I need to say at the top that I almost didn't want to do this one.
You said that about the funeral episode too.
I say it about every episode in this series and then I do them anyway. Daniel's written in with another one for the death series. He says he's putting his superstitions aside, which I appreciate, because Daniel is famously superstitious about exactly this topic and he knows it.
He does know it. He's said so.
He has. And then he goes and sends us this. He's met palliative nurses over the years, and what struck him wasn't anything dramatic. It was the reverence. The humility. People who work in death every day talk about it like it's a thing they're still learning. And not all of them believe in an afterlife. That's the part he keeps coming back to. Some of these workers will tell you the dying see relatives, describe events they had no rational way to know. Some say it's biochemistry, chemicals flooding the brain at the end. Others will quietly tell you they believe it one hundred percent and nothing will move them off that.
And he made an observation about where those conversations happen.
Right. Not at the deathbed, with the family in the room. In the break room. In the parking lot. In some other context entirely, these people are candid. So he's asking three things. Is death still taboo even in the places where it should be least taboo, like a hospice. How do the people who see it every day actually come to think about the dying process. And he's staking his own ground here, he's not hiding it. He says he's on the side of the paranormal. He thinks dying is a process and that the dying can reach out in ways that defy physics. He wants to know how religious traditions, Judaism included, relate to the metaphysical dimension of all this.
He's not asking us to settle whether ghosts are real.
He is not. He's asking what the people closest to it think. And here's the thing that sold me on the episode. That instinct, that the people closest to death have the most interesting things to say about it, it turns out to be backed by an unusually deep research literature. There are surveys. There are chart audits. There are qualitative studies where nurses say things on the record they'd never say at a bedside.
There's a specific number I want to get to, but we should define the territory first.
Define it.
The umbrella term is deathbed phenomena, or end-of-life experiences. It covers a lot more than people assume. Deathbed visions, which is the classic one, seeing a deceased relative. End-of-life dreams and visions, which is the broader category. Terminal lucidity, which is the one that gets people. A patient with advanced dementia who hasn't spoken in weeks suddenly sits up and has a clear conversation. Then there's patients holding on for a loved one to arrive. The wish to go home, which is a phrase that comes up constantly and doesn't usually mean the building they live in. The desire to mend a family rift before the end. And deathbed coincidences, where a distant loved one, sometimes on another continent, senses the death at the exact moment it happens.
That last one is the one that would unnerve me.
It's also the least reported. Forty-three percent in the survey I'm about to cite. Which is still a lot.
So how common is all of this?
This is the number that reframes the whole topic. A UK-wide survey of a hundred and four Marie Curie palliative care staff, run out of the University of Edinburgh and presented in twenty twenty-two, found that ninety-seven percent of them, a hundred and one of a hundred and four, reported at least one type of deathbed phenomenon in the previous five years. The average respondent had encountered seven point seven of them.
In five years.
This isn't a rare event that a few sensitive people notice. This is the texture of the job.
Seven point seven. I want to sit with that for a second.
The breakdown is where it gets interesting. Patient holding on for a loved one to arrive, eighty-six percent. Vision of a deceased relative or friend, seventy-five percent. Dreams of deceased relatives, seventy-two and a half. Wish to go home, seventy percent. Desire to mend family rifts, sixty-six. Sudden lucidity in a dementia or unconscious patient to say goodbye, fifty-six. Changes in pet or animal behavior, fifty-five. Deathbed coincidence, forty-four.
The animal one. That's the one nobody expects on the list.
It shows up in the qualitative studies too. And here's the part Daniel was circling without quite naming it. In that same survey, seventy-six percent of the staff agreed these experiences are of a spiritual nature. And there was no significant difference between the religious staff and the non-religious staff.
Say that again.
Same number. Religious and non-religious staff, same rate of calling it spiritual.
That's the finding that breaks the easy explanation. If this were faith talking, you'd expect a gap. You'd expect the believers to see spirits and the atheists to see hypoxia. Instead you've got nurses who don't believe in God watching a patient have a conversation with their dead mother and going, I don't know what that was, but it wasn't nothing.
Seventy-one percent agreed the experiences comfort the patient and family. Sixty-eight percent wanted formal training on how to handle them. That last number is a tell. You don't ask for training on something you've already dismissed.
So what's the actual landscape of belief among these workers? Is it a clean split, spiritual versus biochemical?
It's messier than a split. There's a Nordic qualitative study, eighteen palliative care professionals, and fifteen of the eighteen had direct experience of patients' end-of-life dreams and visions. The content was dominated by deceased loved ones. Most of the patients perceived them as real and described them with total clarity. And the professionals split three ways. Some of them believed the visions were real and a normal part of dying. A few found them frightening. But most of them landed in the same awkward place. They found the experiences hard to grasp, they tried to explain them medically, and then in the same breath they'd concede that most of these patients were normal and of sound mind.
