doctors who read his book:

This will be good, I tried to tell myself: my first real procedure. My patient—fiftyish, stout, taciturn—was recovering from abdominal surgery he’d had about a week before. His bowel function hadn’t yet returned, leaving him unable to eat. I explained to him that he needed intravenous nutrition and that this required a “special line” that would go into his chest. I said that I would put that line in him while he was in his bed, and that it would involve my laying him out flat, numbing up a spot on his chest with local anesthetic, and then threading the line in. I did not say that the line was eight inches long and would go into his vena cava, the main blood vessel to his heart. Nor did I say how tricky the procedure would be.

During his attempts to insert a plastic catheter into a large vein located just beneath the clavicle, or collar bone, Gawande seems to do everything wrong before finally managing to thread the device into its proper place. The patient, unaware that he is in the hands of a beginner, is subjected to delay and discomfort in the name of medical education. A more experienced resident is standing nearby and could easily have spared him the anguish; but she is there to teach and Gawande is there to learn, and so the patient, who is there to be treated, becomes an object for instruction. The two doctors acknowledge that in helping this man they have also used him:


Outside the room, S. said that I could be less tentative the next time, but that I shouldn’t worry too much about how things had gone. “You’ll get it,” she said. “It just takes practice.” I wasn’t so sure. The procedure remained wholly mysterious to me. And I could not get over the idea of jabbing a needle so deeply and blindly into someone’s chest.

In these ways, the term “medical practice” takes on a little-acknowledged meaning that badly needs to be understood if the process of learning, as Gawande puts it, is not to be “hidden behind drapes and anesthesia and the elisions of language.” Without such practice there would soon be no highly trained physicians to care for the sick. Mishaps will occur that might not have taken place were only the most senior doctor doing every procedure. These are painful truths, but the days of concealing them will be numbered if Gawande has his way. But “everyone is harmed,” Gawande writes, “if no one is trained for the future.”

Mishaps have their usefulness, and this is what the Mortality and Morbidity Conference is all about. Gawande describes one in which the doctors discuss his failure to insert a breathing tube in a surgical opening cut into a rapidly failing woman’s windpipe, a procedure called tracheotomy. At literally the last moment, patient and resident were saved by the arrival of attending physicians who skillfully took over. When the near-tragedy was reviewed at the following week’s conference, every one of the errors leading up to it was described and discussed in an atmosphere where responsibility and correction, not blame, were the themes. As in virtually every such situation, what was clear is that supervision of young doctors by seniors is the key to avoiding the errors that trainees may make.

That medical errors occur and that physicians are culpable in both committing and reporting them is hardly news to the general public. But what is relatively new is the candor with which the profession has begun to deal with it. In one of his many useful references to recent studies, Gawande writes of two much-quoted articles published in the New England Journal of Medicine in 1991, which reported that 4 percent of hospital patients had “complications” resulting from their hospital care, of which two thirds were due to errors in care, and that 44,000 patients died in a year at least partly as a result of that care.

Soon after, the Institute of Medicine of the National Academy of Sciences undertook its own exhaustive study of the problem. When the institute’s conclusions were published in November 1999, the monograph’s title acknowledged the role of human error: To Err Is Human: Building a Safer Health System. Instead of being hidden, the results were made public. To the amazement of many physicians, the report’s frankness has relieved them of some of the burden of personal error. Patients and families have begun to understand the weaknesses built into the systems by which the sick are cared for. And they have also begun to understand the uncertainties of medicine itself, even when it is in the most capable hands. If there is a single characteristic of medicine that every layperson should be aware of when accepting or rejecting an offered medical procedure, it must certainly be its inherent uncertainty.


A hundred years ago, the reigning medical authority of the time, Sir William Osler of Johns Hopkins, in musing on the state of affairs that he called “the uncertainties of medicine,” spoke of “this everlasting perhaps with which we have to preface so much connected with the practice of our art.” Surely, many would say, things are different in the current era of molecular biology. Paradoxically, the opposite is true. The remarkable advances of ultramodern biotechnology have brought with them complexities of such magnitude that medicine sometimes seems in danger of being overwhelmed by forces of increasing intricacy and incomprehension. In certain situations, only the small number of superspecialists who deal in a particular aspect of diagnosis or therapy are equipped to interpret a finding or observation. What conclusions are to be drawn from a hard-to-interpret test of liver function? In what situation is it better to recommend angioplasty rather than coronary bypass? Which of three possible antibiotics is best for a particular resistant bacterium?

