Day three.
Dr. Rowe took the stand at 9:07 AM and sat with the specific upright composure of someone who had practiced being present in this chair. Not stiff — composed. The difference was in the hands: stiff people's hands gripped things. Dr. Rowe's hands were flat on the witness railing, open, still.
"Dr. Rowe," I said. "Can you tell the jury about the first time you noticed something unusual in the Corvalox trial data?"
She could. We'd rehearsed this sequence seven times, and the seventh time it had finally stopped sounding like a sequence and started sounding like memory.
"I was conducting routine data quality review for the Phase III trial data package," she said. "This was standard practice — I reviewed the data at several points during the trial to verify that the collection protocols were being followed correctly. During the April 2011 review, I noticed that the distribution of cardiac adverse events had changed significantly from my February review."
"What specifically had changed?"
"The number of events classified as severity level 2 — moderate adverse cardiac reactions — had decreased by approximately 78% between February and April while the trial population and dosing schedule remained constant." She held the jury's attention the way she always held attention: not by performing engagement, but by clearly conveying that what she was saying was important and true. "A change of that magnitude in that time window was statistically unexpected. Severity level 2 cardiac events don't disappear from a drug trial population without a clinical explanation."
"What did you do?"
"I raised it with my supervisor, Dr. Hendricks. I documented my concern in the quality review log." She paused. "He told me the change was within acceptable parameters and that the review committee had assessed the data."
"Did you have any documentation of that review?"
"I requested it. I was told it existed but was not provided access."
I waited one beat before the next question. The jury was ahead of me — they'd already heard enough to understand that the documentation probably didn't exist. The beat confirmed it without me saying so.
"What happened after you raised your concern?"
"I was reassigned from the Corvalox project to a different research track. The reassignment was framed as a career development opportunity." She looked at the jury. "It was not. I had been on the Corvalox project for three years at that point and had received excellent performance evaluations for each of those years. The first negative evaluation I received arrived two weeks after I escalated the data concern."
I moved through the escalation chronology methodically — the compliance officer meeting, the second documentation request, the second dismissal. The jury was tracking the pattern: concern raised, concern dismissed, consequence for raising it.
By the end of the first day of her testimony, the jury had the full institutional picture: a scientist who had followed every proper channel for six months, documented everything, and been systematically excluded from the project where she'd raised the concern.
Not a disgruntled employee. An employee who was disgruntled for documented reason.
Day four began with Exhibit 47.
I'd been building to this since the opening statement. Every preceding witness, every document, every testimony had been accumulating toward this single piece of evidence.
"Dr. Rowe, I'm showing you Plaintiff's Exhibit 47. Can you tell the jury what kind of document this is?"
"This is a version history log for Meridian's clinical data analysis software. The software is called CDAT — Clinical Data Analysis Tool. Every change made to CDAT's parameters is automatically logged with a timestamp and the credentials of the user who made the change."
"What is a version history log used for in pharmaceutical research?"
"Audit purposes, primarily. It creates a permanent record of how the software's analytical parameters changed over time, which allows research teams and regulatory reviewers to verify that data was processed consistently throughout a trial."
"Is it common for pharmaceutical companies to maintain such logs?"
"It's standard. FDA guidance requires it."
I moved to the exhibit display. Plaintiff's Exhibit 47 appeared on the jury screens — a dense technical document that I'd had reformatted for readability, with the relevant entry highlighted in a box.
"Dr. Rowe, can you direct the jury's attention to the highlighted entry on page seven."
She did. Her voice was completely level.
"This entry is dated March 14th, 2011. It indicates that the adverse event classification threshold for CDAT was modified. The entry shows that severity level 2 cardiac events — the category I had been concerned about in my April quality review — had their classification threshold changed. Any cardiac event rated at severity level 2 or below would now be automatically reclassified as 'unrelated to study drug.'"
"What was the threshold before the modification?"
"Level 1. Only the most minor cardiac incidents — events that would occur in any patient population regardless of medication — were classified as unrelated."
"So the modification expanded the 'unrelated' category to include a whole additional tier of cardiac events."
"Yes. Events that previously would have been reported to the FDA as drug-related adverse reactions were now automatically excluded from the drug's safety profile."
I let that sentence sit in the air.
"What date was this modification made?"
"March 14th, 2011."
"And when was the FDA application for Corvalox submitted?"
"March 20th, 2011. Six days later."
I set down my notes.
"Thank you, Dr. Rowe. No further questions."
I sat down.
The courtroom was not silent — it had the specific quality of a room where people are very carefully not reacting while reacting internally. Two jurors wrote things on their notepads. The retired statistician had stopped writing and was simply looking at the exhibit screen with an expression that said she didn't need further explanation.
