Every Patient Tells a Story (4 page)

More Than Just the Facts

Dr. Hsia was a resident in Yale’s Primary Care Internal Medicine residency training program, where I now teach. She told me about Maria Rogers because she knew I collected interesting cases and sometimes wrote about
them in my column in the
New York Times Magazine
. In thinking about this case, Amy told me she knew from the start that if she was going to figure out what was causing this patient to suffer so, it wasn’t going to be because she had greater knowledge—because Maria Rogers had already seen lots of experts. No, if she was going to figure it out, it would be because she’d find a clue that others had overlooked.

The patient’s story is often the best place to find that clue. It is our oldest diagnostic tool. And, as it turns out, it is one of the most reliable as well. Indeed, the great majority of medical diagnoses—anywhere from 70 to 90 percent—are made on the basis of the patient’s story alone.

Although this is well established, far too often neither the doctor nor the patient seems to appreciate the importance of what the patient has to say in the making of a diagnosis. And yet this is crucial information. None of our high-tech tests has such a high batting average. Neither does the physical exam. Nor is there any other way to obtain this information. Talking to the patient more often than not provides the essential clues to making a diagnosis. Moreover, what we learn from this simple interview frequently plays an important role in the patient’s health even after the diagnosis is made.

When you go to see a doctor, any doctor, there is a very good chance that she will ask you what brought you in that day. And most patients are prepared to answer that—they have a story to tell, one that they have already told to friends and family. But the odds are overwhelming that the patient won’t have much of an opportunity to tell that story.

Doctors often see this first step in the diagnostic process as an interrogation—with Dr. Joe Friday getting “Just the facts, ma’am,” and the patient, a passive bystander to the ongoing crime, providing a faltering and somewhat limited eyewitness account of what happened. From this perspective, the patient’s story is important only as a vehicle for the facts of the case.

Because of that “facts only” attitude, doctors frequently interrupt patients before they get to tell their full story. In recordings of doctor-patient encounters, where both doctor and patient knew they were being taped, the doctor interrupted the patient in his initial description of his symptoms over 75 percent of the time. And it didn’t take too long either. In one study
doctors listened for an average of sixteen seconds before breaking in—some interrupting the patient after only three seconds.

And once the story was interrupted, patients were unlikely to resume it. In these recorded encounters fewer than 2 percent of the patients completed their story once the doctor broke in.

As a result, doctors and patients often have a very different understanding of the visit and the illness. Survey after survey has shown that when queried after an office visit, the doctor and patient often did not even agree on the purpose of the visit or the patient’s problem. In one study, over half of the patients interviewed after seeing their doctor had symptoms that they were concerned about but did not have a chance to describe. In other studies doctor and patient disagreed about the chief complaint—the reason the patient came to see the doctor—between 25 to 50 percent of the time. This is information that can come only from the patient and yet, time after time, doctors fail to obtain it. Dr. George Balint, one of the earliest writers on this topic, cautioned: “If you ask questions you will get answers, and nothing else.” What you won’t get is the patient’s story, and that story will often provide not only the whats, wheres, and whens extracted by an interrogation, but often the whys and hows as well.

Moreover, the interrogation model makes assumptions about the elicited symptoms and diseases. And while these assumptions might be true for most of the people with those symptoms, they may not be true for this particular individual. The great fictional detective Sherlock Holmes talks at length about the difference between the actions and thoughts of the individual when contrasted to the average. Holmes tells Watson that while you may be able to say with precision what the average man will do, “you can never foretell what any one man will do.” The differences between the average and the individual may not be revealed if the doctor doesn’t ask.

“It is much more important to know what kind of patient has the disease than what sort of disease the person has,” Osler instructed his trainees at the turn of the twentieth century. Even with all of our diagnostic technology and our far better understanding of the pathophysiology of disease, research suggests this remains true.

So getting a good history is a collaborative process. One doctor who writes frequently about these issues uses the metaphor of two writers collaborating on a manuscript, passing drafts of the story back and forth until both are satisfied. “What the patient brings to the process is unique: the particular and private facts of his life and illness.” And what the physician brings is the knowledge and understanding that will help him order that story so that it makes sense both to the doctor—who uses it to make a diagnosis—and to the patient—who must then incorporate that subplot into the larger story of his life.

If getting a good history is so important to making an accurate diagnosis, why are we so bad at it? There are several reasons.

First, most researchers, doctors, and patients would agree that time pressures play an important role. A visit to a doctor’s office lasts an average of twenty-two minutes. Although there is a sense that doctors are spending less time with their patients, that number has actually increased over the past twenty years. In 1989, the average doctor’s appointment lasted only sixteen minutes. Despite this extra time, both doctors and patients frequently agree that their time together is still too short.

In response, doctors often depend on a few highly focused questions to extract the information they think will help them make a diagnosis quickly. Yet it is clear that this effort to reduce the time it takes to get a good history increases the risk of miscommunication and missed information. Like so many shortcuts, this information shortcut often ends up taking more time than those interviews in which patients are able to tell their stories in their own ways.

Studies suggest that getting a good history allows doctors to order fewer tests and make fewer referrals—without taking any more time. Indeed, some studies suggest that obtaining a good history can even reduce visit time. In addition, patient satisfaction is higher, adherence to therapy is higher, symptom resolution is faster, lawsuits are less frequent.

Lack of training may also contribute to the problem. Doctors spend two years in classrooms learning how to identify and categorize disease processes, matching symptoms to known disease entities, but until recently very few
programs offered any training on how to obtain that essential information. The assumption seemed to be that this did not need to be taught. And there may have been an unspoken expectation that our improved diagnostic technology would reduce our dependence on this kind of personal information. Studies have shown that neither assumption is true, and now most medical schools offer classes in doctor-patient communication. Moreover, since 2004, medical students are required to demonstrate proficiency in their history-taking skills in order to become licensed physicians. A new generation of physicians may not use these tools, but at least they have them.

