Daring to Emote In Public

She came in expecting a Whipple and a good chance at a cure, but the first peek of inside proved that not-to-be. She woke from the anesthesia, family by her side, and they all heard the news. Surprise metastases, cure is now off the table, the world has tilted on it's axis.

Then she started crying.

Within moments: tissues frantically shoved in her direction, social work consulted, spiritual care consulted, psychiatry considered, the word coping in the chart with a question mark after it. Palliative care consulted for "support."

And we all show up.

Someone paged a gaggle of folks to a room where a woman had just learned she was going to die of her cancer, and the clinical question was whether she was taking it correctly.

The curriculum nobody wrote down

The formal curriculum on emotion is fine. Deliver serious news clearly. Respond to what you see. Explore what matters. Support the decision. Most schools teach it reasonably well.

Then there's what people actually learn.

In 1998, Frederic Hafferty argued in Academic Medicine that reforming the syllabus was theater, because the syllabus was never where the learning happened. What shapes a physician is the hidden curriculum built into the structure of the place. Then he did the useful part and told us where to look — institutional policies, evaluation activities, resource-allocation decisions, and institutional slang.

Run those four over a crying patient and the curriculum comes into focus. Visible emotion is an adverse event, and someone is responsible.

Nobody says that out loud. Nobody has to.

The six words that teach it

A resident runs a family meeting. The family was upset at the end. The attending asks, kindly, in the tone people use when they think they're teaching:

"What could you have done differently?"

A whole curriculum can hide inside those six words. The question presupposes that upset is an outcome the resident produced, that a better resident would have produced less of it, and that the correct amount was lower than what occurred. The grammar does the teaching, which is why nobody notices it happening. Ask a trainee that often enough and they will learn, permanently, that a dry-eyed family is a well-run meeting.

They will then run dry-eyed meetings. Which is the problem.

The repair costs nothing. Change the question. Ask what the family understood at the end that they didn't understand at the start — an answerable, checkable question that aims the trainee at the transfer of information rather than the temperature of the room. A resident who can't answer it has work to do, and the work has nothing to do with tissues.

We drew the line on the wrong axis

Borrow the shape of the inverted-U curve — loosely, since in researching for this post, I discovered (much to my dismay) that my attribution to Yerkes and Dodson's 1908 mice is mostly a century of citation drift — and you get a usable picture. Somewhere on that arousal axis, medicine drew a line and called everything to the right of it pathology.

The line sits at composed. A patient who receives a terminal diagnosis and nods thoughtfully is coping. A patient who receives the same diagnosis and weeps is a consult. Every clinician reading this knows where that band ends, because we were all trained to feel it end.

Generalized anxiety disorder requires excessive worry, more days than not, for at least six months. Prolonged grief disorder, which the APA added to the DSM-5-TR in 2022, requires that the death have occurred at least twelve months ago for adults, plus functional impairment.

And adjustment disorder — the soft landing, the code we reach for when we want to do something with a person's grief — has two criteria that ought to end this conversation. Criterion A gives you three months from the stressor. Criterion B requires that the distress be out of proportion to the severity or intensity of the stressor. Criterion D excludes normal bereavement outright.

Sit with Criterion B. Her stressor is metastatic cancer. What magnitude of distress would be out of proportion to metastatic cancer?

Every threshold in the nosology turns on how long it lasts and whether it fits what happened. Intensity appears nowhere. Whether it made the team uncomfortable appears nowhere.

So say the whole sentence out loud before you page anyone. "I am consulting psychiatry because I believe this woman's response to learning she has liver metastases is out of proportion to having liver metastases." If you can't finish that sentence without flinching, you've learned something, and it isn't about her.

What we pay to make it stop

There is no code for the ninety seconds of silence and coming alongside her that would have made the consult unnecessary.

So it's time to turn that mirror back on ourselves. Palliative care spent two decades making the case that honest conversation doesn't hurt people, and we proved it the only way the system accepts proof — with instruments that don't measure "hurt" unless you're a physician who learned that anxiety=bad.

The Serious Illness Care Program trial is the cleanest example. Bernacki and colleagues reported honestly, including the part nobody quotes: the coprimary outcomes were null, in 64 patients. What gets cited is a secondary outcome. Moderate-to-severe anxiety, 10.2% versus 5.0% at 14 weeks, P = .05. That's a GAD-7 score — an instrument that asks about the last two weeks and scores against DSM criteria for a disorder. It cannot see a family meeting. It does not ask whether she cried.

Then it travels. "Good communication reduces anxiety" becomes "an anxious patient means the communication was bad" becomes an attending asking a resident what they could have done differently. A fourteen-week GAD-7 score turns into a real-time grading rubric for a forty-five-minute conversation. It just loses its denominator on the way.

The words we use when we're uncomfortable

Hafferty thought slang would be the most revealing site. The receipts are worse than he could have known.

