Few can doubt that Western medicine has been a phenomenal success. Heart disease kills two-thirds fewer people now than it did fifty years ago. The frequency of conditions as diverse as stroke and trauma is being gradually checked. Mortality from breast cancer has fallen by a quarter in less than two decades. Doctors would dearly like to attribute these impressive results in Western countries to their accumulated expertise and the advances of science. But as Atul Gawande points out in Better: A Surgeon’s Notes on Performance, his latest collection of lucid essays,1 the residual contradiction is that while medicine succeeds, it never seems to succeed well enough. A doctor’s report card might look creditable today. Yet it nevertheless conceals serious unresolved and unacknowledged weaknesses.
“Science and skill,” Gawande writes, are “the easiest parts of care.” What matters more, he suggests, is diligence, doing right by patients, and ingenuity. Despite these “core requirements” for progress, and the fact that doctors have been remarkably successful in all three, errors are still commonly made. These mistakes badly scar the surface of medicine’s success. “Betterment is a perpetual labor,” Gawande concludes. The trick is to understand one’s limits.
One by one Gawande’s arguments are persuasively made and elegantly illustrated with examples ranging from antisepsis to obstetrics. He describes, for example, Thomas, a previously fit seventy-two-year-old man, who one day was found to have a cancer in his left kidney. The tumor was removed surgically but, instead of Thomas’s recovering, his entire body began to swell. Eventually, he was diagnosed with Cushing’s syndrome, a condition in which the adrenal glands overproduce steroid hormone. No matter what his doctors did, they could not stop the hormone from pouring into his bloodstream. Thomas became sicker and sicker. He could not walk and he had recurrent episodes of pneumonia.
By the time Thomas got to see Atul Gawande, there was only one option left—to remove the source of the offending hormone. Although the operation to take out his adrenal glands carried its own dangers, Gawande told Thomas that his only chance of a normal life was to accept the risk. The operation went well. But Thomas suffered a series of terrible complications, culminating in four months of intensive care, a tracheostomy, and incarceration in a long-term care facility. When he last saw Gawande, he could hardly lift his head. All he could do to express himself was to cry. Gawande implies that this was one operation that should have been avoided.
Many of Gawande’s essays began life as magazine and journal articles. Their original diversity and subsequent reworking to construct a larger thesis about the failure of medicine create a difficult tension. Although the individual foundations that Gawande lays down are strong, the overall architecture of his work lacks form and substance.
By striking contrast, Jerome Groopman, a cancer specialist who, like Gawande, writes for The New Yorker, delivers an altogether sharper and more coherent critique of medicine’s mistaken direction. And while Gawande’s prescriptions are gentle and well-meaning homilies—for example, he urges medical students to “change” and to “write something” about their experience2—Groopman presents a forceful and convincing manifesto that, if implemented, would overturn many conventions of modern medical practice.
Groopman’s central claim is that there is a common flaw that undermines much of contemporary medical education and training, as well as the partnership between patient and doctor and even the professional values of medicine. That flaw lies in the way doctors think. His disquiet originated from the frustration he felt working among his students and residents at Harvard Medical School. Whereas once they would take part in challenging and detailed debates about the patients they met and examined on rounds, they now “too often failed to question cogently or listen carefully or observe keenly…. Something was profoundly wrong with the way they were learning to solve clinical puzzles and care for people.”
Using a technique he has honed in his New Yorker essays, Groopman skillfully mixes stories of patients, interviews with doctors, research evidence, and his own personal experiences as a patient to mount an ambitious assault against several large targets. His ire is raised especially by what he sees as the hubris of “evidence-based medicine.” It would seem axiomatic to a nonphysician that medical practice is based on scientific evidence. Not so, according to the zealous advocates of evidence-based medicine, a movement that has come to dominate clinical practice during the past decade. These advocates, led originally by David Sackett and his colleagues at McMaster University in Canada, argue that doctors have preferred to rely on experience and expert opinion—as opposed to research and statistical evidence—out of laziness and a misplaced deference to the authority of received medical wisdom.
Worse, the same advocates argue that when doctors do consult the “evidence base,” they often do so in ignorance of what makes good and bad science. Groopman views the evidence-based approach, which aims to make clinical decisions follow from statistically valid information—in the form of “systematic” reviews, guidelines, or algorithms—as ill-informed by the realities, complexities, and uncertainties of medical practice. A “rigid reliance” on numbers—the numbers, for example, indicating which medicines have been proven effective for certain kinds of disease—will not meet every need of the patient who sits in front of the doctor. Such numbers are needed, but there can never be a purely rational or exact mathematical solution to a patient’s predicament. Groopman concludes that doctors are “being conditioned to function like a well-programmed computer that operates within a strict binary framework.” He disapproves of this medical scientism.