That's not a position. That's a person standing in two places.
It's the honest position. I've been in that spot clinically. You have a framework that explains almost everything, and then something walks up to the edge of it and you don't throw the framework out, but you don't lie about what you saw either. You just... hold both.
What does the biochemical side actually claim? Because I want to state it fairly before we push on it.
The most common clinical framing is delirium. There's a Swiss paper from twenty twenty-three that says it plainly. In the clinical context, end-of-life dreams and visions are usually neglected, and interpreted and treated as pathological changes in the brain that result in, and from, delirium. That's the default. The experience is a symptom.
And the more specific hypothesis?
Terminal lucidity has a proposed mechanism. A twenty twenty-one paper in Alzheimer's and Dementia offers what it calls a new lucidity hypothesis. Fluctuation of neuromodulators running from the brain stem up to the medial prefrontal cortex and the hippocampus, plus corticotropin-releasing peptides increasing mental clarity. So the sudden return of a dementia patient is a chemical event in a specific circuit.
That's a real hypothesis. It's not a hand wave.
It's a real hypothesis and it's testable in principle, which is more than you can say for a lot of what gets written about this.
Here's where I want to push, though. Does that explanation cover the cases where the patient describes something that later turns out to be true? The relative who died last week and nobody told them. The thing happening in another room.
That's the part the delirium framing doesn't touch. Delirium explains the experience. It doesn't explain the content matching a fact the patient had no access to. And the literature is honest about this. It's not that the studies debunk it. It's that they mostly don't engage with it.
Because you can't put it in a chart.
You can't put it in a chart, and you can't run it as a controlled trial. There's a paper from twenty twenty-four, a scoping review in BMJ Supportive and Palliative Care, that pulled together a hundred and fifteen reports on end-of-life experiences. Seventy percent of them published after twenty ten. Fifty-five percent from the United States. And the most common term across all of them is deathbed visions. The field is growing fast and it's still mostly descriptive. Prevalence across studies ranges from twenty-eight to eighty-seven percent. For end-of-life dreams and visions specifically, thirty to eighty percent, typically showing up weeks to hours before death.
A twenty-eight to eighty-seven range is not a measurement. That's a fog.
It's a fog, and the reason for the fog is definitional. Every study draws the boundary differently. Does a dream count? Does a dream the patient remembers count? Does it count if the patient can't report it and the nurse infers it? Once you fix the definition, the number tightens. But nobody's fixed the definition.
There's a study I want on the table, because it's the one that made me sit up. The Fenwick work.
Twenty ten, Archives of Gerontology and Geriatrics. Thirty-eight nurses, doctors, and carers across two hospices and a nursing home. Five-year retrospective plus a one-year prospective arm. The finding was that these experiences are not uncommon. Visions, coincidences, the desire to reconcile. But the detail that stays with you is the peripheral stuff. Clocks stopping synchronistically at the moment of death. Shapes leaving the body. Light surrounding the body. Strange animal behavior.
Clocks stopping.
And the interviewees were clear on one point. They said these experiences were different from drug-induced hallucinations. They occurred in clear consciousness. That distinction matters, because the lazy dismissal is always, well, they were on morphine. These were people who knew what morphine looked like. They'd seen it a hundred times. And they said this was not that.
What about the death doulas? They're the newest profession in this space and they have no clinical obligation to the delirium framing.
Twenty-six death doulas in a twenty twenty-five study. The finding was that most or almost all of them endorsed spiritual or transpersonal explanations for end-of-life phenomena over scientific or medical ones. And they were comfortable discussing it with clients and families. That comfort is the thing. They're not carrying the same institutional pressure to translate everything into a diagnosis.
So that's what the workers see. The next question is what they do with it. And that's where it gets uncomfortable.
The documentation gap.
Because there's a study that puts the two numbers side by side and the gap is almost comical. Seventy-five hospice nurses, reporting three hundred and sixty-three deathbed communications, averaging four point eight per nurse per thirty days. So roughly one a week, per nurse.
And then the chart audit. Sixty charts reviewed. Five of them recorded the communication. Eight point three percent.
So ninety-two percent of these events happen and vanish.
They happen, they're witnessed, they're significant enough that the nurse remembers them well enough to report them in a survey, and they don't make it into the record. The paper's own conclusion is that these are underreported in patient records and under described in textbooks.
That's the spine of this whole episode, and I want to say why. It's not that the system denies these experiences. It's that the system doesn't have a box for them. And a thing with no box doesn't get written down. And a thing that doesn't get written down doesn't get taught. And a thing that doesn't get taught is a thing every new nurse has to rediscover alone at three in the morning.
