Every Patient Tells A Story Part 1

Every patient tells a story_ medical mysteries and the art of diagnosis.

by Lisa Sanders.

AUTHOR'S NOTE

The stories I tell here are real. In order to respect the confidentiality of these patients who were kind enough to share their stories with me, I have changed their names. In some instances I have altered certain identifying details as well. The doctors featured in these pages described in detail some of their most difficult diagnoses-mistakes and all. They are distinguished not by these errors but by their willingness to discuss them. No one should be punished for simply being honest, and so I have changed the names of these brave doctors.

The use of p.r.o.nouns when you are speaking of an individual remains problematic in writing now that we can no longer just use the generic "he." There is no rule on this at this point, so in this book I will refer to the generic doctor as she and the generic singular patient as he.

INTRODUCTION.

Every Patient's Nightmare.

Barbara Lessing stared out the window at the snowy field behind the hospital. The afternoon sky was dark with yet more snow to come. She looked at the slender figure in the bed. Her daughter, Crystal, barely twenty-two years old and healthy her entire life, was now-somehow-dying. The young woman had been in the Na.s.sau University Medical Center ICU for two days; she'd been seen by a dozen doctors and had scores of tests, yet no one seemed to have the slightest idea of just what was killing her.

It all started at the dentist's office. Crystal had had a couple of impacted wisdom teeth taken out the month before. But even after the teeth were gone, the pain persisted. She'd called her mother halfway across the state just about every day to complain. "Call your dentist," she'd urged her daughter. And she had. Finally.

The dentist gave her a week's worth of antibiotics and then another. After that her mouth felt better-but she didn't. She was tired. Achy. For the next week she'd felt like she was coming down with something. Then the b.l.o.o.d.y diarrhea started. And then the fevers. Why didn't you go to the doctor sooner? the trim middle-aged woman scolded her daughter silently.

Barbara had gotten a call from a doctor in the emergency room of this suburban hospital the night before. Her daughter was ill, he told her. Deathly ill. She drove to Syracuse, caught the next flight to New York City, and drove to the sprawling academic medical center on Long Island. In the ICU, Dr. Daniel Wagoner, a resident in his second year of training, ushered her in to see her daughter. Crystal was asleep, her dark curly hair a tangled mat on the pillow. And she looked very thin. But most terrifying of all-she was yellow. Highlighter yellow.

Wagoner could feel his heart racing as he stood looking at this jaundiced wisp of a girl lying motionless on the bed. The bright unnatural yellow of her skin was shiny with sweat. She had a fever of nearly 103. Her pulse was rapid but barely palpable and she was breathing much faster than normal despite the oxygen piped into her nose. She slept most of the time now and when awake she was often confused about where she was and how she had gotten there.

To a doctor, nothing is more terrifying than a patient who is dying before your eyes. Death is part of the regular routine of the ICU. It can be a welcome relief to the patient, or to his family. Even a doctor may accept it for a patient whose life can be prolonged no longer. But not for a young girl who was healthy just weeks ago. These doctors had done everything they could think of but still there was a fear-a reasonable fear-that they'd missed some clue that could mean the difference between life and death for this young woman. She shouldn't die, but the young resident and all the doctors caring for her knew that she might.

Crystal's thin chart was filled with numbers that testified to how very ill she was. Wagoner had been through the chart a dozen times. Virtually every test they'd run was abnormal. Her white blood cell count was very high, suggesting an infection. And her red blood cell count was low-she had barely half the amount of blood she should have. She'd gotten a transfusion in the emergency room and another after she was moved to the ICU, but her blood count never budged. Her kidneys weren't working. Her clotting system wasn't either. Her yellow skin was covered in bruises and her urine was stained deep red.

Sometimes, if you just work hard enough to keep a patient alive-to keep the blood circulating, the lungs oxygenating, the blood pressure high enough-the body will be able to survive even a vicious illness. These are the miracles brought by technological advances. Sometimes, but not this time. The ICU team gave Crystal bag after bag of blood; they did their best to sh.o.r.e up her damaged clotting system; she got pressers (medications designed to increase blood pressure) and fluids to help her kidneys. She was on several broad-spectrum antibiotics. And yet none of that was enough. She needed a diagnosis. Indeed, she was dying for a diagnosis.