We all know the vocabulary. Difficult family. Unrealistic. In denial. Not coping. Poor historian. None of these are clinical findings. All of them describe how a room made us feel, laundered into the passive voice and parked in a chart.

In 2022, Michael Sun and colleagues ran Hafferty's fourth lens across 40,113 history and physical notes from 18,459 patients at an urban academic medical center. They trained a model to find fifteen negative descriptors. Adjusted for sociodemographics and health characteristics, Black patients had 2.54 times the odds of having at least one in their chart.

Now look at the descriptor list. Agitated. Angry. Aggressive. Challenging. Exaggerate. Resistant. Noncompliant. Refused.

That's an emotion list. Sun set out to measure racial bias in documentation and ended up measuring what happens when medicine converts feeling into text.

The white patient in my vignette cried and got a palliative care consult, which is a soft outcome even when it's the wrong one. Change one variable. She raises her voice instead of weeping, or her son does, and the room reorganizes. Security becomes thinkable. Agitated enters the H&P. And unlike a consult, that word follows her — waiting at the next admission, shaping the first ninety seconds with a clinician who has never met her.

Same underlying event, a clinician who couldn't sit with emotion. The cost scales with the patient's position.

Closer to home: Major-Kincade and colleagues looked for biased language in the charts of PICU patients carrying a palliative care consult. Underpowered at 63 patients, they found no significant demographic effect. They did find the language, in our charts, and physicians and social workers used more of it than anyone else on the team. Those are the two consultants in my vignette. We are called to metabolize the family's emotion, and we are writing it down worst.

The distinction we trained out of ourselves

Bad communication causes distress. A resident who explains a Whipple to a woman who thinks she's there for a hernia has caused distress, and that distress is a finding about the resident.

Good communication reveals distress. The cancer was there before the meeting. The prognosis was there before the sentence. The grief was assembled and waiting before anyone said the word. All the conversation did was make it impossible to keep not knowing.

The two look identical from the doorway. Same tears, same tissues, same page to social work. Medicine, unable to tell them apart from outside the room, resolved the ambiguity toward its own comfort and decided all of it was error.

Here's the cost. Among patients on palliative chemotherapy who didn't understand it wouldn't cure them, Jane Weeks and colleagues found that the ones who rated their physician's communication most favorably were more likely to be wrong. I've written about the collusion mechanism behind that, so I won't relitigate it.

Read it through this lens and it says something specific: the calm room might be the misinformed room. If distress is your quality signal, the signal is wired backward, and the meetings you rate highest are the ones where the least got through. Which means the composed patient should make us question ourselves.

Some people receive terrible news with equanimity, and I admire them. But if we don't teach the work to distinguish equanimity from the possibility that she didn't hear you, we've built a curriculum that rewards us for not checking.

Consult anyway

Offer social work, chaplaincy, psychiatry, and specialist palliative care generously — and offer them because a person might want help carrying something heavy. Even if the problem lives in the trigger: a consult placed because the team got uncomfortable, documented as a question about whether the patient is responding correctly, we will show up. We might slow walk the consult for a few hours, though.

Where I might be wrong

In roughly increasing order of how much it would cost the argument.

The curve is borrowed goods. I've used the shape of a relationship whose provenance is a century of misattribution, in a piece about intellectual honesty. It earns its place as a picture rather than a warrant, and I've said so. A less charitable reader will call that having it both ways.

The duration criteria cut both ways. Acute grief intensity in the first year predicts prolonged grief disorder later. If early distress carries prognostic information about who will struggle, the team's instinct at minute twenty is premature rather than defensive — it's tracking something real. My argument needs the timescale distinction to hold cleanly, and the literature is messier than that.

I may be making a real problem harder to name. Sometimes the family is upset because the meeting was botched, and "what could you have done differently?" is exactly the right question. If this lands as permission to stop asking it, I'll have traded one bad reflex for another. The distinction between causing and revealing carries enormous load here, and it isn't always available in the moment.

And the SICP findings might be more than a rhetorical prop. If serious illness conversations really do reduce clinical anxiety at fourteen weeks, honest communication is therapeutic rather than merely tolerable — a stronger claim than mine, and one that partly undermines it. Someone should push back that those numbers are also just good news.

What this is really about

The surgical oncologist explained the findings clearly and did nothing wrong. The indictment belongs to a learning environment that taught all of us — down to the palliative care physician who took the page — to read a woman's tears as a problem with a solution rather than as evidence that our words landed.

She has a right to her anxiety. Rights don't come with target values. The correct amount of dread about your own death was never ours to set.

She was crying because she understood.

I am a palliative care physician, educator, and professional strategery expert. Known for turning rounds into rants and rants into teaching points. Rounds & Rants represents my views — not those of any organization or professional membership organization where I hold a role. I don't write on their behalf and they don't vet what I publish.