Which is exactly the mechanism Larson described forty years ago.
Daniel asked whether death is still taboo even in hospices. Larson's study is the sharpest answer anyone has ever given to that question.
Nineteen eighty-five. Two hundred hospice workers, anonymous confessions. And I want to read a few of these because paraphrasing them loses the punch. "I feel that I'm a fake." "Sometimes I hate the people I work with." "I am terribly afraid of my own death." "I hope patients won't die on my shift." And this one. "Sometimes I pray that God takes this patient because he's suffering so much. I hope he dies soon. I feel guilty."
That last one is the whole job in five sentences.
And Larson's explanation for why none of them said it out loud is the part that generalizes way beyond hospice. He called it the fallacy of uniqueness. The individual's false assumption that he or she is the only one responding in this way.
So every hospice worker in that study thought they were the only one who felt like a fraud, the only one afraid of their own death, the only one who sometimes wanted a patient to die. And because each of them thought they were alone, none of them said anything, which confirmed to everyone else that they were alone.
It's a self-sealing silence.
That's the taboo. It's not that hospice workers can't talk about death. They talk about death all day. It's that they can't talk about their own reaction to it. The taboo isn't on mortality. It's on the worker.
And it extends to the phenomena. There's a finding in the Nordic study that patients could be reluctant to report their visions because of a fear of being considered crazy. So the silence runs in both directions. The patient doesn't want to sound mad. The nurse doesn't want to sound superstitious. And the family is sitting right there, so nobody says anything, and the experience goes into the chart as agitation or not at all.
Do we know how often staff withhold the paranormal reading from families specifically?
Not directly. Nobody's measured that exact gap. What we have is indirect and it's strong. Larson's secrets. The fear-of-being-considered-crazy finding. And the repeated calls for training. In one systematic review, normalization came up as a theme in half the studies, and training in about forty-two percent. When half the literature is asking for something to be normalized, that's a signal it currently isn't.
Let's talk about what these experiences do for the dying, because that's the argument that moves it out of the metaphysical and into the clinical. If it were purely a curiosity, fine. But there's data.
There's good data. In the Lawrence and Repede work, eighty-nine percent of nurses said patients who had a deathbed communication had a peaceful, calm death. Compare that to forty point five percent of patients who didn't have one.
Eighty-nine against forty.
That's a massive difference. And it's not a controlled trial, so we can't say the communication caused the peace. It could be that people who are at peace are more likely to have the experience. But either way, the presence of the experience is a strong signal about how the death is going to go.
It's a prognostic marker.
It's a prognostic marker that nobody writes down. There's more. A study of fifty-nine hospice inpatients found most reported at least one dream or vision, nearly all felt real, and dreams of the deceased were significantly more comforting than dreams of the living. That's a clean result. The dead show up in the dream and it helps.
The dead are better company than the living. In the dream.
In the dream, specifically. And then there's the posttraumatic growth work. Hospice patients with end-of-life dreams and visions scored significantly higher on personal strength, on spiritual change, and on overall posttraumatic growth. Those are measured, they're significant.
And the families?
Two hundred and twenty-eight bereaved family caregivers. Comfort from dreams correlated with accepting the reality of loss, working through grief, and continuing bonds. All three correlations in the half-a-point range, which for this kind of survey work is strong.
So the experience helps the patient die calmer and helps the family grieve better. And it goes in the chart eight percent of the time.
Eight percent of the time.
And this is where Judaism offers something different from the Catholic model most people have in mind. Daniel asked about this specifically.
It does, and the contrast is sharp. Judaism has a formal deathbed confessional. It's called the Viddui. It's recited when death is imminent, and it can be said by the dying person, by family, or by a rabbi, in Hebrew or in English. The Shulchan Arukh is explicit that it should never be imposed. The dying person decides.
Read the line about the timing.
"If you feel death approaching, recite the Viddui. Many have said the Viddui and not died, and many have not said the Viddui and have died. If you are unable to recite it aloud, say it in your heart."
It's telling you not to treat this as a mechanism. You say it and you might live. You skip it and you might die. It's not a lever you pull to get an outcome.
That's the explicit contrast with the Catholic last rites. The Viddui does not insure the soul's place in the world to come. It doesn't tempt fate. It's a confession, not a transaction. And the Shema, "Hear O Israel," is the last thing a Jew is supposed to say before death. The final word of the Shema is Echad. One. And that word is read as the soul's ultimate reconciliation.
There's a framing in the Jewish sources that I find more honest than most of what I've read on this. Judaism teaches an afterlife. It doesn't use that teaching to minimize the reality of loss. No euphemisms. You don't say passed away. You say died.