Often patients have conditions or combinations of conditions that do not easily match the supposed evidence. Sometimes patients have problems that are not easy to study scientifically. A strict requirement for evidence before acting may mean that physicians will stop thinking, stop evaluating each patient as a unique human being, and stop applying their knowledge to the particularities of the person before them. Groopman rails several times against the “bean counters” of medicine—doctors who recommend treatments that are seemingly supported by statistics but may not be appropriate for the person they are facing.
There is a still deeper fault line within medical practice. On average, about 15 percent of a doctor’s diagnoses are inaccurate. Groopman directs a well-aimed arrow at a system of medical training that more often than not fails to investigate why these diagnoses are missed. Doctors are rarely taught to ask how an error could have taken place, let alone how it could be avoided in the future. Most are unaware of their mistakes. Even if patients remain unwell, no systematic effort is made to find out where doctors may have gone wrong. Doctors are uncertain about their own uncertainties. (Although for some doctors, such as radiologists, Groopman cites alarming research that shows the worse their performance, the more certain they seem to be that they are right!)
Amid these wide-ranging attacks, the pharmaceutical industry does not escape Groopman’s scrutiny. The discovery of new medicines has delivered huge benefits for patients. But the incentives offered to doctors, and too often accepted by them—gifts, airfares, hotel accommodation, and expensive meals—distort their ability to make unbiased treatment decisions. In a recent survey of over 1,600 American physicians, for example, nine out of ten reported a relationship with a drug company.3 The benefits they received ranged from drug samples to tickets for sporting events, from payments for speaking to money in exchange for persuading patients to join a clinical trial.
The pressure of increasingly aggressive marketing tactics—which, Groopman shows, can amount to overt harassment—by pharmaceutical sales representatives only adds to a climate of acute misunderstanding. Most doctors receive information about new drugs directly from the pharmaceutical companies. Rarely do they investigate what is known about a drug for themselves. This reliance on biased information leaves the doctor poorly equipped to make balanced judgments.
But Groopman reserves some of his most bitter criticism for his colleagues within academic medicine. They have fostered a belief that anyone can take care of patients. This “arrogance” has created a culture at academic medical centers where research is applauded and teaching is taken for granted, where writing scientific papers (for journals like The Lancet) takes precedence over developing clinical skills. He very emphatically offers two examples of inferior, some might say even cruel, care at Memorial Sloan-Kettering hospital in New York City. In one case, he describes a patient who desperately sought treatment at this prestigious institution. But after his cancer failed to respond to chemotherapy, his doctor simply abandoned him, refusing even to return his calls. In a second example, a Sloan-Kettering oncologist told a woman in her fifties with spreading bladder cancer that there was no scientific evidence that would support any further treatment. He spoke of protocols, data, percentages, the statistical likelihood of her survival, and the technicalities of research studies. He was oblivious to her needs, to her hopes and fears. And he left her deeply distressed. Doctors are trained to deal with success, not failure. This hospital may be a cancer center of high international standing, yet Groopman makes the fair argument that the values, attitudes, and behavior of a doctor matter far more than the reputation of the institution at which he works. And here Memorial Sloan-Kettering, at least on the basis of these two instances, falls short of Groopman’s high standards.
So much for the prevailing environment of medicine today. But matters take a more sinister turn when one asks just how well doctors think.
Groopman draws extensively on the emerging cognitive science of medicine, which seeks to understand the mistakes doctors make in evaluating the information they gather from a patient’s history, the physical examination, and results of investigations. He reviews the errors and biases that most doctors unconsciously succumb to when thinking about what their findings mean for a patient’s diagnosis and treatment.
There is a rich and rather disturbing variety of human weaknesses to consider when watching a doctor at the patient’s bedside. Physicians can be easily led astray by seeing the patient from only one—and often very negative—perspective, independent of what the clinical findings suggest. Patients might be stigmatized if they are thought to have a mental health problem, or caricatured if they are judged to have engaged in self-harming behavior, such as alcoholism. This kind of mistake is called “attribution error.” “Availability error” occurs when a doctor makes a decision based on an experience that is at the forefront of his mind but which bears little or no relation to the patient before him. For instance, a specialist in gastroenterology may only think of the gut when evaluating a woman with abdominal pain. He may not think of gynecological causes for her symptoms. The ready availability of his own specialized experience in his assessment of what is wrong with a patient can seriously bias a doctor’s judgment.