This book is about the process of making that diagnosis, making any diagnosis. So often this crucial linchpin of medicine goes unnoticed and undescribed, yet it is often the most difficult and most important component of what physicians do. As pervasive as medicine has become in modern life, this process remains mostly hidden, often misunderstood, and sometimes mistrusted. In movies and novels it's usually the one-liner that separates the fascinating symptoms from the initiation of the life-saving therapy. On television it's the contemporary version of Dr. McCoy's (Star Trek) magic diagnostic device (his tricorder) that sees all, tells all. But in real life, the story of making a diagnosis is the most complex and exciting story that doctors tell. And these are stories that doctors tell. Just as Sherlock Holmes or Nick Charles (the hero of the Thin Man Thin Man mysteries) or Gil Grissom ( mysteries) or Gil Grissom (CSI) delights in explaining the crime to victims and colleagues, doctors take pleasure in recounting the completed story of their complex diagnoses, stories where every strange symptom and unexpected finding, every mystifying twist and nearly overlooked clue, finally fit together just right and the diagnosis is revealed. In this book I'll take you into those conversations and onto the front lines where these modern medical mysteries are solved-or sometimes not. delights in explaining the crime to victims and colleagues, doctors take pleasure in recounting the completed story of their complex diagnoses, stories where every strange symptom and unexpected finding, every mystifying twist and nearly overlooked clue, finally fit together just right and the diagnosis is revealed. In this book I'll take you into those conversations and onto the front lines where these modern medical mysteries are solved-or sometimes not.

Just a hundred years ago, journalist and acerbic social critic Ambrose Bierce defined the word "diagnosis" in his Devil's Dictionary Devil's Dictionary as "A physician's forecast of disease by [taking] the patient's pulse and purse." And that was true for most of human history. Until very recently, diagnosis was much more art than science. as "A physician's forecast of disease by [taking] the patient's pulse and purse." And that was true for most of human history. Until very recently, diagnosis was much more art than science.

But since Ambrose Bierce wielded his rapier pen, there has been a revolution in our ability to identify the cause of symptoms and understand the pathology behind them. In the era in which Bierce wrote, Sir William Osler, considered by many to be the father of American medicine, was able to write a comprehensive summary of all the known diseases in his 1,100-page masterwork, The Principles and Practice of Medicine The Principles and Practice of Medicine. These days each tiny sub-branch of medicine could provide as many pages on its super-specialized knowledge alone.

At the birth of medicine, millennia ago, diagnosis (the identification of the patient's disease) and prognosis (the understanding of the disease's likely course and outcome) were the most effective tools a doctor brought to the patient's bedside. But beyond that, little could be done to either confirm a diagnosis or alter the course of the disease. Because of this impotence in the face of illness, the consequences of an incorrect diagnosis were minimal. The true cause of the illness was often buried with the patient.

In more recent history, medicine has developed technologies that have transformed our ability to identify and then treat disease. The physical exam-invented primarily in the nineteenth century-was the starting point. The indirect evidence provided by touching, listening to, and seeing the body hinted at the disease hidden under the skin. Then the X-ray, developed at the start of the twentieth century, gave doctors the power to see what they had previously only imagined. That first look through the skin, into the inner structures of the living body, laid the groundwork for the computerized axial tomography (CT) scan in the 1970s and magnetic resonance imaging (MRI) in the 1990s. Blood tests have exploded in number and accuracy, providing doctors with tools to help make a definitive diagnosis in an entire alphabet of diseases from anemias to zoonoses.

Better diagnosis led to better therapies. For centuries, physicians had little more than compa.s.sion with which to help patients through their illnesses. The development of the randomized controlled trial and other statistical tools made it possible to distinguish between therapies that worked and those that had little to offer beyond the body's own recuperative powers. Medicine entered the twenty-first century stocked with a pharmacopeia of potent and effective tools to treat a broad range of diseases.

Much of the research of the past few decades has examined which therapies to use and how to use them. Which medication, what dose, for how long? Which procedure? What's the benefit? These are all questions commonly asked and that can now be regularly and reliably answered. Treatment guidelines for many diseases are published, available, and regularly used. And despite concerns and lamentations about "cookbook medicine," these guidelines, based on a rapidly growing foundation of evidence, have saved lives. These forms of evidence-based medicine allow patients to benefit from the thoughtful application of what's been shown to be the most effective therapy.

