by Lisa Sanders
“Excellent,” agreed Braverman. Linda Friedlaender, curator of education, spoke briefly about the painting (The Death of Chatterton, Henry Wallis’s rendition of the suicide of the seventeen-year-old poet of the eighteenth century, Thomas Chatterton) and then they moved on to the next painting.
After the class, Braverman and I talked over coffee about his innovative teaching technique. “Astute observational skills are usually acquired only after several years of being in medical practice,” Dr. Braverman said. “Suddenly, all of the accumulated experience leads doctors to see things they have not been taught before. They become terrific observers—eventually. With this course, I hope to jump-start these special diagnostic skills right from the beginning.” Even though they’re looking at paintings, not patients, what they learn here can be applied to medicine, Braverman asserted.
He knows this because he tested it. For two years Braverman had participants write a description of what they saw in a dozen photographs picturing individuals with visible abnormalities. After the class they were given a different set of photographs with the same instructions. Tests were scored based on the description of specific aspects of the photographed abnormalities. Correctly identifying the disease or condition did not affect the score; identifying and describing the visual data was all that counted. Before-and-after test scores were compared, and students improved by an average of 56 percent after spending this afternoon in the museum.
To ensure that this was not simply due to better test-taking skills the second time around, the same two-part test was given to a group of students before and after a lecture on physical examination. These students also improved—you don’t get to medical school if you can’t learn how to take a test—but not nearly as much.
Even before I heard about this study, I knew from personal experience that these skills could be taught. I was in my third or fourth year of medical school when I suddenly started seeing people with abnormalities everywhere. It was as if I had suddenly been transported into a world populated with the ill, the injured, the aberrant. Of course they were there all along—why hadn’t I seen them? Certainly knowledge plays a role. When you learn a new word or name, it suddenly seems to be everywhere.
But it’s more than that. We are trained from a very early age to avert our eyes from abnormalities. Children are fascinated by people whose appearance differs from what they’ve come to expect. And we teach them to ignore that interest. My daughter, Tarpley, once asked a cashier if she was a man or a woman. My husband flushed with shame for the discomfort it caused the homely, hirsute woman. He apologized but recognized that the damage was done. Afterward he explained to our daughter just how much that kind of comment must have hurt the woman. She doesn’t ask those kinds of questions anymore. She’s learned not to stare.
Medical school forces you to undo that training. You mustn’t avert your eyes from abnormality. You need to seek it out. You need to figure it out. And it doesn’t just turn off when you leave your office. I frequently (quietly, I hope) point out to my husband pathology that I see on the street—the rolling gait of a man with an above-the-knee prosthesis; the strange gray-toned tan of a man with iron overload syndrome, known as hemochromatosis, the schizophrenic woman’s restless lips and mouth, a long-term side effect of many antipsychotics. I now live in a world filled with abnormality. It’s fascinating.
How is it possible to see something without noticing it? Dr. Marvin Chun, a professor in the Visual Cognitive Neuroscience Lab at Yale, has devoted his career to trying to answer that question. When I visited him on a warm fall afternoon, he invited me to view a video already quite famous in his field of vision and attention. On a monitor I saw six adults standing in the midst of some strange game, their actions frozen by technology. There appeared to be two teams—one dressed in white, one in black. Each team had a basketball. Strangely, they weren’t on a court but in the corridor of an unidentified office building. Closed elevator doors were clearly visible in the background.
My task, once the video started, was to watch the white team and keep track of how many times the ball was passed between players—keeping separate counts of when it was passed overhead and when it was bounced from person to person. The image started to move and I kept my eyes glued to the white team’s basketball as it was passed silently among the moving mass of black and white bodies. I got up to six overhead passes and one bounce pass and I lost track. Determined not to give up, I kept going until the thirty-second video was complete.
Eleven overhead passes and two bounce passes? I ventured. I told Chun that I got a little confused in the middle. Despite that, I’d done a good job, he told me. I missed only one overhead pass. Then he asked, “Did you see anything unusual in the video?” Other than the unusual setting for the game, no, I saw nothing at all out of the ordinary.
“Did you see a gorilla in the video?”
A gorilla? No, I had definitely not seen a gorilla.
“I’m going to show you the video again, and this time, no counting, just look at the game.” He restarted the video. The white and black teams sprang back into action. Eighteen seconds into the game—around the time I lost my concentration—I saw someone (a woman, I find out later) in a gorilla suit enter the hallway court on the right. She strolled casually to the middle of the frame, beat her chest like a cartoon gorilla from a children’s TV show, then calmly exited out of the left side of the picture. Her on-camera business lasted eight seconds and I hadn’t seen her at all.
If you had asked me if I thought that I could miss a gorilla—or even a woman in a gorilla suit—strolling through the picture, I would have agreed that it was impossible to overlook such an extraordinary event. And yet I did. So did more than half of those who were given the same task by Daniel J. Simons in his lab at the University of Illinois at Urbana-Champaign. How is that possible?
