By Justin Yamashita, MSc. Benchtop, site, CRO: three levels of basic and clinical research, explained without spin.
THE ROOT ROOM
Where the characters work through the week's evidence.
How the tracks work together: The Root Room is where you watch a bad move actually work on somebody, and where misinformation starts with our version of Black Box Warnings. It shows you how to stay in the room and have a hard health conversation, using the same evidence Plain Talk breaks down and The Informed sources.
Last issue, Sam the Skeptic sat in on a site selection meeting and watched a map lose an argument to a calendar. That was about who gets studied. This one is about who is watching once the study is running, and what they’re allowed to do with the data very few others see.
The Meeting That Ended a Study 3 Years Early
On 31 May 2002, a small group of biostatisticians and physicians met and looked at data from a trial of 16,608 women, concluding that it shouldn’t keep going.
None of them worked for the trial, not for the 40 clinical centers running it, not for the institute funding it, not for the company supplying the pills. The doctors (investigators) who saw those women every few months had no idea the meeting had gone that way. They could not have known. They had been kept blind to those numbers from the first day, on purpose, by design.
The women found out by mail. Letters started going out on 8 July 2002 telling them to stop taking their study pills.
That group had a name almost nobody outside this industry has heard: a Data Safety Monitoring Board (DSMB), sometimes called an Independent Data Monitoring Committee (IDMC). It’s the meeting you’re not allowed to sit in. Neither is anyone running your trial, at least not for the part that matters.
Root to Rx is free, three times a week, written by a clinical researcher. Subscribe at RootToRx.com, or help us grow and send it to someone you know. I run this on my own, so every share helps more than you would think.
The Report She Will Never Get to Read
The person who builds the safety record is not the person who gets to read what it says, as Coordinator Nani and Rooty the Researcher explain to Dennis what happens to his wife’s data.
"Who actually looks at this," says Dennis, whose wife enrolled 6 weeks ago, nodding at the screen. "Or does it sit in a drawer until the company decides it likes the answer?"
He isn’t being rude. He’s read enough from the internet to be worried about the wrong things.
Nani, the coordinator on this study and a nurse before that, doesn’t look up from the table she’s cleaning. She’s been doing this for 19 years, which means she’s had this conversation in this doorway more times than she can count.
"I help collect the data and build the record," she says. "I don't get to read the ending until the study closes and the results are published."
"What's that supposed to mean."
"It means every adverse event, every lab value, every visit your wife has, I write it down accurately and I send it forward. What I don't do is see how the groups compare, and plenty of studies have more than 2. Nobody here does. Not me, not the pharmacist, not the doctor running this site."
Dennis actually sits down. "You're telling me the people running the study don't know if it's working."
"Not yet, no. That's not a mistake. That's the design, and it's how we protect the quality of the study."
Sam the Skeptic, whose whole job is asking how do we know, leans toward Rooty the Researcher, who brings the receipts. "Okay, that sounds bad. That sounds like nobody's minding the store."
"It sounds like that until you ask why," Rooty says. "If the doctor treating you knows which group you're in, and knows the drug is looking good so far, everything changes. What she writes down changes. What she asks you changes. What she talks you into changes, and she won't even notice it happening. Blinding isn't secrecy. It's the thing that keeps the record honest while it's being written."
"Then who does see it?"
"A different room, and I don’t mean one down the hall. Statisticians and physicians with the relevant expertise and no financial stake in how this turns out. They get the unblinded comparison at set intervals through the study, and they’re the only ones who do."
Nobody Votes. They Look.
Dennis is still on the part that bothers him. "And if they see something wrong, what, they write a memo or the drug company hides the data?"
"They can recommend the trial stop," Nani says. "That's the actual mechanic and it's worth getting right, because people oversell this. The board doesn't have a switch on the wall. It reviews, it makes a recommendation, and the sponsor acts on it."
"So the sponsor can just ignore them."
"On paper. In practice, try explaining later why your own independent board told you to stop and you kept going anyway." She turns the laptop away. "In 2002, a national hormone therapy trial enrolled more than 16,000 women and was planned to run 8.5 years. At a regularly scheduled meeting, the DSMB saw invasive breast cancer cases cross a line that had been set before the first woman enrolled. They recommended ending that arm. It ended, 3 years early. The people running it found out afterward like everybody else."
"Because they were blind."
"Yes, because they were blind."
Sam frowns. "Wait. If the board can only recommend, and only the sponsor can act, is anybody required to have one of these at all?"
"Good question and the answer is annoying," Rooty says. "The ethics review board that clears a study before anybody enrolls, the IRB, is required by federal regulation. The safety monitoring board isn’t, in general. FDA strongly recommends one when people in the trial face serious harm or death, and there’s exactly one situation where the regulation flatly requires it. Otherwise it’s a very strong norm with a very short leash."
"So it's a habit."
"It's a habit with 50 years of scar tissue behind it, and habits can be broken. That's the part worth carrying out of here."
The Line Was Drawn Before Anyone Enrolled
Sam isn’t done. "Who decided what counts as bad enough to stop?"
"They did, and they did it first," Nani says. "Before the first site opens for enrollment, the team running the trial writes a charter and the board has to agree to it. The charter says how often they meet, what they look at, and where the lines are. These are people with real expertise, medical and statistical, in the disease being studied. So the line might be, if breast cancer cases in the treatment group reach this level, we act, and that level gets set by people who actually treat breast cancer. That number is on paper before a single person signs a consent form."