But effective therapy depends on accurate diagnosis. We now have at our disposal a wide range of tools-new and old-with which we might now make a timely and accurate diagnosis. And as treatment becomes more standardized, the most complex and important decision making will take place at the level of the diagnosis.

Often the diagnosis is straightforward. The patient's story and exam suggest a likely suspect and the technology of diagnosis rapidly confirms the hunch. An elderly man with a fever and a cough has an X-ray revealing a raging pneumonia. A man in his fifties has chest pain that radiates down his left arm and up to his jaw, and an EKG (electrocardiogram) or blood test bears out the suspicion that he is having a heart attack. A teenage girl on the birth control pill comes in complaining of shortness of breath and a swollen leg, and a CT scan proves the presence of a ma.s.sive pulmonary embolus. This is the bread and b.u.t.ter of medical diagnosis-cases where cause and effect tie neatly together and the doctor can almost immediately explain to patient and family whodunit, how, and sometimes even why.

But then there are the other cases: patients with complicated stories or medical histories; cases where the symptoms are less suggestive, the physical exam unrevealing, the tests misleading. Cases in which the narrative of disease strays off the expected path, where the usual suspects all seem to have alibis, and the diagnosis is elusive. For these, the doctor must don her deerstalker cap and unravel the mystery. It is in these instances where medicine can rise once again to the level of an art and the doctor-detective must pick apart the tangled strands of illness, understand which questions to ask, recognize the subtle physical findings, and identify which tests might lead, finally, to the right diagnosis.

To the doctors caring for Crystal Lessing, it was not clear if the mystery of her illness was going to be solved in time to save her life. Certainly there was no shortage of diagnostic data. There were so many abnormalities it was difficult to distinguish between the primary disease process and those that were the downstream consequences of the disease. The doctors in the ER had focused on her uncontrolled bleeding. Why wasn't her blood clotting? Was this disseminated intravascular coagulopathy (DIC)-a mysterious disorder that frequently accompanies the most severe infections? In this disease the fibrous strands that make up a clot form w.i.l.l.y-nilly inside blood vessels. These tough strands slice through red blood cells as they course through the artery, releasing the oxygen-carrying contents and strewing the torn fragments of cells into the circulation. Yet careful examination of Crystal's blood didn't reveal any of these cell membrane fragments. So it wasn't DIC.

And why was she yellow? Hepat.i.tis was the most common cause of jaundice in a young person. But the ER physician found no evidence of any of the several viruses that can cause hepat.i.tis. Besides, the blood tests they'd sent to check how well her liver was working were almost normal. And so, they concluded, it wasn't her liver.

Once Crystal was transferred to the ICU, the doctors there had focused on the b.l.o.o.d.y diarrhea. She'd had two courses of powerful antibiotics for a dental infection before the diarrhea and fever started. That fit the pattern for an increasingly common infection with a bacterium called Clostridium difficile Clostridium difficile, or C. diff C. diff., as it's known around the hospital. The use of antibiotics can set the stage for this bacterial infection of the colon, which causes devastating diarrhea and a severe, sometimes fatal, systemic illness.

The ICU team had looked for the dangerous toxin made by the C. diff C. diff. bacteria but hadn't found it. Still, that test can miss up to 10 percent of these infections. In fact, it's standard practice to retest for the bacterial toxin three times before believing that the disease isn't present when suspicion for the disease is high. The ICU team started Crystal on antibiotics to treat C. diff C. diff. anyway-the story of antibiotics followed by b.l.o.o.d.y diarrhea made that their leading diagnosis.

But Dr. Wagoner, the resident caring for the patient, was unsatisfied with the diagnosis. Too many pieces didn't seem to fit. The antibiotics and diarrhea made sense but the diagnosis left too many of her symptoms unexplained.

That Friday afternoon-forty-eight hours after Crystal was admitted to the hospital-Wagoner did what doctors often do when faced with a complex case: he reached out to a more experienced physician. Despite all the available technology, the tools doctors often rely on most are the most old-fashioned-a phone, a respected colleague, a mentor or friend.