We have tremendous faith in our ability to see what is in front of our eyes. And yet the world provides us with millions of examples that this is not the case. How often have you been unsuccessful in looking for an object and recruited the help of someone who finds it immediately right in front of you? Or had the embarrassing encounter with a friend who confronts you angrily after you “ignored” his wave the night before while scanning for an open seat in a crowded movie theater? According to the Federal Highway Administration, there are over six million car accidents every year. In many of these crashes, drivers claim that they had looked where they were going and simply hadn’t seen the object with which they collided—evidence that people are regularly capable of not seeing what’s in front of their eyes, what Sherlock Holmes might have called seeing without noticing.
Researchers call this phenomenon “inattention blindness” because we often fail to notice an object or event simply because we are preoccupied with an attentionally demanding task. Our surprise when experiencing this very common event derives from a fundamental misunderstanding of how the brain works. We think of our eyes like movie cameras, capturing all that is before us as we choose what to focus on at the moment. We might not be paying attention to everything, but we assume, first, that we will be able to recognize any important event that occurs and, second, that, if necessary, we can always rewind the movie and play it back in the theater of the mind. What we missed the first go-round would be noticed when we remembered the event.
Of course, that’s not how it works. When asked about the gorilla in the basketball game, I had no memory of the beast. I searched my memory but I didn’t remember him because I didn’t see him. My attention was directed elsewhere.
There are qualities that make an object more likely to be seen. Chun tells me that if a naked man or woman had walked into the frame instead of a gorilla, the chance that I would notice the unexpected image would be much higher. Or if the gorilla had been bloody, or if he had moved or acted like a gorilla, I would have been more likely to see him. That’s because there are some fundamental images that the mind recognizes as important.
So what
’s going on here? Clearly the information is traveling through the eyes to the retina. And a functional MRI—one that reveals which areas of the brain are working in any given task—shows that the neurological signaling is getting the information to the right part of the brain—so you’re definitely seeing it. But before this image can enter your awareness, another part of the brain jumps in to try to decide if this information is worthy of attention. And that judgment all depends on what you’re looking for.
As it turns out, most of the time we see what we want to see, what we expect to see. Our ability to see objects or events that are unexpected and dissimilar to those that we are looking for is extremely limited.
To go back to the experiment with the ballplayers and the gorilla, my task was to follow the white-clad players and keep track of how often they threw the ball. Most viewers given that task fail to notice the gorilla. In the same experiment, subjects instructed instead to follow the ballplayers wearing black did see the gorilla. Because the gorilla was also black, it was closer to what they were looking for and so the image was able to get past the brain’s gatekeepers and be noticed.
What happens to the visual information that enters the brain but doesn’t get the attention of the subject’s consciousness? Is it stored there, waiting for a second chance, like a delicious detail in a rerun episode of The Simpsons? Most research suggests not. If the sight doesn’t capture one’s attention initially, it’s gone forever.
Based on research like this, Chun and many other researchers in this area now believe that the expectations of the viewer are the primary shapers of what is seen, and that the unexpected will often be missed. We become better seers when we have better expectations. When you are given a specific task—follow the ball as it’s passed between members of the white team—you can predict what the expectations might be, and that observers are unlikely to see the passing gorilla because it’s not in their set of expectations.
What about in situations where you are looking but the task is more complex—the way it is in real life, or in the hospital taking care of patients? If their theory is true, what you see and what you don’t see will be shaped by what your experiences have led you to expect. Perhaps Osler was mistaken when he said that more diagnoses were missed because of not seeing than not knowing. Perhaps not knowing is what caused not seeing. Certainly that played a role in the case of Michael Kowalski.
Great Expectations
Michael Kowalski was not a man who was easily frightened. And he could count on one hand the number of times he’d cried as an adult. But when Dr. Keith Stoppard entered the room, he heard muffled, ragged breathing and, as his eyes adjusted to the dim light, he could see the massive man lying huddled in the bed. As unlikely as it seemed, Michael Kowalski, a fifty-two-year-old former college boxer, ex-army man, father of a marine, and all-around tough guy, was crying like a baby.
His wife, Maureen, a redheaded Valkyrie, stood by her husband’s bed. Her face was darkened with freckles and lined by fatigue as she tenderly placed a cool, wet cloth on her husband’s forehead. His short salt-and-pepper hair and unwaxed handlebar mustache lay plastered to his skin, and his round face was flushed and gleaming with the combination of sweat and tears. “Doc, I’m scared,” he said, his raspy voice nearly a whisper. “Can’t you tell me what’s wrong?” The woman squeezed her husband’s hand in silent reassurance.
Stoppard, a third-year resident, didn’t know what to say. He was worried. Mr. Kowalski had been in the hospital for three days and Stoppard was no closer to figuring out what was making him so sick than he had been on the day the patient had been admitted.
On that first day, this had seemed like a pretty straightforward admission: a middle-aged outdoorsman sent in by his regular doctor for what looked like Lyme meningitis. Stoppard had spoken to the patient’s doctor earlier that day and the case had seemed easy enough—get the lumbar puncture to confirm the diagnosis, then start intravenous antibiotics and watch him get better. But since then nothing had gone as he’d expected, and now he wasn’t sure what to think or what to expect.