The opinions of the highly specialized consultants called to address such problems sometimes conflict. The responsible attending physician who must actually make the major decisions may not fully comprehend every assumption and each piece of medical evidence that should enter into them. And once having chosen a course of action, he must then trust to an unknown number of others that it will be properly carried out. These range from the surgeon doing an operation to the maintenance man servicing a piece of equipment or the orderly cleaning it. We are caught in a spiral of uncertainty that is only magnified by the increasing range of our capacities.

Not only that, but medical doctors may be less capable than they think. The medical literature is full of impressive-seeming statistics and opinions that are subject to the frailties of the methods by which they are reached, and the frailties of the human beings responsible for them. Were that not so, women and their doctors would not now be so utterly confused about the usefulness of mammography in dealing with breast cancer. Gawande writes of a forty-two-year-old lawyer who refuses a biopsy in spite of a suspicious mammogram. She has already had, she truculently points out, three biopsies in the past five years, always for similar findings and always negative for cancer. Gawande reviews for us all the strategies he might use to convince her to accept having yet another.

But in one of his rare lapses, Gawande never seems to consider that he has no definite grounds for making his recommendation. Statistics for large numbers of patients support his advocacy of urgent biopsy, but this woman’s history does not; and it is specifically this woman for whom he has convinced himself that his judgment is correct. The fact is that he has chosen to recommend a biopsy in the face of the uncertainty that is inherent not only in her situation but that is always present when dealing with the vagaries of diagnosis and treatment of possible breast cancer. Neither he nor anyone else really knows the best way to deal with that infinitely variable disease. If there is a condition that exemplifies medicine’s ultimate uncertainty, malignancy of the breast is surely it.

There have been many clinical debates and about-faces concerning this and other conditions in recent decades; they are all traceable to the hubris of believing that medicine is anything more than what Gawande calls it in the subtitle of his remarkable book: an imperfect science. I would go further even than that. It is not a science at all. It remains what it has always been and will ever be—an art that uses science as well as it can, and too often incorrectly or inconsistently.

All of this was put neatly in a single sentence written in 1967 by Dr. Alvan Feinstein, widely acknowledged as one of the world’s leading authorities on the clinical use of statistics, and the founder of the form of research called clinimetrics. In his book Clinical Judgment he wrote:

At a time of potent drugs and formidable surgery, the exact effects of many therapeutic procedures are dubious and shrouded in dissension—often documented either by the unquantified data of “experience” or by grandiose statistics whose mathematical formulations are so clinically naive that any significance is purely numerical rather than biologic.

In other words, much of what a well-trained physician does, and is convinced he knows, is not supported by valid scientific evidence applicable to the actual situation of the indi-vidual sick person. Declarations such as Feinstein’s demolish the claims that the new so-called “evidence-based medicine,” in which decisions are said to emerge from a review of all pertinent studies and literature, can somehow convert diagnosis and therapeutics into an exact science.

A convinced advocate of medical certitude might reply that none of this applies to our current era. But it does. Alvan Feinstein and the few biostatistically trained bedside doctors equipped to analyze such matters have found no reason to modify their verdict. Until his death only a few months ago, Feinstein was still speaking and writing in the same vein, and many controversies, such as those over mammography, have borne him out; so have the numerous reversals in authoritative opinion every doctor has seen in professional journals and in his experience. Is radical mastectomy the best treatment for breast cancer? Is drinking coffee associated with an increased risk of pancreatic malignancy? Should every ruptured spleen be removed? Is a low-fiber diet the best treatment for chronic diverticulitis? Is acid production by the stomach the key factor in peptic ulcer? Should every man, or nearly all men, with prostate cancer have surgery? Are most cases of impotence psychosomatic? The answer to every one of these questions was once “Yes” and is now “No.” Feinstein himself debunked the studies linking coffee to pancreatic cancer.