Dr. Tom Manis was one of the most highly regarded doctors in the hospital. A nephrologist, he was called in because of Crystal's kidney failure. But as Wagoner presented the patient to the older doctor, it was clear he was hoping that Manis could help them figure out more than just the kidney.

As Manis read through the chart, he too became alarmed. Wagoner was right-this diagnosis didn't fit well at all. For one thing, C. diff C. diff. colitis is usually a disease of the sick and elderly. The patient was young and had been healthy. But even more to the point, C. diff C. diff. wouldn't account for the profound jaundice and the anemia that persisted despite multiple transfusions. So Manis did what the resident had done-"I called every smart doctor I knew," and told them each the perplexing tale of Crystal Lessing-again, using those irreplaceable tools, a phone and a friend. One of those friends was Dr. Steven Walerstein, the head of the hospital's Department of Medicine.

It was early evening by the time Walerstein had a chance to see the patient. He didn't read her chart. He never did in tough cases like this. He didn't want to be influenced by the thinking of those who had already seen her. Far too often in these difficult cases something has been missed, or misinterpreted. And even if they had collected all the pieces, they had clearly put the story of this illness together incorrectly.

Instead he went directly to the patient's bedside.

Walerstein introduced himself to the young woman and her mother. He pulled up a chair and sat down. Getting the whole story is essential but it can take time. Can you tell me what happened, from the beginning? he asked the sick girl. Like the cla.s.sic detective in a mystery novel, he asked the victim to go over the crime once more. "I've told this story so many times," Crystal protested. Her voice was thick with fatigue, her words slurred. Couldn't he just read it in her chart? No, he told her gently but firmly. He needed to hear it from her, needed to put it together for himself. Slowly the girl began telling her story once more. Her mother took up the tale when the girl became confused or couldn't remember.

Once the two women had gone through the events that brought each of them to the ICU, Walerstein asked the mother for a little more information about her daughter. Crystal had just graduated from college, she told him. She was working as a nanny while she tried to figure out what she wanted to do with her life. She didn't smoke or drink or use drugs. And she'd never been sick. Never. She roughly brushed away tears as she described her daughter to this kindly middle-aged doctor. He nodded sympathetically. He had a daughter.

Then Walerstein turned back to the young woman in the bed. Her yellow skin was now hot and dry. Her lips were parched and cracked. Her abdomen was distended and soft, but he could feel the firm edge of the liver, normally hidden by the rib cage, protruding a couple of inches below. She moaned again as he put pressure on this tender and enlarged organ.

Only then did he allow himself to look through her chart. He skipped over the notes and buried himself in the myriad abnormal test results that had been collected over her two days in the intensive care unit.

Walerstein was a general internist, admired for his broad knowledge of medicine and his clinical ac.u.men. If he didn't know the answer right off the bat, he was known to ask questions that would lead to the answer. And this young woman needed an answer, or she would die. Having thoroughly examined the patient and her chart, Walerstein took a moment to step back and look for some kind of pattern buried in the chaos of numbers and tests.

The ICU doctors had focused on the b.l.o.o.d.y diarrhea and had gotten nowhere. Indeed, although the girl had seen blood in her stools at home, since arriving at the hospital she had very little diarrhea. It didn't seem to Walerstein to be the most important of her symptoms. Instead, Walerstein went back to the striking feature that had caught the ER doctor's eye-her blood would not clot.

The liver makes most of the proteins that cause blood to coagulate. Could it be that her liver was no longer making these proteins? Could it be that her liver wasn't working at all? Could it be that her liver wasn't working at all? That would account for both the bleeding and the jaundice. But liver failure is usually marked by dramatic elevations in certain enzymes that are released when liver cells are destroyed, and those enzymes had been nearly normal since she'd come to the hospital. Her doctors had taken that to mean that the liver was not involved in this deadly process. That would account for both the bleeding and the jaundice. But liver failure is usually marked by dramatic elevations in certain enzymes that are released when liver cells are destroyed, and those enzymes had been nearly normal since she'd come to the hospital. Her doctors had taken that to mean that the liver was not involved in this deadly process.