It was nearly midnight when Stoppard saw Kowalski in the emergency room that first night. The patient told him he’d started feeling sick about a week earlier. At first, he’d figured it was just the flu. He’d felt tired, his body stiff and achy. “I was like an old man—I could barely get around,” he told the doctor in his low growl. But after two or three days of feeling lousy, he developed a strange, patterned fever. “You could set your watch by these fevers,” he explained. “Around four every afternoon I’d get real cold. I’d be shivering like mad. I’d load on the blankets but nothing I could do warmed me up. Then suddenly I’d be hot as hell. Sweaty. It was crazy.” His fevers would get up to 103°–104° every night, his wife, a nurse, added. By four in the morning he’d wake up drenched in sweat and have to change pajamas. By dawn, the fevers would subside—only to have the whole pattern repeat itself that afternoon.
Besides the fever, he told the resident, his neck felt stiff and painful, his head pounded, and a cough had left his throat raw. The joints in his legs, arms, and hands felt tight and sore. It was hard to move, to even get out of bed. Finally he went to see his regular doctor, Dr. Dennis Huebner. After hearing the story and examining him, Huebner figured it was probably just a virus but decided to send off blood to test for Lyme to be safe. He knew the patient was at risk for the disease. Kowalski was an avid outdoorsman, spending as many weekends as he could hunting and fishing just outside Old Lyme, Connecticut, where the disease was endemic.
The patient had pulled off many ticks over the years, he acknowledged. But, he added, none lately; he’d been too busy to get out to the woods the past few months. Still, Huebner considered Lyme one of the diseases you just don’t want to miss. If you get it early you can blast it with a week of antibiotics and it’s gone. Miss it and the patient may need months of care. Huebner told the patient it was probably just something going around and he should call if the fevers persisted. He’d let him know if the Lyme test came back positive.
That night the fever came, right on time, and the next day the patient called Huebner, who reluctantly started the patient on doxycycline. “He told me it was probably a virus,” the patient reported, “but I felt like I was sick enough to need antibiotics. And he was okay with that. I took the pills, but the fever just kept coming. After a couple of days, the doc says to me: ‘Look, you’re not getting better. You gotta go to the hospital.’”
The patient considered himself a pretty healthy guy. He’d done his time in the army “in the last war” (Vietnam), and now drove a truck for a local company. He had high blood pressure and his cholesterol was “worse than the doc says it should be,” but he took his medicines regularly and had felt well “until this crap started up.” On exam, in the emergency room, he had a fever of 103° and his heart was beating rapidly. The muscles of his neck were painful to the touch, but he could move his head freely. Just below his jaw he had several enlarged, painful lymph nodes. The joints in his hands and his knees were markedly tender but not red or swollen. Blood work sent by the ER showed an elevated white blood cell count and mildly abnormal liver enzymes.
The fever, painful neck, and pounding headache certainly pointed toward meningitis—a serious, potentially fatal infection. And untreated Lyme disease can progress to the brain, causing meningitis. But it wasn’t a perfect fit: as awful as this guy felt, he wasn’t as sick as the patients Stoppard had seen in the past with meningitis. With a fever this high, those patients were often too sick to talk. Despite the high fever, this patient was at times irritable, at other times funny, but very much awake and alert. The liver abnormalities weren’t typical either. Well, maybe it was a viral meningitis—its course is much less severe than its bacterial counterpart and sometimes could drive up liver enzymes. In any case, they’d need to do a lumbar puncture. That would tell them if this was a meningitis and, if so, what was causing it.
But when Stoppard recommended this procedu
re, the patient blew up. He was already sick, already in pain, and now these doctors he’d never met before wanted to stick a needle in his back? No way. He would have to talk to his doctor. The patient’s wife tried to persuade him but he was adamant: no procedures until he cleared it with his doctor—period. Huebner’s partner was on call that night—would he speak with him? The patient sat up in the ER gurney and glared fiercely at the young resident: he would speak to his doctor and no one else. Defeated, Stoppard added high-dose intravenous antibiotics to the doxycycline he was already taking and waited anxiously for the morning and the certainty of the lumbar puncture.
Stoppard reached the doctor the first thing the next morning and he immediately called the patient. He needed this procedure, Huebner told him. They had to know if this was meningitis. The patient agreed, reluctantly, and the uncomfortable test was done. The results came back almost immediately—they were normal. There was no evidence of an infection in his brain. The Lyme test sent by the doctor days before came back that morning as well—it was also normal. He didn’t have meningitis; he didn’t have Lyme disease. They were back at square one.
One technique doctors use to make a diagnosis is to group symptoms, physical exam findings, and lab data and identify which are the most important and use them to try to find a recognizable pattern. This patient had many symptoms, but which were most important? Stoppard felt that the fever was key—it was extremely high and had this very distinct pattern. He wasn’t so sure about the rest of them. But the fever, in combination with the enlarged lymph nodes and the elevated white blood cell count, clearly pointed to an infection. So where was this infection? What had they missed? Kowalski was on two strong antibiotics—but were they the right ones? At this point the team had no way of knowing. All they could do was keep looking.