“Search satisfying error” is yet another source of misshapen medical thinking. It takes place when a doctor stops looking for an answer to the patient’s problem as soon as he discovers a finding that satisfies him, albeit incorrectly. He gives up too soon. “Confirmation bias” intrudes when the doctor selects only some parts of the information available to him in order to confirm his initial judgment of what is wrong. “Diagnostic momentum” takes over when the doctor is unable to change his mind about a diagnosis, even though there might remain considerable uncertainty about the nature of a patient’s condition. And “commission bias” obstructs good clinical thinking when the doctor prefers to do something rather than nothing, irrespective of clinical clues suggesting that he should sit on his hands.
Doctors are not routinely taught these cognitive pitfalls. Nor are they trained to learn from their effects. Yet these errors and biases can prove fatal. Most doctors are unaware that their thinking is prone to predictable mistakes. Our systems of medical practice neither seek to detect these mistakes nor feed their lessons back to doctors to prevent their recurrence.
An all too typical error is that doctors simply stop observing the patient carefully. Add to that inattention a tendency to hurry the consultation—the economic incentives and pressures to see more patients in less time are immense—and cognitive errors become common. In the face of acute time pressure, doctors will come to rely more and more on shortcuts to make judgments. Pattern recognition based on an instantaneous appraisal of the patient will become the norm. Indeed, the capacity for spot diagnosis is a revered skill among clinically minded physicians. But speed may well create the conditions for further error.
Despite his greater knowledge, the specialist is not immune from these missteps. Specialization can confer undue and sometimes dangerous confidence in those who possess such knowledge. Groopman invites doctors to question any expert who dismisses an unusual cluster of symptoms and signs on the grounds that “we see this sometimes.” Specialists’ appeals to some kind of mystical diagnostic skill owe more to the high opinion they have of themselves than to any kind of clinical reality.
As most of us who have sought medical help will testify, miscommunication is not uncommon between patient and doctor. Many physicians have demonstrable gaps in their ability to convey important information to patients—they are often especially bad in giving advice about how to use prescription medicines. Electronic decision aids—devices that supposedly help doctors to arrive at the correct diagnosis—are unlikely to help, even though many extravagant claims are made for the impact of information technology on health. Groopman believes such electronic fixes might actually encourage more mistakes. They are a distraction. They promote a reductive and unthinking kind of checklist behavior. And they divert the doctor away from what should be his primary focus: the patient’s own story.
Groopman draws very clear and precise conclusions from his review of the surprising array of cognitive dangers that face doctors. Most important, perhaps, is his claim that “competency is not separable from communication.” Communication skills for doctors are nothing new. There is already great attention paid to communication during medical training. But some critics still argue that an emphasis on communication is not relevant to all branches of medicine. If you are going to come under the surgeon’s knife, surely, these skeptics say, you would rather have a surgeon who can cut accurately than one who can converse beautifully. Groopman anticipates this charge. Based on his own labyrinthine experiences as a patient who had operations and many other treatments for pain in the lower back and, later, his hand, he dismisses the accusation of irrelevance: “the surgeon’s brain is more important than his hands,” he affirms. In truth, of course, a good surgeon needs both a sound brain and steady hands.
Doctors must also learn to think differently. A solid base of medical knowledge is not enough to be a good physician or surgeon. Research into cognitive errors in medicine reveals that most mistakes are not technical. They stem from mistakes in thinking. Intuition, a clinical sixth sense, for example, is unreliable. But equally, the assumption that medicine is a totally rational process is also wrong. Doctors may be reasonably smart, but they repeatedly fall into common and well-defined traps.
Physicians can guard against these traps by heightening their sense of self-awareness and becoming conscious of their own feelings and emotions, responses, and choices. All too few doctors have this skill today. Indeed, a doctor’s training can instill utterly contrary traits—confidence and certainty, in particular, which might close off an awareness of one’s usually unconscious weaknesses. Instead, uncertainties should be acknowledged. Unveiling what we are unsure about would not only be more honest, it would also likely promote a degree of collaboration between patient and doctor that has hitherto been lacking.