Which is a strange kind of comfort. It's the comfort of not being lied to.
Now the clinical side of that. How well do clinicians actually handle Jewish end-of-life practice?
Poorly, and there's data. A twenty nineteen study in the Journal of Palliative Medicine found clinicians scored a median of six out of ten on Jewish end-of-life knowledge before a brief educational intervention. After one hour of education, they went to nine out of ten.
One hour.
One hour moved them three points. Which tells you the gap wasn't depth of understanding. It was exposure. Nobody had ever told them. And there's a separate paper on the ethical challenges of caring for Orthodox Jewish patients at the end of life, which is its own set of problems around autonomy, around who makes decisions, around what's permitted at the moment of death.
So we've got a system that doesn't chart the experiences, doesn't train on the experiences, and also doesn't know the religious frameworks the patients are bringing to the bed. That's three gaps stacked on top of each other.
And the patients are the ones standing under all three.
Before we get to Hilbert, I want to name the thing that's been sitting under this whole conversation. Daniel staked his ground. He said he's on the side of the paranormal. And what the research actually shows is not proof of anything. But it shows that the people with the most exposure to death are the least dismissive of it. That's not evidence for the paranormal. It's evidence about who's paying attention.
Hilbert: Nobody surveys the night shift. I worked nights at a nursing home. Not a nurse. Orderly. Eleven to seven. Moved residents, changed sheets, cleaned up. On my floor, for most of the shift, I was the only one awake. That's the point I've been sitting on.
Go on.
Hilbert: The nurses get interviewed. The doulas get interviewed. The day staff get the training sessions, they get the pamphlets. The orderly who was actually in the room at three in the morning when a resident sat up and had a conversation with somebody who wasn't there, nobody asks him anything. I've been waiting years to correct the record on this.
So correct it.
Hilbert: There was a woman. Advanced dementia. Hadn't spoken in weeks. No verbal communication at all. I'm in the room, doing what I do, and she sits up. Straight up. Looks at the corner of the room, over by the window, and says, clear as anything, "I'm not ready yet." Then she lay back down. Didn't speak again. Died four days later.
Did you tell anyone?
Hilbert: No. I didn't want to be the guy who made it weird. You tell a story like that on a night shift and suddenly you're the guy. So I didn't. I told my brother-in-law, years later, and he said the same thing happened to his aunt, and then he said something I've thought about since, which is that you can't trust him on this stuff because he exaggerates everything.
You're relaying it anyway.
Hilbert: I'm relaying it with the caveat. That's the honest way to do it.
The detail I keep coming back to is the four days. She came back long enough to say one sentence and then she was done with the conversation.
Hilbert: That's how it went. And then I quit, because the coffee machine on nights was always broken. Every night. Somebody would put a note on it. Nothing ever happened. So I left.
The coffee machine.
Hilbert: You can't work a shift like that without coffee. Anyway. There's a phone call I'm expecting and I'm not taking it in here.
The night shift thing is real, though. It's not just his grievance. The Fenwick study interviewed carers across two hospices and a nursing home. Somebody was on those floors at night. And the Marie Curie survey got seven point seven encounters per staff member on average, which means somebody was there for all of them. It's just that when the study gets written, the person who fills out the questionnaire is the one with the title.
So the record is built by the day shift and the events happen on the night shift. That's a data problem. It's also just true.
One thing I want to flag before we close, because it surprised me. We went looking for a dedicated peer-reviewed paper on the DMT or ketamine or hypoxia explanation for deathbed visions. The kind of thing you see cited constantly in popular writing.
And?
It's not there. The search came back empty. The closest documented skeptical framing is the delirium interpretation and the neuromodulator lucidity hypothesis. Both real, both in the literature. But the specific chemical story that gets repeated everywhere, the one that sounds so settled, doesn't have a paper behind it that we could find.
So the biochemical explanation is real but less settled than it's presented.
Considerably less settled.
Which leaves us with the actual shape of this. Seventy-six percent of the people who work with the dying privately call these experiences spiritual. The clinical default codes them as delirium. And eight percent of them make it into the chart. That gap isn't a research problem. You can't fix it with a bigger study. It's a cultural one. It's about what a profession is willing to write down.
Daniel put his flag down on the paranormal side. The research doesn't settle that. It doesn't need to. What it does show is that the people closest to death are not the ones dismissing it. They're the ones asking for training.
And that's a strange kind of answer to a question about ghosts. The most credulous people in the room are the ones who've been in the room the most.
Thanks as always to Hilbert Flumingtop, who produces this show and who has now put the night shift on the record.
If you want to send us something, email us at show at my weird prompts dot com. Reviews help too, wherever you listen.
This has been My Weird Prompts.
We'll be back soon.