As a veteran of emergency rooms, post-operative wards, and intensive-care units, I find it nothing less than astonishing that a physician as new to his art as Gawande (he is near the end of his residency training in surgery) should be so forthcoming about the actual experience of moment-to-moment hospital work. Most relative neophytes are so awed by having been accepted into the priesthood of specialty medicine and so reluctant to cause themselves trouble in the institutions in which they will work for the coming decades that they would be hesitant to risk offending their seniors. And there is even more to be astonished at. Realizing that he was risking professional retribution when he contemplated publishing an article on medical errors in The New Yorker, which now appears in Complications, he approached Dr. Michael Zinner, his chief of surgery, for permission. Though the article frankly described doctors making some of the mistakes I have mentioned, Zinner immediately agreed that it could be published:

…I gave him the manuscript and then, a few days later, walked into his office braced for the worst. As it turned out, he didn’t love it. How could he? No hospital public relations department in the world would have let an essay like that go out. But he did a remarkable thing: he supported me anyway. The article could easily backfire, he warned me, with the public or other doctors. But if there was flak, he would help me, he promised. And he let me go ahead.
In the end, there never was any flak. Even when my colleagues from work have disagreed with what I’ve written, they have been constructive and engaged and have held nothing against me. We are all, I’ve found, in the process of trying to understand how much of what we do is good, how much of it can be better.

I find it additionally astonishing that the residents and attending staff of a great American hospital (Gawande never gives its name, but Michael Zinner is the chief of surgery at Boston’s Brigham and Women’s Hospital) have not only tolerated the descriptions in Gawande’s book, but have been supportive of the man who has made them. At one of Harvard’s elite teaching hospitals and probably in most similar institutions, it is the public relations department and not the doctors who are troubled about the disclosure of error. The doctors are gradually coming to know that understanding and decision-making can only be improved when the public shares the knowledge of medical fallibility.

These are the matters of which Gawande writes, and he does not hesitate to use personal anecdotes to describe them, including those of his own failures in diagnosis or therapy. We read, for example, of the pseudonymous Mr. Jolly, about to be discharged from the intensive care unit after treatment of an infection in the leg wound of one of the many arterial operations he has undergone. While sitting up watching TV, he suddenly fell back dead. Gawande was so sure that the cause was a blood clot thrown to the lung (called a pulmonary embolus) that he criticized a more junior colleague who had one day earlier ordered a dose of vitamin K to speed up the clotting process. Gawande “all but accused him of killing the patient.” But an autopsy showed the man’s chest contained six pints of blood, caused by his having ruptured an undiagnosed aortic aneurysm. Every one of the attending physicians, including the radiologists who read the chest X-rays, had missed it.

It is perhaps because Gawande is so candid in telling of his own mistakes that colleagues have been willing to accept his disclosure of the inevitability of error in every doctor’s practice. It is not only individual error to which he refers. He correctly makes the point that most of the real disasters in medicine are the result of a single human mistake being compounded by a string of other mishaps, none of which by itself would have resulted in the unhappy outcome had they not tumbled one upon the other. Medical tragedies are most commonly the result of a rapidly evolving process, rather than a single adverse event. They tend to happen because entire systems misfire, and the fail-safe mechanisms built into them break down.

Gawande’s failed tracheotomy is a case in point. It involved a heavy-set woman brought into the emergency room unconscious, following the rollover at high speed of a car in which she was the driver. With blood in her nose and throat obstructing the airway, an attempt by the emergency room physician to insert a breathing tube into her larynx was unsuccessful; the vigorous efforts to do so resulted in more bleeding and in swelling of the windpipe. With the airway thus severely blocked, the inexperienced Gawande began hectically to attempt a tracheotomy, but not until he had waited too long in hope that the intubation might succeed. The lighting was bad, the patient’s neck was fat, Gawande made the wrong choice of incision, he cut into an unseen vein deep in the thick tissues and the suction tubing became clogged with clotting blood; and when he finally cut into the trachea, he was unable to get the tube through the opening. Most egregious of all, he had not called for more experienced help. Just as all seemed lost, the senior attending surgeon appeared and took over. But by then, the wound was such a mess that even he could not do much. And then, at a moment that could have been fatal, an anesthesiologist who had just arrived was somehow able to insert a pediatric breathing tube between the swollen vocal cords, and the patient’s life was saved.