The corollary of admitting uncertainty is that doctors should be more aware of their errors and should more freely and openly disclose them. Only then will they be able to evaluate and learn from their mistakes. This statement sounds too obvious even to deserve mention. Yet the prevailing medical culture is still heavily weighted against revealing even the possibility of error. Disclosing uncertainty and error will demand a deep change in medicine’s attitude toward emotion. Most physicians fail to recognize, let alone analyze, their own emotional states in clinical encounters. This repression of feeling misses an important variable in the assessment of a patient’s experiences and outcome. The emotional temperature of the doctor plays a substantial part in diagnostic failure and success.
Possibly the most radical proposition that Groopman advances—from the doctor’s point of view, anyway—is that the physician should seek a new ally in helping to correct the cognitive errors and biases inherent in his makeup. This new ally is the patient. Patients can ask questions that pull doctors away from the traps they might otherwise fall into. Groopman concludes:
For three decades practicing as a physician, I looked to traditional sources to assist me in my thinking about my patients: textbooks and medical journals; mentors and colleagues with deeper or more varied clinical experience; students and residents who posed challenging questions. But after writing this book, I realized that I can have another vital partner who helps improve my thinking, a partner who may, with a few pertinent and focused questions, protect me from the cascade of cognitive pitfalls that cause misguided care…. That partner is my patient or her family member or friend who seeks to know what is in my mind, how I am thinking.
What makes a good doctor? Physicians like to think of themselves as members of a profession. But definitions of profession and professionalism change. A century ago a doctor was considered to be part of a social elite. He—and medicine was then very much a masculine endeavor—had a unique mastery of a special body of knowledge. He professed a commitment to levels of competence and integrity that he expected society to respect and trust. This commitment formed the basis for a social contract between the profession and the rest of the community. In return for the moral values, knowledge, and technical skills displayed by doctors, society bestowed on them the authority, autonomy, and privilege to regulate themselves. This version of professionalism is now moribund.
Doctors are no longer masters of their own knowledge. For a start, in many Western countries women now outnumber men at medical schools. The public is also far more educated than it was a century ago. Patients have access to the same information as doctors. They may know more than most doctors about their own condition. Meanwhile, doctors increasingly work in teams. Their responsibilities are shared with many other professionals—nurses, therapists, and pharmacists, for instance. The clinical hierarchy might still favor the doctor. And it is true that the doctor still takes final responsibility for a patient’s care. But the notions of absolute mastery and control no longer hold.
Ideas of privilege, autonomy, and self-regulation are also outdated. For usually good reasons, doctors have been cut down to size in our society. Partly this graying of their public image is because doctors are now seen as fallible. Society is less willing to bow to a doctor’s once sacred authority. As a result, doctors are being made more accountable than ever before to the public. This process has not been without pain. In some countries, such as the UK, they have finally lost the power to govern themselves. Instead, public agencies have the final responsibility for judging their performance.
Competence, knowledge, judgment, commitment, vocation, altruism, and a moral contract with society remain at the heart of what it means to be a doctor. But there are new dimensions to professionalism which herald something of a revolution in the philosophy of medicine. It is these domains that underpin the cognitive science set out by Groopman.
The patient is a far more powerful force in a doctor’s professional life today than in past generations. The patient expects to be more the equal partner of the doctor. Medicine’s goal is not only to cure or palliate disease. It is also to promote a person’s well-being and dignity. Many patients want to be engaged participants in a doctor’s thinking, not just its passive recipients. Whereas once doctors spoke of the doctor–patient relationship, they now increasingly talk of the patient–doctor interaction. The inversion is significant as well as symbolic. It denotes a shift of power from professional to patient. Interaction also better indicates the greater equality in their alliance. The word “relationship” often carried a strong hint of paternalism.
The expectations society has of medicine have changed. Doctors have duties to society, as well as to patients and themselves. They are part of an expensive system of health care which has to be managed responsibly. Doctors have to be good stewards of that system and not merely practitioners working with single patients. These wider responsibilities sometimes run counter to a doctor’s well-developed sense of independent identity. The pace of change in medicine is also so fast that doctors must demonstrate their continuous ability to keep up-to-date as knowledge advances. They should be willing to concede that they are part of a multidisciplinary health team. And patients expect a little compassion to leaven their doctor’s technical expertise.
In research conducted in Britain, doctors seemed to value this more modern description of professionalism, despite their inherent conservatism.4 Physicians have also developed a strong sense of social commitment, despite their having less power and authority. In the US, for example, nine out of ten doctors rate community activity, politics, and patient advocacy as important aspects of their work.5 Doctors seem to be adapting to changing social mores. However, rather than expect doctors to somehow absorb these values randomly during the meandering course of their training, some medical educators are now designing programs to teach professionalism, assess and evaluate it, and identify the best conditions for strengthening and protecting professional values in often highly pressured clinical settings.6 This is why Groopman’s argument is so timely.