What if, instead, the liver had already been destroyed by the time Crystal came to the hospital? What if these markers of liver injury (known as transaminases) weren't elevated because there were no more liver cells left to injure, if all the liver cells had already been destroyed? No one in the Emergency Department or in the ICU had made this leap. And yet if you looked at it in this way, as Walerstein did, everything made perfect sense. It all fit.

He then turned his attention to the profound anemia that had been noted from the start. Despite multiple transfusions, Crystal still had only half as much blood as she should. She was bleeding-her red-tinted urine showed that-but she wasn't bleeding that much. It was clear that her red blood cells were being destroyed within her body. Deep within her chart there was a test that showed this but Walerstein noted that the team caring for her hadn't considered this in their search for a diagnosis.

Too often information you don't initially understand is simply set aside, especially when there is such a wealth of information. Walerstein understood this phenomenon. And once set aside it's often forgotten. It happens all the time. But Walerstein also knew that in a difficult case like this one, data that has been set aside often holds the key.

So Crystal had liver failure and red blood cell destruction. That combination stirred something deep within his memory. Walerstein could feel the pieces slowly come together like the cogs in some ancient machine. And then suddenly he knew what this was.

The internist hurried to the library to check his hunch. Yes! He was right. This combination-liver failure and red blood cell destruction-was an unusual manifestation of an unusual inherited illness: Wilson's disease.

In Wilson's disease, the liver lacks the machinery to regulate copper, an essential mineral found in the diet. Without these chemical tools, excess copper builds up in the liver and other organs and slowly, insidiously breaks them down. Usually this process takes place over decades, but occasionally, for reasons that are still not understood (though it is often a.s.sociated with the use of antibiotics, as it was in Crystal's case), the copper blasts out of the liver-destroying the organ in the process-and a lifetime of the stored mineral floods into the bloodstream. Once there, all h.e.l.l breaks loose: the copper demolishes red blood cells on contact. The kidneys work hard to clear the cell fragments from the circulation but are gravely injured in the process. Meanwhile the high levels of copper in the bloodstream attack virtually every organ in the body. In this form, the disease is rapidly and universally fatal unless the patient receives the only possible cure-a new liver to replace the one destroyed by the jailbreak of copper, a liver that has the machinery to dispose of the excess mineral. If this was Wilson's disease, this patient needed a transplant immediately.

But first Walerstein had to confirm the diagnosis. It was late on a Friday night and so it would be impossible to measure the amount of copper in her blood-in any case, his hospital laboratory didn't even do that test. But there was another way to diagnose this disease. Patients with Wilson's will often acc.u.mulate copper in their eyes-a golden brown ring at the very outermost edge of the iris. Walerstein hurried back to the ICU. He carefully examined the girl's eyes. Nothing. He couldn't see the rings, but maybe an ophthalmologist with his specialized equipment could. "It's not often that you call the ophthalmologist at nine p.m. on a Friday" to do an emergency examination, Walerstein told me. But he related the girl's story one more time-this time with a likely diagnosis, if only he could confirm it. "I'm sure he thought I was nuts, until he saw the rings." As soon as Walerstein had the results, he hurried into the patient's room to tell the girl and her mother what they'd found.

Crystal Lessing was transferred by helicopter that night to New YorkPresbyterian Hospital. Patients with the greatest need get priority in the transplant line. Without a new liver, Crystal would die within days and that put her at the front of the line. She received an organ the following week and survived.

Crystal's story is every patient's nightmare: To be sick, even dying, and have doctor after doctor fail to figure out why. To be given the wrong diagnosis, or no diagnosis at all, and to be left to the ravages of disease with nothing more than your own endurance and the doctors' best-guess therapy to rely on. To live or die in a modern hospital filled with the promise of treatment and yet without a diagnosis to guide its use.

How was Walerstein finally able to make a diagnosis after so many others had failed? How do doctors make these tough diagnoses? Walerstein is modest about his role in the case. "I think I was just lucky enough to know about this rare form of this rare disease. No one can know everything in medicine. I happened to have known about this," he told me. It's sometimes a mysterious process-even to the doctors themselves. "A bell went off and the connection was made," Walerstein told me. "That's all I know."

