Zain Khawaja, an emergency medicine resident at Northwestern University, begins his First Opinion essay in STAT with a scene any emergency physician would recognize. A woman arrives with what looks like an oncologic emergency, and the room is tight with fear. He introduces himself and shakes her hand, then the hands of her family. The tension visibly eases. The patient smiles. What he is describing is not a nostalgia piece about bedside manner. It is an argument that medicine discarded one of its cheapest, oldest tools and has never replaced it.
The essay's title asks medicine to start shaking hands again, and the response it will get is predictable. One side will talk about humanity; the other side will talk about germs. Both sides will miss what the evidence actually says, which is that the handshake was never the thing patients were asking for.
The evidence is old and consistent
The study Khawaja cites has been sitting in the literature for almost two decades. In 2007, Gregory Makoul and colleagues at Northwestern surveyed 415 adults about greeting preferences and reviewed recordings of 123 outpatient visits. More than 78 percent of patients wanted their physician to shake their hand, and in the recorded visits doctors and patients shook hands in about 83 percent of encounters. Patients also had preferences about names: about half wanted their first name used, and a majority wanted physicians to introduce themselves by first and last name. In more than half of the recorded visits, the doctor never used the patient's name at all.
The study's age is the point. The preference it documents outlived the practice it documents. Before the pandemic, the handshake was near universal and mostly unexamined. Then Covid arrived, and with it distancing, gloves, gowns, and a defensible fear of transmission, and the gesture disappeared from most encounters without any formal decision to remove it. It simply did not come back. The preference, by every available measure, did not change.
The preference has detail in it
The 2007 numbers are worth reading in full, because they describe a greeting rather than a ritual. The 78 percent who wanted a handshake were the majority, not the totality: 18 percent did not want one, and the rest were unsure, which is exactly why the study's authors told physicians to offer the gesture while watching for nonverbal cues that say no. Older patients wanted the handshake less than younger ones, at 74 percent against 87 percent, a reminder that a single gesture lands differently across a waiting room. The name findings cut the same way. About half of patients wanted their first name used, a minority preferred the last name, and the rest wanted both, which is a preference distribution, not a mandate.
What the videotapes showed is the gap between the ritual and the greeting. In the 123 recorded visits, doctors and patients shook hands 83 percent of the time. But in half the visits the physician never used the patient's name, and in more than one in ten the physician never introduced himself at all. A handshake with no names attached is a transaction. A name with no handshake is still an acknowledgment. The study's real subject was never the grip. It was whether the person in the gown gets recognized as a person, and the data show the handshake was merely the most common vehicle for that recognition, not the only one.
Why the handshake became the battleground
The hygiene case against the handshake is real, and Khawaja concedes it rather than dodges it. A handshake is inappropriate when isolation precautions are in place, when a patient is significantly immunocompromised, or when either person may have a transmissible illness. Hospitals contain all three conditions constantly, which is why blanket policies took hold so easily. The counterargument is not that germs do not matter. It is that the risk can be managed with the same hand hygiene and clinical judgment that already govern every other form of contact in a hospital, and that eliminating the greeting while retaining everything else was never an evidence-based decision, just an easy one.
That is where the debate usually ends, with humanity on one side and infection control on the other, each side quoting its own intuitions. The problem is that the battleground is wrong. The handshake is a proxy.
The benchmark was never the handshake
Khawaja himself supplies the sentence that dissolves the argument: "The goal here is not the handshake itself." What the patient wants, his essay argues, is deliberate human connection, a moment in which the person behind the chart is acknowledged by name and by touch. The handshake happens to be the most efficient instrument for that. It is not the only one.
The best evidence for this comes from a study about furniture. Researchers examining outpatient encounters found that placing a chair near the bedside made physicians more likely to sit during visits, and that sitting was associated with higher patient satisfaction and better communication scores, without adding a minute to the encounter. Nobody asked for a chair. What patients experienced was a physician who stayed, at eye level, for the conversation. The structural change produced the same class of benefit the gesture produces: the felt sense that the doctor is present with this patient, now, on purpose.
That is the benchmark. Not the handshake, not the chair, not the smile. The deliberate moment. The instruments are interchangeable; the thing being measured is not.
Khawaja says as much with his own list of substitutes: a warm verbal greeting, a hand placed over one's own chest, eye contact, the patient's name used on purpose. He extends the responsibility beyond the doctor to every member of the care team, which is the practical version of the point. A physician can do everything right in the first minute and lose the encounter to a cold registration desk or a brusque discharge. Connection is not a doctor-only deliverable. It is a property of the whole encounter, and the handshake is one instrument in an orchestra that includes everyone who speaks to the patient that day.
The real competitor to connection is not germs
The essay's own list of obstacles is the giveaway. Khawaja names the electronic medical record, rising patient volumes, frequent interruptions, and pressure for rapid decisions. Not one of those is an infection control policy. The competition for connection in a modern encounter is not hygiene. It is the screen, the census, and the clock.
A physician can eliminate every handshake in the hospital and still fail to connect, and can sit, make eye contact, and use the patient's name without touching anyone and succeed. The infection control debate is real but narrow. The attention problem is broad, and it is the one that actually determines whether patients feel seen. The handshake fight is loud because it is easy to have. Redesigning encounters so that a clinician can afford sixty seconds of presence is a structural problem, and structure does not argue well in essays.
The economics point the same way. The electronic record demands documentation, the schedule demands throughput, and the documentation and the throughput together consume the minutes that connection used to live in. A handshake takes four seconds and requires no template redesign, which is part of its appeal as a proposal. But the same is true of the chair study's finding: a piece of furniture, placed deliberately, changed the whole geometry of the visit at no time cost at all. When a chair moves the needle as much as a gesture, the problem was never the gesture. It was that nobody had built the room, or the schedule, or the workflow, around the patient being the point of it.
Khawaja's closing thought, in his own words, is that the more technologically advanced medicine becomes, the more intentional it must be about preserving its human elements. That is the actual proposal, and it survives the loss of the handshake entirely. Warm verbal greeting, a hand placed over one's own chest, a moment of eye level contact: the point of every alternative he lists is that the connection is the requirement and the gesture is merely the delivery system.
The demand underneath the gesture
The patient in the emergency department did not arrive with a request for a specific ritual. She arrived terrified, and the handshake was the instrument that told her, faster than words, that she was being treated as a person experiencing one of the most frightening days of her life. What she wanted was that message. Any deliberate gesture that carries it will do.
So the argument about whether to bring back the handshake is misaimed, and both sides can be right without settling anything. The handshake can stay gone and connection can still be restored, through visit design, eye contact, names, and time. The handshake can come back and connection can remain absent, if the gesture returns alone while the screen keeps the doctor's eyes. The question worth arguing is not what patients' hands should do. It is what the doctor's attention should do, and the research, old and new, keeps answering the same way: attention, on purpose, is the product.
Primary sources
- Zain Khawaja's First Opinion essay in STAT for the emergency department anecdote, the argument for reviving the handshake, the infection control caveats, and the alternatives he lists.
- The 2007 greeting preferences study by Gregory Makoul and colleagues at Northwestern, published in the Archives of Internal Medicine and covered by ScienceDaily, for the survey and videotaped visit figures.
- The 2023 chair-and-sitting study published through PubMed Central for the outpatient encounter figures.