Good doctoring is about listening and observing, establishing a trusting environment for the patient, displaying authentic empathy, and using one’s skills and knowledge to deliver superb care. But a neglected aspect of this professionalism is getting doctors to think about their own thinking. Only by doing so are doctors likely to reduce the number of errors they make. What should they do?
Encouraging patients to tell and retell their stories is essential. (Patients’ fears about what might be wrong or their anxieties about the future course of their illness should be drawn out into the open. Whatever the doctor’s own attitudes about the patient, it is a critical element of any mutually respectful therapeutic partnership that the doctor acknowledges the patient’s version of the truth of his or her story. This acknowledgment may mean repeating tests or reconsidering a long and strongly held diagnosis.
In their encounters with patients, irrespective of the financial incentives to be more efficient and productive, doctors must try to remain systematic and thorough when they take a patient’s history and conduct physical examinations. Shortcuts are dangerous. Thinking requires the investment of time. Groopman repeats the same lesson again and again: slow down. The more time a doctor takes, the fewer cognitive errors he will make. (It is just this kind of slowing down that may be more and more difficult under the time pressures imposed by some HMOs and other insurance organizations as part of their demands for efficiency.) And once a decision is made, always retain an element of doubt. That sliver of uncertainty will leave the doctor not only better able to recognize failure early but also free to revise his opinion as new information comes to light.
This change in behavior cannot be brought about at the flick of a mental switch. Groopman gives a telling example of what might be involved in changing the cognitive culture of medicine. Victoria Rogers McEvoy is a former Wimbledon tennis star who now works as a pediatrician in Boston. To ensure that her self-awareness—and so her ability to detect her own cognitive errors—is maximal, she prepares herself psychologically before each clinical encounter, just as she used to do before every tennis match. If doctors understood the biases they were prone to make and briefly but formally prepared themselves to be ready for those biases before seeing a patient, a great deal of medical error might be prevented.
But while we are expecting more insight from doctors, it is also fair to ask: What makes a good patient? Posing questions to improve a doctor’s thinking is certainly part of Groopman’s answer.7 If the consultation with a doctor is going badly, the patient might ask, “I feel that we are not communicating very well with each other. What is going wrong? How can we do better?” When considering a diagnosis, the patient might suggest, “What else could it be? Is there anything you have discovered that doesn’t quite fit? Is it possible that I have more than one problem? What other parts of my body are near where I am having my symptoms?” And when a treatment is being prescribed, the patient might inquire, “How well tested is this drug for the condition I have?”
Groopman’s investigation into how doctors think has important potential to recalibrate the way medicine looks at itself. Doctors are imperfect, to be sure. But their errors carry valuable information that can be put to good clinical use. A doctor’s mistakes are perhaps best seen as signs of a mind at work. The patient and doctor together share a common purpose in getting this mind thinking straight.
May 31, 2007
Metropolitan, 2007. ↩
His advice includes: “don’t complain,” “ask an unscripted question,” and “count something.” ↩
Eric G. Campbell et al., “A National Survey of Physician–Industry Relationships,” New England Journal of Medicine, April 26, 2007, pp. 1742– 1750. ↩
The quite substantial recent changes in British ideas of professionalism are set out more fully in Doctors in Society: Medical Professionalism in a Changing World (London: Royal College of Physicians, 2005). Full disclosure: I was a member of the working party that gathered evidence for this inquiry, and I wrote its final report. ↩
See Russell L. Gruen and colleagues, “Public Roles of US Physicians,” JAMA, Vol. 296, No. 20 (November 22/29, 2006), pp. 2467–2475. ↩
See Frederic W. Hafferty, “Professionalism—The Next Wave,” The New England Journal of Medicine, Vol. 355, No. 20 (November 16, 2006), pp. 2151–2152. ↩
Again through a revealing story, Groopman endorses faith as an additional means to help someone become “a productive partner in the uncertain world of medicine.” Although trained in science, many doctors will resist those pure scientists who consider religion an outmoded and medieval relic of our age. Religious belief many have a powerful and valuable point to play in some patients’ response to their illness. See, for example, James Randerson, “The God Disunion: There Is a Place for Faith in Science, Insists Winston,” The Guardian, April 25, 2007, p. 3. ↩