This book is about that bell-how doctors know what they know and how they apply what they know to the flesh-and-blood patient who lies before them. It can be a messy process, filled with red herrings, false leads, and dead ends. An important clue may be overlooked in the patient's history or examination. An unfamiliar lab finding may obscure rather than reveal. Or the doctor may be too busy or too tired to think through the case. Even the great William Osler must have had his bad days.

And the patient, by definition sick, frequently tired and in pain, inarticulate with distress, is given the essential task of telling the story that could help a doctor save his life. It's a recipe for error and uncertainty. It is "an inferential process, carried out under conditions of uncertainty, often with incomplete and sometimes inconsistent information," says Jerome Ka.s.sirer, former editor of the New England Journal of Medicine New England Journal of Medicine and one of the earliest and most thoughtful modern writers on this unruly process. and one of the earliest and most thoughtful modern writers on this unruly process.

It's a wayward path to an answer filled with unreliable narrators-both human and technological-and yet, despite the unlikeliness, that answer is often reached and lives are saved.

Often, but not always. The possibility of error is ever present.

It's certainly not news that medical errors are common. In 1999 the National Inst.i.tutes of Health (NIH), Inst.i.tute of Medicine, released a report on the topic-To Err Is Human. In that now famous report the authors concluded that there were up to 98,000 patient deaths due to medical errors every year-the same number of deaths we would see if a jumbo jet crashed every day for a year. That set off a national effort to reduce the rate of errors in medicine that is still bearing fruit.

That report did not look at errors in diagnosis. And yet errors in diagnosis make up a large chunk of the errors made in medicine. Depending on which study you believe, it is the first or second most common cause of medical lawsuits. Studies suggest that between 10 and 15 percent of patients seen in primary care specialties-internal medicine, family medicine, and pediatrics-are given an incorrect diagnosis. Often the error has no effect-people get better on their own or return to their doctor when the symptoms get worse-but doctors and patients alike worry about the possibility of a diagnostic error that hurts or even kills. In a study of over thirty thousand patient records, researchers found that diagnostic errors accounted for 17 percent of adverse events.

Doctors are getting better at making diagnoses. Tests and imaging have made possible diagnoses that were in earlier times only knowable at autopsy. And while postmortem studies done in this country suggest that the rate of unsuspected diagnoses has been remarkably stable over the past several decades, that statistic is skewed by the diminishing number of autopsies performed. A study done at the University Hospital in Zurich, Switzerland, where there is a 90 percent autopsy rate, shows that over the past few decades the number of missed or erroneous diagnoses has steadily dropped. Another study done for the Agency for Healthcare Research and Quality (the AHRQ, a research arm of the NIH) shows a similar trend in this country if you account for the ever shrinking pool of autopsies.

Still, the fear of getting it wrong is always present for both doctors and patients. As a result, there is a new and growing interest in better understanding diagnostic errors in medicine. The first-ever conference on the topic-one of the earliest signs of growing research interest-was held in Phoenix in 2008. And the AHRQ, the government agency charged with improving the quality of health care in this country, offered its first grants for research on the topic in the fall of 2007.

Research into diagnostic error, like research into the diagnosis process itself, is still a very new field. There is even difficulty in deciding what const.i.tutes a diagnostic error. What a thoughtful patient may consider an error is not necessarily the same as that which his equally thoughtful doctor might consider an error.

For example, when a patient comes to my office with a sore throat and a fever, I might check for strep, and if it's not present I'll probably send him out with a diagnosis of a viral illness. But I share with all such patients what I expect to happen over the next few days-that they should start to feel better within a day or two. And if not, I tell them to call me and let me know. Because, while the odds are overwhelming that this is simply a viral syndrome, it's not 100 percent certain. I might be wrong. The test might be wrong. It might be mono. It might be some other kind of bacterial tonsillitis. It might be cancer.

I can't just check under the hood and see if the spark plugs need to be replaced-the way a mechanic diagnoses the funny noise your car is making. Instead, I have to listen to the engine and, based on the indirect evidence I can collect, make a thoughtful and well-informed guess as to what is probably probably going on. going on.

If I send that patient home with a diagnosis of a viral syndrome and he doesn't get better and has to come back, would that be a diagnostic error? I suspect the patient would think so. And certainly it wasn't a correct diagnosis. But did I make an error? Should I have done something different?

