I had reviewed the slides four times before I left the room at noon.
Not because I didn't know the material. I had known the material for two years — I had built the material, brick by brick, starting with a dataset that took eight months to compile and a statistical model that I had rebuilt from the ground up after the first rejection. I knew the serum level cutoffs in my sleep. I knew the confidence intervals without looking. The slides were not for my benefit.
The slides were for the room. The room needed entry points: clarity at the beginning, the data compressed into graphics that told the story before the labels did, the conclusion arriving like the last piece of a structure that had been visibly under construction the whole time.
Priya had looked at them on Monday evening and said: "The slide about inflammatory markers needs a transition sentence." I had added one. I had written it, hated it, rewritten it, and then looked at it for twenty minutes before deciding it was correct.
The East Hall was filling when I arrived at one-thirty.
I put my things on the presenter's table, clipped my clicker to my hand, and looked out at the seats. Rows of white chairs, the kind that are slightly too upright, arranged in a gentle arc facing the screen. The light from the high windows came in at an angle that hit the first two rows directly.
I did not immediately look at the first row.
I was arranging my notes. I was checking the slide connector. I was doing several entirely legitimate tasks that required my attention.
I looked up.
He was in the front row, center seat. Notepad balanced on one knee, pen uncapped, entirely still. He had not brought a laptop, which meant he was planning to write by hand, which meant he was planning to write quickly and specifically, the way you write when you are taking notes on things rather than things for their record.
He was not looking at me. He was reading something in the program.
I looked back at my notes.
The session chair introduced me at two-oh-three. She read my credentials in the brisk, respectful way of conference chairs who have learned to compress a CV without losing the weight of it: post-doc, Harvard; prior publication record; the paper under discussion today, revised and resubmitted to Circulation after — brief pause — the original peer review process.
Ninety people in the room. I had counted.
I hit the first slide.
The first five minutes are the ones I had practiced until they were involuntary. The framing — why cardiac biomarker prediction in women under fifty is a different problem than the literature currently treats it — had to land before I got to data, because if I led with the data, a third of the room would start forming objections before they understood the premise. You have to build the house before you let anyone inside.
I heard myself speaking and I was in it — the way you are in it when the preparation has stacked up behind you like scaffolding and what you're saying is not performance but structure. The cohort comparison landed where it should. The serum level data opened up the way it should. The room was quiet in the attentive way.
I did not look at the front row.
I was aware of the front row. I had peripheral vision and a complete understanding of where every person in my immediate sightline was positioned. I was not looking at him. I was looking at the room, the way you look at a room when you are presenting to it — tracking faces, reading engagement, watching for the moment when someone's expression shifts from listening to disagreeing so you can address the objection before it becomes a question.
The inflammatory marker slide: transition sentence. It landed clean.
The revised methodology section — the section that had absorbed fourteen months and a complete reconceptualization of my stratification model after his notes had forced me to acknowledge the weakness in the original — I presented it with the particular confidence of someone who knows exactly how solid the ground is because they laid every stone themselves.
I did not look at him.
Then I looked at him.
He was writing.
His pen moved in short, specific strokes across the notepad. Not continuous notes — the kind of discrete notation that means this point, this number, this claim. He was not watching me. He was watching the slides and writing down things he intended to do something with later.
I went back to the last section.
Q&A opened with three minutes left in the session.
The first question was from a researcher in the third row — a clarification about the age stratification methodology that I had been expecting; I answered it in under a minute. The second was from a woman near the back who wanted to know about the exclusion criteria. Standard. I gave her the answer and the slide number for reference.
The third question came from the left side of the room, a conference veteran whose name I recognized from two papers I had cited in my footnotes. A good question — pushing on the sample size in the subgroup analysis. I answered it with the confidence of someone who had pre-loaded an answer to this specific challenge because she had read all the relevant literature before she submitted.
Then he raised his hand.
The room shifted. Not dramatically — a small collective adjustment, the kind that happens when attention finds a focal point. I saw two people near the front glance from him to me.
He waited until the chair acknowledged him.
"Dr. Morrow, Brigham and Women's."
He paused, briefly, to let the room arrange itself around what was about to happen.
"Dr. Callahan, the revised biomarker threshold you're proposing — the cutoff at the ninety-second percentile, rather than the standard ninety-fifth — that's the load-bearing beam of your entire clinical recommendation. If the threshold shifts, the clinical downstream shifts with it. My question is this."
His pen was down on the notepad. He was looking at me directly.
"Do you believe this data is sufficient to recommend a protocol change? Not 'is it publishable' — I'm asking whether you, with the full dataset in front of you, believe this is enough to move medicine."
The room was very quiet.
It was the question I had been afraid of. Not because I didn't have an answer — I did; I had had it for eight months. But because it was the question underneath all the other questions, the one that the methodology, the cohort design, the statistical model, every layer of the paper had been building toward or away from. It was the question that asked, without asking, whether I was certain. Not professionally certain. Not the evidence is within acceptable parameters certain. Whether I, Nora Callahan, standing in front of ninety people with two years of my life on a screen behind me, was certain.
I looked at him.
He was waiting, with the stillness of someone who was not performing a challenge but asking a question he genuinely wanted answered.
I opened my mouth.