During my medical residency, I spent a month on the obstetrics ward, learning how to deliver babies. On one occasion, I was paged to a delivery room down the hall. Inside, I saw the mother-to-be propped up in bed, her husband sitting next to her. Then I noticed the crowd. In addition to the obstetricians and delivery nurses, a special team of nurses and pediatricians from the neonatal I.C.U. had gathered.
“Is this a high-risk delivery?” I asked one of the obstetrics residents.
“Maybe,” he whispered back. “The baby had a positive screen for cri du chat.”
Cri du chat, which means “cat’s cry” in French, is a rare genetic syndrome affecting one in fifteen thousand to fifty thousand infants. It was discovered by Jérôme Lejeune, a French geneticist, in 1963. Early in fetal development, genes situated on the short arm of chromosome 5 are accidentally left out, like parcels falling off the back of a delivery truck; affected children can have poor muscle tone, low birth weight, a cleft palate, unusually small heads, speech delays, learning disabilities, and heart problems. As many as one in ten children with cri du chat don’t survive the first year of life. The anatomy of the vocal cords is often affected, as well, resulting in a highly abnormal cry that sounds startlingly like the mewing of a cat.
Because the syndrome can cause such significant handicaps, pregnant women with risk factors are often offered a screening test. By sampling fragments of fetal DNA that have migrated from the placenta into the mother’s blood, doctors can sometimes detect the fingerprints of a genetic disorder. But, because the test doesn’t inspect the infant’s DNA directly, it can’t provide a perfect picture of fetal genes and chromosomes; it’s more like a slightly blurry snapshot. It’s possible for a result to be incorrect—falsely positive or falsely negative. If the screening test does come back positive, another test is needed to confirm the diagnosis. This second test is more invasive, and involves taking samples of fetal or placental tissue. And yet, although it’s more accurate than the screening test, it’s not perfect, either. Even the confirmatory tests can sometimes be misleading.
As we stood near the back of the room, the obstetrics resident told me that the laboring woman had undergone some of these more advanced tests. The results had been reassuring, and the team had told the parents that the fetus was most likely normal. But the tests couldn’t say for sure, and neither could we. The neonatal-I.C.U. team was there on standby because, despite our most sophisticated tests, the nature of the child would remain a mystery until the moment of its birth.
In the room, that moment unfolded in stages. First, we glimpsed the child’s head; then the shoulders, then limbs, then a body. The room was strung on a wire as the crowd of nurses and doctors watched the delivery in silence. The obstetrician lifted the newborn free of the birth canal. The cord was clamped and cut, and the newborn opened its mouth, filled its lungs, and cried. The exhausted mother beamed as she held her child in her arms. She didn’t hear what everyone else heard. The cry was not a normal cry. Shrill and plaintive, it sounded like the mewing of a cat.
Some medical problems are obvious. When the sidewalks in Boston are icy, I see a lot of patients who’ve lost their footing and fallen onto their outstretched hands; I can usually tell at a glance if they’ve broken a wrist. But most medical problems aren’t obvious. They arise from hidden processes that occur within the body, in tissues, cells, enzymes, or genes. They manifest only indirectly, through symptoms or signs. As the American Medical Association noted, in 1912, internal medicine is concerned primarily with “abstract problems” and the “intangible struggle against unseen infections.” Choosing an appropriate treatment depends on discerning the cause of an illness—and yet there are a number of possible causes for most symptoms. How do doctors connect a symptom to a cause?
The answer, of course, is that we test. To test is to examine something critically, to put it to the proof. The word is often thought to derive from the Latin testari, meaning to testify. But the Reverend Abram Smythe Palmer, a respected nineteenth-century lexicographer, placed its origins with the Old French test, which referred to a vessel used for cupellation, the extraction of precious metals with heat. “ ‘To test’ a thing,” Palmer wrote, in 1882, “is properly to submit it to the crucible or melting pot, to assay the quality of its metal.” The emergency department often feels like a crucible, where we approach our patients, undifferentiated, as a minter approaches ore. We apply our tests to find out what lies within.
Today, amid the coronavirus pandemic, we are thinking about medical tests more than usual. Often, we have a fairly simple vision of how tests work. We picture them as high-tech and definitive; we see them cutting through the ambiguities of an often asymptomatic virus. We hope that, by speedily distinguishing between the sick and the well, tests might help us establish defensive cordons around schools, workplaces, and public events. Colleges and universities have used coronavirus tests to sort students into different dormitories as they return to campus. Sports teams, too, have created testing-based “bubbles” within which they hope something like normal life can go on. We envision simple steps—a nasal swab, a sample tube, an expensive machine—followed by bad news or an all-clear.
But physicians tend to approach testing more cautiously, and in an incremental fashion. In fact, we are always testing, often in ways that don’t involve technology. One of our most important tests is one of our simplest: the visual assessment, what we call “eyeballing” a patient. There’s a double meaning to the statement “the doctor will see you now”; just laying eyes on someone can yield a huge amount of information. We can quickly tell whether a patient is critically ill or stable; we can often recognize critical illness from the way someone looks in a doorway. We may not know the cause, but we can sense the severity.
A lot of testing happens through language. Around 100 A.D., Rufus of Ephesus, a Greek physician, published the first treatise on taking a medical history; he described aspects of the patient interview that medical students still learn today, such as asking about the location, duration, and character of pain. I learned many of these principles in medical school but didn’t realize until my residency that interviewing patients is actually a way of testing them. “Think about it this way,” one of my supervisors said. “When you question a patient about their symptoms, do their answers influence your suspicion of potential causes?” They do, just as the results of a blood test would.
In the centuries after Rufus, doctors pioneered new ways of testing the body. “Water casting,” or inspecting urine, became the diagnostic test of choice in medieval Europe. The Jerusalem Code of 1090 made doctors liable to public beatings if they failed to examine it. A blood-pressure measurement was taken for the first time in 1733, when an English clergyman inserted a brass pipe into the artery of a horse (he found that the animal’s blood pressure rose by a factor of four when it began struggling). In the seventeen-fifties, Leopold Auenbrugger, an Austrian physician, developed a groundbreaking technique called percussion, which is still in use today. After observing his father tapping wine casks to determine how full they were, he realized that a similar method could be employed to localize diseases, such as pneumonia, within the body of a living patient. He discovered that a healthy lung, when rapped with a couple fingers, sounded like “a drum covered with a thick woolen cloth,” whereas a diseased region was “entirely destitute of the natural sounds.”