I could have been more certain. I could have sent my patient to an Ear, Nose, and Throat specialist who could have looked down his throat with a special scope. I could have even asked for a biopsy of the red and swollen tissue to confirm my diagnosis. That would be time-consuming and painful for the patient and ridiculously expensive. But even then, the diagnosis would not have been 100 percent certain. In medicine, uncertainty is the water we swim in.

The chance of being wrong is overwhelming when dealing with something more complicated than a sore throat. Doctors-far more than the patients they care for-recognize that some error is inevitable. From the first moment a doctor sets eyes on a patient, she begins to formulate a list of possible causes of the symptoms-what's known as a differential diagnosis. As the story emerges, that list is modified-diseases on the list disappear to be replaced by new ones that more closely adhere to the patient's story, or exam, or sometimes test results. By the end of the encounter the doctor has a list of likely suspects.

If the doctor has worked through the problem well, there's a very good chance that one of these possible diagnoses will be right. The rest though, by definition, will be wrong. We are regularly wrong in the pursuit of being right. It's important to have a list of possibilities because medicine is complicated and diseases and bodies differ. We frequently have a diagnosis that we consider most likely, but we're taught also to come up with a plan B because our patients don't always have the most likely disease. The question we are taught to ask ourselves is, if it isn't that, what else could it be?

As a collector of diagnostic stories, I find myself frequently asking why one doctor was able to make the diagnosis when others before her couldn't. Where were the errors? How were they made? What can we learn?

Sometimes the problem is a lack of knowledge. This was certainly the case in Crystal Lessing's story. She had a rare presentation of an unusual disease. One of the human limitations in medicine is that no one can know everything.

There were errors in thinking in Crystal's case as well. Recognizing that the fundamental problem was liver failure was an essential step in Walerstein's thinking process-a realization that eluded all of the doctors who had seen her initially.

There were also mistakes in some of the data collected from the patient. Walerstein noted that the patient's "b.l.o.o.d.y diarrhea" consisted of a couple of episodes of b.l.o.o.d.y stools the day she came to the hospital. And Walerstein was also the first to note the enlarged and tender liver when he examined the young woman-a hint that the organ wasn't as normal as the blood tests suggested. Abnormalities uncovered by testing were also not interpreted correctly. Crystal's jaundice was initially attributed to the destruction of the red blood cells. Yet when further testing revealed that this red-blood-cell ma.s.sacre was not the result of an abnormal immune system improperly attacking the cells, Walerstein was the first to consider other causes of red cell destruction. Research suggests that diagnostic errors-like this one nearly was-are often due to a mult.i.tude of missteps made along the way.

The solution to this case, as with so many cases, lay in the proper use of all the tools we have at our disposal. Walerstein took a careful history, performed a thorough physical examination, and identified the important lab abnormalities. Only then was he able to connect the information about that patient with the knowledge he had to make the diagnosis. Only then did the pieces of the puzzle come together.

In telling you these stories I try to put you, the reader, in the front line, in the shoes of the doctor at the bedside-to know that feeling of uncertainty and intrigue when confronted with a patient who has a problem that just might kill him. I try to show you the mind of the physician at work as she struggles to figure out exactly what is making the patient sick. To do this I have divided this book according to the steps we take in the evaluation of each and every patient we see. Each chapter focuses on one of the tools of our trade, how it's supposed to work, and how errors send us astray. As physicians become more open about what we do, we make it easier for patients to understand what they can do to more fully partic.i.p.ate in their own care.

This book has its roots in a column I have written for the past six years for the New York Times Magazine New York Times Magazine. The column has been my opportunity to share with general readers my personal collection of fascinating diagnostic histories. It's a collection I began (unwittingly) to a.s.semble years ago, while my own medical career was still in its formative stages.

I came to medical school as a second career. The first I spent in television news, mostly covering medicine, mostly for CBS. I hadn't planned to go to medical school; it wasn't some long deferred dream. But one day, while filming with television correspondent Dr. Bob Arnot, I watched him save an elderly woman's life. He was supposed to be shooting a stand-up on whitewater rafting when he suddenly disappeared from the raft I was watching in the monitor. The cameraman and I searched the distance and saw him on the banks of the river, pulling an elderly woman onto the rocky sh.o.r.e. The cameraman refocused on this new image and I watched with fascination as Bob performed basic cardiopulmonary resuscitation (CPR) and brought the nearly drowned woman back to life.