"Why does that matter so much?"
Rooty takes this one. "Because if you set the line after you’ve seen the data, you’ll set it somewhere convenient. Everybody would. Not out of malice, out of human nature. You look at a graph, you see a bump, and suddenly you have a very reasonable story about why that bump isn’t the kind of bump that counts. Writing the rule down first is how you stop yourself from doing that later. Moving it afterward has a name, moving the goalposts, and it isn’t how clinical research works."
"So the rule protects them from themselves."
"The rule protects the answer from all of us. Same reason a referee isn’t allowed to bet on the game, and same reason you decide the household budget on a Sunday and not while standing in the store."
Dennis has been quiet. "And nobody at this office can see any of it."
"There’s some nuance, as there always is. In phase 1, first in human work, there are safety review meetings where the investigators and the study team sit down together, look at the adverse events and the findings so far, and decide whether to move up to the next dose level. But for a phase 3 study like this one, nobody at this office can see any of it," Nani says. "I’ll know how your wife is doing. I’ll never know how the study is doing. Those are 2 different questions and I only get one of them."
She enters the data, and sponsor and CRO (generally) create the report. It goes out sealed. She never sees what it says about the group her patients are in.
"Nobody told me that meeting room existed," Dennis says.
"Most people never hear about it," Nani says. "You did, because you asked. That's the whole difference. You asked, and now what you believe is built on something you can check."
The Black Box Warning: The Empty Room
The Empty Room. The claim works because you cannot see into the room, not because the room is empty.
Black Box Warning
Tactic: The Empty Room.
Tool: point at a process nobody outside it can see, and let the reader's imagination fill the space.
Running it: "nobody is independently watching a trial once it starts, they can do whatever they want."
Here’s why this one is worth learning. The tactic never says the room is empty. It just points at a closed door, and closed doors do the rest of the work.
Move 1. A wellness influencer can pick something genuinely invisible from the outside. Not hidden, just invisible. Nobody’s concealing a data monitoring committee exists. There’s simply nothing for you to look at, because the whole design depends on almost no one seeing the data.
Put it this way. You can’t see the kitchen from your table at a restaurant. That doesn’t mean nobody’s cooking.
Move 2. Say what’s not there instead of what is. "Nobody’s watching" can’t be disproven by a person standing in a hallway, and nobody scrolling past is going to stop and look it up. That’s the entire trick.
Put it this way. If I tell you there’s no lifeguard at the beach, you’d have to walk the whole shoreline to prove me wrong. Nobody does that, so the claim just sits there.
Move 3. Collect the agreement. Everybody in the room has been on the outside of a closed door somewhere, and they nod.
Put it this way. Every one of us has sat in a waiting room while people talked about us behind a door. That feeling is real, and that is exactly what gets borrowed.
The move to keep
Turn "nobody is watching" into a question with an answer. There are 2 named rooms here. One clears the study before anyone enrolls, and it’s required. One watches the data while the study runs, and it’s strongly recommended rather than required for most trials. Ask which of those 2 the person means. Almost nobody making the claim can name either, and the ones who can are usually making a much narrower point that deserves a real hearing.
Sam the Skeptic: "You can't check an empty room. You can check a name."
Take This One With You
You’ll hear both halves of this, probably from people who care about you.
When somebody says nobody independently checks a trial once it starts:
"There are 2 separate groups, and neither answers to the people running the study. One clears the design and the consent form before anyone enrolls. The other looks at the safety data while it runs and can recommend stopping it, which is exactly what happened to a national hormone trial in 2002."
When somebody says oversight is airtight, so relax:
"An ethics board is required. A safety monitoring board mostly isn’t, outside high-risk trials where the FDA strongly recommends one, and an ethics board doesn’t watch the data. The FDA has said in its own rulemaking that IRBs are not in a position to follow subjects through a study."
Both of those hand the other person something checkable. Neither one calls anybody stupid. That’s the entire method, and it’s the reason people keep talking to you after you use it.
Next issue: the manufacturing floor, and why the same molecule made in 2 different plants is not automatically the same product.
Send this to one person
Not "someone who should know." One person. The friend whose sister just enrolled in something. The coworker who says trials are the wild west. Wednesday's Plain Talk is the one that names exactly what each room checks, and it’s free at RootToRx.com.
Want to practice the move before you need it? Stay in the Room is our 2 minute game on holding a hard health conversation without either person walking out. Free, no signup, at openlabelmedia.com/games.
Refer a friend - it helps a lot
You just watched the people behind your medicine do the work. Send this to someone who needs to trust them, and collect a Series 1 card while you're at it.
New here? Root to Rx runs three pieces a week. Monday you watch a real health claim try to work on someone. Wednesday you get the tools to run it yourself. Thursday you get the full evidence so you can check my work. Written by a clinical researcher, free, at RootToRx.com.
For educational purposes only. Nothing in this newsletter is medical advice. Talk to your doctor before making any health decision.
Instagram: instagram.com/roottorx | LinkedIn: linkedin.com/company/roottorx | Facebook Lab: facebook.com/groups/theroottorxlab | Subscribe: RootToRx.com
You are receiving this because you subscribed at roottorx.com.
Disclosure: Root to Rx is published independently by Open Label Media LLC. Views expressed are personal views of Justin Yamashita and do not represent his employer or any affiliated organization. No employer resources or proprietary information are used. Every claim is sourced from publicly available materials.
Root to Rx | 228 Park Ave S, #29976, New York, New York 10003, United States