She almost missed it.
That was the part that stayed with her afterward; not the finding, but the almost-missing. The way it had been sitting there in the peripheral vision of her research for months, a data point she had filed and not returned to because it wasn't about him, wasn't about the Fighter or Elena or any of the architecture she had been building. Just a number that didn't match another number in a document she had pulled for a different reason entirely.
She had almost closed the tab.
She hadn't.
It started with Marcus's folded paper.
Not the medication itself: the brand. She had memorized it in four seconds in an elevator and it had sat in the back of her mind the way all information sat, filed, patient, waiting for something to connect to. Sertraline. Northside Wellness Pharmacy. 100mg. She had used it as a tool, as a lever, as the mechanism of a plan she had been executing with the specific focus of someone who had stopped asking whether the plan was right and had only kept asking whether it was working.
But she was still, underneath everything, a journalist.
And journalists notice when numbers don't match.
The document was a quarterly report from Helix Pharmaceutical, the company that manufactured the generic version of the medication she had taken from the Fighter's cabinet. She had pulled it three weeks ago looking for something else entirely — a contact, a source, a name she needed for a different thread — and had skimmed it and filed it and moved on.
She went back to it at 2 AM on a Tuesday because she couldn't sleep and her plan was in its final phase and there was nothing to do but wait and she had never been good at waiting and her hands needed somewhere to go.
She opened the document.
Read it properly this time.
The number that didn't match was small. Easy to miss. A discrepancy between the efficacy rate cited in Helix's public-facing materials — their website, their prescriber information sheets, the literature that went to pharmacies and doctors and the patients who asked questions — and the efficacy rate buried in table 7 of their internal clinical trial summary, which was technically a public document because it had been submitted to the FDA as part of a regulatory filing, but which was formatted in a way that assumed nobody would read it. Seventeen-point-three percent versus twenty-four-point-one percent. The difference between what they told the world the drug did and what their own data said it did.
She sat with this for a moment.
Then she started pulling threads.
By 4 AM she had eleven documents open.
By 6 AM she had twenty-three and a legal pad with handwriting that covered both sides, her handwriting getting faster and smaller as the picture got larger, the specific compression of someone who was writing faster than they could think because the thinking was going faster than it usually did.
By 8 AM she understood what she was looking at.
This was structural.
Helix had not falsified their clinical trial data. That was the elegant, terrible part of it. They had designed the trials to produce the numbers they needed — not by manipulating results but by manipulating methodology. Patient selection criteria written to exclude the populations where the drug underperformed. Trial duration set to the window where efficacy peaked before the plateau. Comparison baselines chosen from older studies with different measurement standards.
Everything technically defensible. Everything technically true.
And underneath it, in the gap between the technically true and the actually true: seventeen-point-three percent of patients for whom the drug worked as advertised. Twenty-four-point-one percent in the internal data. The difference represented, when she did the math, approximately 340,000 people in the United States currently taking a medication that was performing significantly below what they had been told to expect.
340,000 people making decisions about their mental health based on numbers that had been architecturally arranged to look better than they were.
She put her pen down.
Looked at the legal pad.
Picked the pen back up.
The prescriber data came from a different source — a healthcare analytics firm whose data she had accessed through a contact she had cultivated three years ago for a story about insurance billing practices. She had not used the contact since. She used it now.
What she was looking for was whether the efficacy gap showed up in the real world the way it showed up in the internal data. Whether there was a signal in actual prescription outcomes that matched what table 7 was quietly saying.
There was.
It took her four hours to find it because it was distributed across seven different datasets that had never been aggregated, that existed in separate silos maintained by separate institutions that did not talk to each other. She aggregated them. She built the picture.
The signal was there. Patients on Helix's generic sertraline were being switched to alternative medications at a rate significantly higher than patients on comparable formulations from other manufacturers. The switching rate had been climbing for three years. Nobody had connected it to the efficacy data because the efficacy data was in a regulatory filing that assumed nobody would read it.
Nobody had read it.
She had read it.
She thought about 340,000 people.
She thought about them specifically, not as a number but as the thing a number was made of. People who had gone to their doctors and described something difficult and been given a prescription and filled it at a pharmacy and taken it every morning next to a glass of water and waited, the specific waiting of someone who has been told that the thing they are doing will help, the waiting that requires a particular kind of faith in the system that is supposed to protect you.
She thought about what it meant to wait like that and have the thing not work as well as you'd been told.
She thought about the Fighter's cabinet above the glasses.
She sat with that for a long time.
The thing she noticed, sitting at her desk at noon with the legal pad and the twenty-three documents and the city going on outside her window, was the specific quality of how she felt.
It was just true.
340,000 people. A methodology designed to make numbers look better than they were. A regulatory filing nobody read. A signal in the data that had been climbing for three years.
She had found this.
Not because she was hunting the Fighter. Not because she was building a case against Elena. Not because of the plan or the list or the name she had written Tomorrow next to three times. She had found it because she was a journalist and she had been looking at documents at 2 AM and she had almost closed the tab and she hadn't.
She thought about the prescription she had taken from his cabinet.
She thought about it for a long time.
She picked up her phone.
Not to call Marcus. Not to call the Cut Ties group. Not to call anyone who was part of the architecture of what she had been building.
She scrolled to a number she had not called in eight months. A health policy reporter at the Tribune who had covered pharmaceutical regulation for twelve years, who was meticulous and careful and had the kind of sourcing network that took decades to build, who she had always respected even when she hadn't liked him and who would know, immediately, what table 7 meant and who else needed to see it.
She held the phone.
She thought about the Fighter's cabinet.
She thought about seventeen-point-three percent.
She thought about 340,000 people doing the specific waiting of someone who has been told the thing they are doing will help.
She thought about Viktor in his room. About Amore's careful face. About Mika's hands in her lap and the pen clicking and the word tomorrow written three times in her own handwriting.
She thought about all of it.
Then she called David Chen at the Tribune and when he answered she said: I have something. It's not what I've been working on. It's bigger. I need you to look at a document.
He was quiet for a moment.
Then he said: Send it.
She sent it.
She sat at her desk afterward and did not move for a while.
Outside, Chicago was doing what it did. The city going on without asking anyone's permission. She had lived in this city for thirty years and it had never once paused for her and she had never once expected it to and she did not expect it to now.
She looked at the legal pad. At the twenty-three documents. At the number she had written in the margin in the specific compression of her fast handwriting:
340,000
And then, underneath it, the thing she hadn't said out loud in a very long time, the thing that existed before the plan and the list and the name written Tomorrow and the bottles and the empty pharmacy shelves:
Not the Fighter. Not Elena. Not the mythology or the dismantling of it or the satisfaction of being the one who pulled the curtain back on something comfortable.
This. A number in a table that didn't match another number. A methodology designed to look better than it was. 340,000 people waiting for something to work.
She had been a journalist for thirty years.
She had been other things too.
But she had been a journalist first.
She looked at her phone.
Looked at the plan. At the name. At the last Tomorrow.
Then she looked at the legal pad.
The legal pad did not ask her anything. It just held what she had found, patient and specific, waiting for her to decide what to do with it.
Outside the penthouse, somewhere in the city, the Fighter was missing.
The prescription was still in her bag.
The gallery where Elena was showing was three weeks away.
Viktor's letter was somewhere in the world, moving.
She looked at the legal pad for a long time.
Then she opened a new document and started writing the story that was true in a way that had nothing to do with her.
They always did, when the thing was real.
340,000 people waiting. One journalist who almost closed the tab. The words came fast. They always did, when the thing was real.