I didn't quit television right then and there and head off to medical school, but it planted an idea and revealed a hidden dissatisfaction with my role in TV. Television reaches millions, but touches few. Medicine reaches fewer but has the potential to transform the lives of those it touches.

So I did two years of premed at Columbia University, then applied and was accepted at the Yale School of Medicine. I completed my residency training at Yale's Primary Care Internal Medicine program and have stayed on here to care for patients and teach new generations of doctors.

When I started medical school I thought I would be most interested in pathophysiology-the science behind what goes wrong when we get sick. And, in fact, I loved that subject and still do. But what captured my imagination were the stories doctors told about their remarkable diagnoses-mysterious symptoms that were puzzled out and solved. These were the stories I found myself telling my husband and friends at the dinner table.

Covering medicine for as many years as I had, I thought I understood how medicine worked. But these stories revealed a new aspect of medicine-one well known to doctors but rarely discussed outside those circles. In writing my columns and now this book, I try to share a face of medicine that is both exciting and important. Exciting because the process of unraveling the mystery of a patient's illness is a wonderful piece of detective work-complicated yet satisfying. Important because any one of us might someday be that patient. The more you know about the process, the better you will be prepared to a.s.sist and understand.

CHAPTER ONE.

The Facts, and What Lies Beyond.

The young woman was hunched over a large pink basin when Dr. Amy Hsia, a resident in her first year of training, entered the patient's cubicle in the Emergency Department. The girl looked up at the doctor. Tears streamed down her face. "I don't know if I can take this any longer," twenty-two-year-old Maria Rogers sobbed. Since arriving at the emergency room early that morning, she'd already been given two medicines to stop the vomiting that had brought her there-medicines that clearly had not worked.

"I feel like I've spent most of the last nine months in a hospital or a doctor's office," Maria told the doctor quietly. And now, here she was again, back in the hospital. She'd been perfectly healthy until just after last Christmas. She'd come home from college to see her family and hang out with her friends, and as she prepared to head back to school this strange queasiness had come over her. She couldn't eat. Any odor-especially food-made her feel as if she might vomit. But she didn't. Not at first.

The next day, on the drive back up to school, she'd suddenly broken into a cold sweat and had to pull over to vomit. And once she got started, it seemed like she would never stop. "I don't know how I made it to school because it seems like I had to get out of the car to throw up every few minutes."

Back at school she spent the first few days of the semester in bed. Once she was back in cla.s.s her friends joked that she was just trying to get rid of the extra pounds from the holidays. But she felt fine and she wasn't going to worry about it.

Until it happened again. And again. And again.

The attacks were always the same. She'd get that queasy feeling for a few hours, and then the vomiting would start and wouldn't let up for days. There was never any fever or diarrhea; no cramps or even any real pain. She tried everything she could find in the drugstore: Tums, Pepcid, Pepto-Bismol, Prilosec, Maalox. Nothing helped. Knowing that another attack could start at any moment, without warning, gnawed insistently in the back of her mind.

She went to the infirmary with each attack. The doctor there would get a pregnancy test and when it was negative, as it always was, he'd give her some intravenous fluids, a few doses of Compazine (a medicine to control nausea), and, after a day or two, send her back to the dorm. Halfway through the semester she withdrew from school and came home.

Maria went to see her regular doctor. He was stumped. So he sent her to a gastroenterologist, who ordered an upper endoscopy, a colonoscopy, a barium swallow, a CT scan of her abdomen, and another of her brain. She'd had her blood tested for liver disease, kidney disease, and a handful of strange inherited diseases she'd never heard of. Nothing was abnormal.

Another specialist thought these might be abdominal migraines. Migraine headaches are caused by abnormal blood flow to the brain. Less commonly, the same kind of abnormal blood flow to the gut can cause nausea and vomiting-a gastrointestinal equivalent of a migraine headache. That doctor gave Maria a medicine to prevent these abdominal "headaches" and another one to take if an attack came anyway. When those didn't help, he tried another regimen. When that one failed, she didn't go back.