In 2019 video visits with physicians were available on a limited basis but represented a small slice of total patient–physician encounters. Then came the pandemic, and because virtual, or “telehealth,” appointments were such a good way to ensure the safety of patients and medical staff, regulations were loosened and payors such as Medicare and private insurance raised compensation for virtual visits to parity with in-person ones. By April 2020 the percentage of doctor visits that were online in the United States was 86 times what it had been a year earlier. By that May, 95% of all patient encounters at Johns Hopkins’s neurology department were online. I had my first-ever virtual doctor visit during that time, but notwithstanding the convenience and safety, at no point was I warned that I was giving up one of the three pillars of diagnosis. In a video visit my doctor could still talk with me (the first pillar), he could still order tests (the second pillar), but he could not reach through a screen to look, listen, and feel for objective clues to my condition—the physical examination.

One reason the transition to telehealth happened so quickly was that the health care industry was already headed that way. In 2015 a $54 million facility billed as the world’s first “hospital without beds” opened in Chesterfield, Missouri, founded on the idea of improving patient access to specialized medicine and, in some cases, moving patient care out of the institutional setting and into the home. The latter scenario uses remote monitoring devices and a network of roving health service providers who may or may not be employees of the virtual hospital. If things get worse, the model calls for a mobile emergency response team or an ambulance ride to a brick-and-mortar hospital.

By the time the Missouri program was launched, “direct to consumer” pharmacy start-ups such as Lemonaid had begun peddling prescription drugs online without an in-person visit or any previous relationship between the prescribing doctor and patient. Patients, particularly younger, more tech-savvy ones, were already conducting a lot of their activities online. A Norman Rockwell world of family physicians holding tongue depressors and patients saying “Ahhh” was ripe for disruption, and Covid-19 simply pulled the trigger.

Patients responded favorably to the convenience of virtual visits, though at least initially many still preferred in-person ones. In surveys, some clinicians questioned the accuracy of online diagnosis. One reported that prior to the pandemic, about three-quarters of all records of primary care visits included the patient’s blood pressure; by June 2020, without the typical taking of vitals at the beginning of office visits, less than half of them did. This was not insignificant. Hypertension is a leading cause of death that could more than double a patient’s risk of hospitalization for severe Covid. Of course, remote sensing of vital signs is one of the suite of services in the “hospital without beds” model pioneered in Missouri, but pending such a technological patch for routine online doctor visits, the surveys revealed a larger weakness of a video chat: its selective focus on patients’ subjective accounts of “symptoms,” to the exclusion of a medical professional’s physical examination for what are known in medicine as “signs.”

The distinction between symptoms and signs is taught at every level of medical training, from emergency medical technician classes to elite medical schools. Symptoms are the patient’s subjective experience of their body. Signs are objective observations a medical care provider can see, hear, feel, and even smell.

Attention to signs begins when a provider knocks and enters the exam room. Does the patient seem alert and well oriented? Does she move with ease or difficulty? Is her skin sallow or pale, her breathing noisy or labored? The inquiry continues in a formal examination during which the medical professional surveys the body; feels parts of it for signs of injury or illness; listens through a stethoscope to the lungs, heart, and bowel; shines a little flashlight into the eyes, throat, and ears; asks the patient to grip their hands; and checks tendon reflexes with a little rubber hammer. The exam can give evidence about the patient’s chief complaint, but it can also reveal something the patient is unaware of or disclose something the patient would rather not discuss.

If I had not been able to listen to his heart and lungs, and examine his jugular vein and lower extremities, I may have put too much weight on the patient’s lack of concern and missed the diagnosis.

“A thorough physical examination provides a measure of objectivity that can help me rethink a patient’s narrative,” observed Paul Hyman, a Harvard-trained pediatrician and internist, in a 2020 essay in the Journal of the American Medical Association. As telehealth visits soared during the lockdown, Hyman had already been worrying about the extinction of the physical exam, which, even before telehealth, some physicians were skipping under pressure from the number of patients seen and the limited time they had with them.

Although board certified in two specialties, Hyman practices as a primary care doctor in Brunswick, Maine. His office is in what was formerly a small suburban hospital converted into a clinic. Outside is a pine forest and a duck pond, and he can smell the conifers and hear the ducks as he walks from the parking lot. About a third of his patients are children. Of the adults, some work in the shipyards at Bath. Others sew the rubber bottoms onto Maine Hunting Shoes at L.L.Bean. Some are lobster fishers; some, professors at Bowdoin College. A lot are retired people from the cities.

Maine is well known for its long, cold winters. The state has its share of stoics, Hyman observed in his JAMA opinion piece. One of his patients reported that he was feeling a bit tired, but guessed it was nothing. He was probably working too hard, he said. However, when Hyman conducted his physical examination he found signs that the man had congestive heart failure.

“If I had not been able to listen to his heart and lungs, and examine his jugular vein and lower extremities, I may have put too much weight on the patient’s lack of concern and missed the diagnosis,” reported Hyman. Objective data from the physical exam can also mediate between what the patient wants and what a doctor judges to be efficacious, he argues, as in the case of a patient with a respiratory infection asking for antibiotics that in Hyman’s view are not indicated.

“If I communicate that results of their lung examination are clear and that their oxygen saturation levels are normal, they often feel more reassured that they do not need medication,” he wrote.

Hyman’s anecdotal observations are supported by a 2019 study of doctor visits for acute respiratory infections in children. The study compared outcomes for patients seen in person by primary care providers, those seen at urgent care clinics, and those receiving care from “direct to consumer” (DTC) telehealth companies that give medical advice and treatment online without a prior doctor–patient relationship. In this study, reported in Pediatrics, 52 percent of the DTC encounters resulted in prescriptions for antibiotics, while only 31 percent of the primary care visits did. Urgent care was in the middle. Moreover, the telehealth doctors were far more likely to prescribe antibiotics outside established guidelines—perhaps as a precaution in the absence of prior knowledge of the patient and objective data from a physical exam.

Whatever the trade-offs of telehealth, patients will go where they can be seen in the face of a growing shortage of primary care doctors. A growing and aging population is creating more demand for medical services even as physicians, too, grow older and reach retirement. Direct-to-consumer telehealth is ready to take up the slack, offering nearly instantaneous online access to a physician somewhere who functions like a driver for a ride-sharing app. Venture capital has been pouring into telehealth. Big players like CVS, Walmart, and Amazon are jumping into the market. Hyman gets a lot of unsolicited email asking him to come to work in DTC telehealth operations.

For all his concerns, Hyman says telehealth has its place. “For patients who live in a rural area where there’s no primary care physician, if telehealth enables them to receive care in a certain way, that could be a good thing,” he told me during a phone call. And beyond primary care, there are medical specialties that are, and probably always will be, unrepresented in remote, less populated regions. A recently retired neonatologist I talked to marveled at his ability to perform telehealth consultations for tiny patients in remote Alaskan hospitals from an intensive care nursery unit in Anchorage. There are other, more subtle benefits. For example, the onscreen view of a patient’s home could afford the physician new insight into how the patient lives.

However, physical touch and observation have their place too, and if they may soon be carried out by robots, we may wonder if that’s really what we want. The stoical Mainer in whom Hyman saw signs of heart failure during a physical exam was treated and survived. When I last spoke to Hyman, doctor and patient were about to get together in person.