Eleanor Marsh — cross-examination
She took her time getting to the podium.
The delay was tactical: let the room settle from the exhibit's impact, signal that she was not alarmed by what she was about to address.
"Dr. Rowe, you've testified about your scientific credentials at some length. You hold a doctorate in pharmaceutical chemistry, correct?"
"Yes."
"Not biostatistics."
"Correct."
"The analysis of clinical trial statistical methodology — specifically the determination of appropriate adverse event classification thresholds — falls within the field of biostatistics, not pharmaceutical chemistry, isn't that right?"
"The design of statistical thresholds requires biostatistical expertise, yes. The evaluation of whether a threshold change serves a legitimate scientific purpose can be performed by any scientist with sufficient familiarity with the data."
"Can it?" Marsh's tone was genuinely curious rather than aggressive. "Or are you saying that because you believe your familiarity with this particular data is sufficient?"
"I'm saying that a threshold change affecting 4.7% of a trial population, implemented six days before an FDA filing, does not require a doctorate in biostatistics to evaluate. It requires reading the version history and the submitted FDA data and comparing the two numbers."
"Or it requires the professional expertise to understand why a threshold change might be scientifically justified."
"If there had been a scientific justification, I would expect it to appear somewhere in the documentation."
"And you didn't find one."
"I looked for one. I didn't find one."
Marsh pivoted. "Dr. Rowe, your promotion to Senior Research Director at Meridian was denied in February 2011. Do you recall that?"
"Yes."
"And your concerns about the trial data began to develop in April 2011."
"My documented quality review concern was April 2011. The data change I was concerned about occurred between February and April."
"So between a denied promotion and your external whistleblower action, you also found a data problem that, if substantiated, would create significant legal exposure for the company that denied your promotion."
"I also found a data problem because I was doing my job. Quality reviews were a defined part of my role."
"Is that your testimony?"
"That is the fact."
[ Argument Crusher: Marsh cross assessment. Credential attack: partially effective (two jurors showed uncertainty). Motive attack: landed weakly — Dr. Rowe's redirect of the timeline was clean. Redirect needed: Chen's data corruption distinction will address credential limitations. Dr. Rowe's composure under sustained attack: exceptional. ]
Marsh pressed the motive angle for another twenty minutes without breaking Dr. Rowe's composure. She moved through the credential challenges, the alternative interpretation framework, the characterization of the software modification as routine maintenance.
Dr. Rowe answered each question with the calm precision of someone who had prepared for this conversation for months and whose material was true.
At one point Marsh pushed: "Is that your interpretation of the software log, or the standard scientific interpretation?"
Dr. Rowe looked at the jury rather than at Marsh. "I can explain the statistical consensus on adverse event reclassification thresholds if you'd like, counsel. There is a documented body of literature on what constitutes appropriate threshold adjustment in Phase III pharmaceutical trials. I'd be happy to walk through it."
Marsh declined the invitation.
She knew when a witness wasn't breaking.
After the testimony ended, I walked Dr. Rowe to the hallway.
"How do you think it went?" she said.
"The jury liked you," I said. "More importantly, they believed you. That's all that matters in there."
She nodded, slowly. She was tired — two full days of intense examination and cross-examination produced a specific kind of fatigue, the bone-deep exhaustion of sustained careful truthfulness.
"The program modification evidence," she said. "When you showed Exhibit 47. The courtroom—"
"I know."
"Did you see their faces?"
"Yes."
She looked down the corridor. Forty-seven plaintiffs who had filed into that courtroom every morning for four days. Who would be there until the verdict.
"It's not enough," she said. "What we're doing here. Even if we win. The drug is still on the market—"
"The FDA safety review opens next week," I said. "Cahill's team filed for emergency review status. Based on what's come out in testimony." I paused. "The trial and the regulatory track are running simultaneously. One of them will produce a market action."
She looked at me.
"You planned that," she said.
"We planned that," I said. "With Cahill. The public testimony is part of the FDA pressure track."
She was quiet for a moment.
"Thank you," she said.
It was the most simple version of the thing she'd been trying to say since January 2015 when she'd walked into my office with a banker's box.
"That's the job," I said.
A LOT More chapters reign FREE upon unwrittenrealm.com.
• Normal | $8 — 15 chapters ahead
• Advanced | $11 — 20 chapters ahead
• Overlord | $19 — 25 chapters ahead, plus request a refinement (plot, character, system) on any of my novels
patreon.com/TheFinex5
▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰
― DECREE ―
More chapters reign FREE upon unwrittenrealm.com.
The throne acknowledges.
▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