When the case was discussed at the Mortality and Morbidity Conference, it was clear that the accumulating errors, and unavoidable circumstances, had led to the near-disaster. They followed one upon the other as though in a series, each growing out of the previous ones: the patient’s thick and fatty neck; the consequent inability of the emergency room physician to intubate her and the trauma resulting from his attempts; Gawande’s not calling for help and his too-long wait before commencing tracheotomy, a procedure at which he had little experience; his poor choice of incision; the inadequate lighting; the clot-obstructed suction tubing and the unavailability of a replacement; the senior surgeon’s failure to come to the bedside on his own, having lingered unnecessarily long elsewhere. If any of these factors had not occurred along the way, the patient would very likely not have come so close to dying. The people involved failed and the instruments failed, but most importantly of all, the system for dealing with extreme emergencies did not stand up to the stress placed on it.

It is not enough, Gawande and so many others have pointed out, to improve the performance of everyone involved—it is the entire complex system by which care is rendered in hospitals that requires attention; and even that will not result in the complete elimination of tragedy. “No matter what measures are taken, doctors will sometimes falter, and it isn’t reasonable to ask that we achieve perfection,” Gawande writes. “What is reasonable is to ask that we never cease to aim for it.”

Gawande is as unflinching in his approach to medical uncertainty as he is when dealing with fallibility. His discussion of the problem, like his treatment of all other matters in his important book, alternates between the anecdotal and the scholarly. Like the thorough clinician he is, he seeks guidance and documentation in the medical literature and in the experience of colleagues. But his own bedside encounters have made him skeptical of the ultimate value of strict adherence to the principle of autonomy, which holds that patients should make informed decisions for themselves. Many factors lead him to question just to what extent the principle of autonomy should apply—among them the admittedly confused atmosphere in which decisions are often made; the reluctance or even inability of many patients to participate in such decisions; and the patient’s confidence in the doctor’s traditional ethical commitment to doing his best to provide a cure.

Like so many physicians whose experience vastly exceeds his own, and like so many patients too, Gawande doubts that the principle of autonomy can be universally applied. Even when it is clear to a dispassionate reader (as it is in the case of the woman to whom he recommended breast biopsy) that he is guided only by his own interpretation of the laws of probability, he looks for ways to convince patients that the doctor really does know best. In this, he is no different from most of his more experienced colleagues, myself included, in doubting—sometimes with dubious reason—the unquestioned value of total reliance on a patient’s ability to choose wisely:

…The conundrum remains: if both doctors and patients are fallible, who should decide? We want a rule. And so we decide that patients should be the ultimate arbiter. But such a hard-and-fast rule seems ill-suited both to a caring relationship between doctor and patient and to the reality of medical care, where a hundred decisions have to be made quickly…. The doctor should not make all those decisions, and neither should the patient. Something must be worked out between them, one on one—a personal modus operandi….
…Medicine will continue to struggle with how patients and doctors ought to make decisions. But, as the field grows ever more complex and technological, the real task isn’t to banish paternalism; the real task is to preserve kindness.

The second section of Gawande’s book—entitled “Mystery”—is the least satisfying, but only by comparison with the other two. Here he writes of patients whose problems were obscure, dramatic, or mystifying because they were rare or simply because they were inexplicable. Older readers will be reminded of the essays of Berton Roueché in The New Yorker starting in the 1940s, but there is a difference. Roueché wrote of medical detective work, in which mysterious problems were addressed by clever investigators. Presented as puzzles, they were sometimes solved elegantly and to the satisfaction of every reader. The problems Gawande tries to deal with are hardly as neat. The men and women of whom he so engagingly writes suffer from a variety of debilitating problems: pain of unknown origin, pathological blushing, the uncontrollable vomiting of pregnancy, morbid obesity—none of them conditions that yield easily to standard medical measures. But the stories of the patients and physicians grappling with these recondite or resistant symptoms are fascinating, and Gawande tracks them down to their ultimate resolution in diagnosis and therapy, even when the outcomes are less satisfactory than might be wished.

Reading Complications, we become aware of the emergence of a new medical voice, and a welcome one. Here we find clinical perception, a wide-ranging knowledge of the pertinent literature, and the precocious wisdom of a young physician confronting the realities of one of America’s leading hospitals. He writes with directness and lucidity—and humility as well—that lift the veil of obscurity and obfuscation behind which so many of the most far-reaching dilemmas of today’s medical care have been half-hidden. The writings of Atul Gawande convey the quiet assurance and tone of the doctor acting as both observer and participant. This is clinical watchfulness at its best.

This Issue

July 18, 2002