Medicine has its own “have a nice day.” It is six words long, and most of us could say it in our sleep. Some of us probably have.
“Let us know if anything changes.”
It may be the only instruction in medicine that both people in the room hear as goodbye. That is harmless on every day but one. Some thirty times, nothing much was at stake. Today bleeding is. It sounds exactly the same.
He will be counseled about it out loud, in a kind voice, and it will be written in his chart. Then he will get the same goodbye as always, and he will nod. Nothing in the last thing he hears will tell him that today is different.
Real physician voices, twice a week
Free, and one click to unsubscribe.
The case is composite, and I have built it to be ordinary. A man in his seventies has come to the same office every three months for eight years. Every visit ends the same way. The physician stands, a hand goes to the door, and the sentence is said. He nods.
Today differs in one respect. He is leaving with a first prescription for an anticoagulant, and for the first time in years there is a particular change he must act on the day it happens. She knows this. Earlier in the visit she talked with him about bleeding, and the pharmacy will add a leaflet in six-point type.
Then the visit ends the way the others did. Nothing wrong was said. The sentence is accurate, kind, and what we are taught to say. The counseling earlier in the visit was the warning. But the close is the last thing he hears, and it gives today the same ending as thirty visits when nothing much was new.
What I think he hears is that the visit is over, and that it was like the others. I do not think he is at fault. Same words, same place, same moment, thirty times. Nothing in the thirty-first goodbye tells him it is different.
I think the physician has stopped hearing it too. A sentence said at every door becomes part of the door. She will chart that he was counseled, and she will be right.
The cost is easy to state. I am not reporting an outcome. Suppose he notices a black stool on a Saturday. He may remember the talk about bleeding. But it came in the middle of a visit that closed like every other, and the sentence at the door did not tell him that this is the “anything.” He has a leaflet and a nod. She has a note that says she warned him.
My hypothesis is that a sentence said in the same words on every occasion comes to be heard as the occasion, not as what it says.
Medicine has studied this kind of advice, under the name safety-netting. In 318 recorded U.K. consultations in a 2019 British Journal of General Practice study, it was given in about two-thirds, and was specific fewer than half the time. In the same group’s 2026 telephone study, patients recalled being given it in about two-thirds of the consultations where it was. I have not found a study of what repetition does to it.
The nearest research on repetition is about computer security warnings, in a 2015 brain-imaging study and a 2018 follow-up in MIS Quarterly. The brain’s response to an identical warning dropped sharply by its second exposure. Anyone who has ever clicked “I agree” will not need the brain scan. Warnings that kept changing appearance were tuned out more slowly, but still tuned out. A screen is not a door, and in the later study attention partly recovered between days, while my patient’s visits are three months apart. Carrying this to the clinic is an analogy, and it is mine.
What I propose is small. Keep the sentence. It is a courtesy, it is true, and on ordinary days it does no harm. But on the day the risk changes, interrupt it. Before the usual close, mark that today is different, in words this patient has not heard from you. At that door it might be:
“Today is different from our other visits. This medicine can make you bleed. If your stool turns black, or a nosebleed won’t stop, go to the emergency department. Don’t wait for Monday, and don’t wait for your next visit.”
Then the old sentence, as always. This is ordinary safety-netting with one addition. It marks the visit. She has already told him the symptoms once. Saying them again at the door, in new words, tells him that this goodbye is not the usual one. It is not a substitute for teach-back, which checks what he understood and has evidence this does not.
It is not free. It is thirty seconds standing at a door, in a schedule that was already behind when the day began. It breaks a rhythm the patient finds reassuring, and he may ask whether he should be worried, which is a fair question and another five minutes.
And it has a limit from the same studies. The changing warnings were tuned out too, only more slowly. Say the specific version at every visit and it becomes the next thing he nods at.
The chart records what we said. The patient goes home with what he heard. Only one of those will be with him on a Saturday.
A goodbye he has heard thirty times cannot carry a warning he has heard once.
Gurdev Singh is a physician-writer whose work centers on one idea: Most misunderstandings don’t occur because no one spoke, but because meaning did not transfer. Singh studies the gap between what is said and what is heard, especially in high-stakes settings such as illness, aging, and difficult decisions, and explores how trust is formed or lost within seconds, why reassurance sometimes increases distress, and how presence can succeed where explanation fails.
Singh writes literary nonfiction, research-informed essays, and fiction, and gives talks on empathy, communication, and human dignity. The aim is not to teach scripts, but to deepen awareness of what people carry away from an encounter, because that, more than information, shapes outcomes. In Singh’s words, “Where words end, meaning begins.” Singh’s writing appears on Substack at The Two Wells.
Your KevinMD author page is one of the top results for your name. Turn it into a premium profile, publish in 1 to 3 days, and get a guaranteed podcast slot.