Finally, many doctors are uncomfortable with the emotions that are sometimes associated with illness. When patients present their stories, they often look for cues from the doctor as to what type of information they should give. The interrogation format tells the patient that what’s needed from them are the facts and only the facts. And yet illness is often much more than a series of symptoms. The experience of being sick is frequently interlaced with feelings and meanings that shape and color a patient’s experience and perception of a disease in ways that are unimaginable, and unanticipated, by the doctor. A family history of heart disease or cancer may lead a patient to minimize a symptom. I recently got a phone call from a friend, a man in his late fifties whose father had heart disease. My friend was having chest pains when he walked up a hill. He wondered if this was his childhood asthma returning. He was shocked when I suggested he see a cardiologist. He had two blocked arteries, which were opened with complete resolution of his pain. The same history might cause another to focus on a symptom well beyond its actual severity. I have a few patients who have had many stress tests because of their concerns over their chest pain. The fact that previous tests have not shown heart disease provides them with no comfort or reassurance. Financial concerns may likewise affect how patients tell their stories.

Worries about the social meaning of symptoms can complicate even a straightforward diagnosis. I learned this the hard way. A patient I saw when I was a resident came for a school physical. She was young and healthy. As I was finishing up and preparing to move on to the next patient, she suddenly
asked me about a lesion on her buttocks. Could it be from doing sit-ups on the hard floor? she asked somewhat anxiously. I quickly looked at the lesion. It looked like a small blister, located in the cleft between the buttocks. Certainly, I reassured her, glancing at my watch. I noticed that she seemed worried about the lesion, but I didn’t ask any further questions or do a more thorough exam because I was running behind schedule. Only when the lesion reappeared months later did she acknowledge that her boyfriend had had a breakout of genital herpes on a vacation they’d taken together and she hadn’t insisted on his use of a condom. The reappearance of the lesions made herpes the likely culprit. I completely missed a straightforward diagnosis because I was too rushed to address her anxiety and she was too embarrassed to offer this other history. It happens all the time.

Everybody Lies

Several years ago I got a call from a producer named Paul Attanasio. He had created a television show based in part on my column in the
New York Times Magazine
and wanted to know if I would be interested in being a consultant for this new show. It was a drama, he told me, about an ornery doctor who was a brilliant diagnostician. I agreed to work on the show, thinking that it wasn’t going to last long. The show, called
House M.D.
, quickly found an enthusiastic audience.

In this show, Dr. Gregory House doesn’t value patient history. Indeed, he frequently tells his trainees that they should not believe a patient’s version of his illness and symptoms, because “Everybody lies.” In the context of the show, there is a certain truth in that. Patients frequently lie to House and sometimes his staff—not because the patients are intrinsically deceitful but because of who House is. As portrayed (brilliantly by Hugh Laurie), House is far from the kindly and gentle doctor whose presence invites trust and confidences. Instead, he is narcissistic and arrogant, a drug addict, and something of a pedant. He is a darker, more bitter version of Conan Doyle’s brooding detective Sherlock Holmes. House’s demeanor tells patients that
the feelings and meanings illness may have for them are not important and so they don’t tell him about them. As a result, House often gets only part of their story.

The mystery is solved only when the rest of their story is revealed—either from evidence found when his staff break into the patient’s home (a quirky twist on getting a thorough history) or when the patient is finally forced to reveal his hidden truths. House acknowledges the importance of a thorough patient history but concludes that the problem is the patient who lies rather than the doctor who fails to establish a relationship in which difficult, embarrassing, or distressing truths can be told.

Amy Hsia knew from the start that if she was going to figure out the cause of Maria Rogers’s cyclic episodes of vomiting, it would be because of some key piece of history that she was able to get that others had overlooked. But sitting outside the patient’s room that afternoon, she wasn’t sure she’d found it. She went through the thick charts, reading the notes and test results collected by all of the other doctors involved in the same exercise in previous hospitalizations, trying to make it all make sense. Nothing leaped out at her. The sketchy description of the symptoms and history provided nothing she hadn’t already found out from the patient herself.

Hsia considered the differential diagnosis once more. Nausea and vomiting have a very long list of causes: ulcers, gallstones, obstruction, infection. Hepatitis, pancreatitis, colitis, strokes, and heart attacks. None of them seemed to fit in a case of a young woman with multiple episodes of vomiting and lots and lots of tests showing no abnormalities. Maybe she wasn’t going to be able to figure this patient out either. She ordered a new medication to relieve the nausea and then moved on to see her next patient.

The next morning, when Dr. Hsia, her supervising resident, and the attending physician—the troika of the modern hospital medical patient care team—visited Ms. Rogers, the girl’s bed was empty. The sound of the shower told them where she’d gone. That caught the young resident’s attention. When she had come by a couple of hours earlier to examine the girl, she’d
been in the shower then too. She remembered that Rogers had told her that her nausea improved when she took a shower. What kind of nausea didn’t get better with the traditional nausea medications—by now she’d been on most of them—but improved with a hot shower?

Hsia posed the question to the team. Neither had heard of such a syndrome. So, after the team had finished seeing all the patients they were caring for, Hsia hurried to find a computer. She went to Google and entered “persistent nausea improved by hot showers.” She hit the enter key and less than a second later the screen was filled with references to a disease Hsia had never heard of: cannabinoid hyperemesis—persistent and excessive vomiting (hyperemesis) associated with chronic marijuana use (cannabinoid).