Just getting it isn't scalable

I came to research as a consultant. A retina practice in Hampton Roads had one coordinator running its studies, and I was hired to help her find the inefficiencies and scale the department. She was doing every part of the job: finding potential subjects, running the visits, submitting the data, building tomorrow's binder. I offered to be her back end. She would run the visits, collect the data, and ship the samples; I would do the EDC entry and the binder prep for the next day. She agreed.

Then she quit, without warning. I found out from an email: an updated delegation log with her end date filled in as that same day. I was at Disney World, on my kids' first trip. I called her from in front of Cinderella's Castle and begged her to tell the CRAs it was an error. She said she was done.

So I took over. There were three subjects in two studies, and I had no idea what I was doing.

The first thing I did when I got back from Orlando was call the CRAs, explain what had happened, and ask for hand-held guidance on running these studies as accurately as possible with as few deviations as possible. The goal was to keep the subjects enrolled while I got access and got trained.

The vacation problem

Seven years later the department had enrolled hundreds of subjects across dozens of studies and the team had grown to seven-plus coordinators (plus techs, sub-investigators, photographers) and a referral network that ran from Northern Virginia to Elizabeth City. Phase 1 through 4, gene therapy, surgical, device, observational, investigator-initiated. I am proud of that. It is not the point of this note.

The point is a thing that happened over and over the whole time, and that I heard from coordinators at almost every investigator meeting I attended, back when those were in person.

A coordinator goes on vacation and asks a colleague to cover a visit. The colleague is good. They've run hundreds of visits. So they do the visit the way visits are done: take the blood pressure once, the way most studies take it. This protocol wanted it three times, five minutes apart, seated, averaged. Or they use the visit worksheets from the version of the protocol they last worked on, which an amendment replaced in March. Nobody notices in the room. It surfaces days later when someone enters the visit into the EDC and the field throws a query.

That is not a competence problem. Everyone in that story is competent. It is a handoff problem, and the handoff failed because the thing being handed off lived in one person's head.

The MVP and the Jaguar

I used to say our coordinators were each the LeBron James of their study. Meant as a compliment, and it was true: each one carried the whole thing, and you could not replace them. I also used to say they were like a Rolls-Royce. There is a reason those cars are rare and expensive. They are hand-built, they take forever, and you cannot scale a hand-built car.

Both comparisons are the same observation from two sides. The coordinator who "just gets it" is irreplaceable, and irreplaceable is exactly the property you do not want in an operation that has to survive a vacation, a resignation, or a third study starting on the same Monday.

I heard "you just have to get it" as the answer to how someone learns this job more times than I can count. It is true and it is not scalable, and I do not think clinical research has to work that way.

What "getting it" actually is

When I pulled apart what an experienced coordinator knows the night before a visit, it came down to a short list of questions.

1. Is the version of every document I am about to use the current one?
2. Is everyone who will touch this visit trained on it and delegated for it, with both dates before tomorrow?
3. Are the people I need actually the people on site tomorrow, certified, correctly masked, and on the log?
4. How do we do each procedure here, at this site, on this machine, with the stuff on the shelves?
5. What order do the steps go in, by the protocol's rules and by what works in our rooms, and is the visit inside its window?
6. Why is each step the way it is, so that when it does not go to plan, the person in the room can think?

The experienced coordinator answers all six from memory. The covering coordinator cannot, because the answers were never written down anywhere they could find them in the time they had. And the first answer to "where is that written" is usually "in the binder," which means: find the binder, find the tab, find the section, find the paragraph, hope it is the current version.

The lock

It is like locking the door at night and then lying in bed wondering whether you did. What you want is to see the latch turned with your own eyes. Better, to check it from your phone without getting up, going downstairs, and turning on the lights.

That is what I set out to build, and it is what CRC Run Sheet is. Not a CTMS. I used CRIO for three years and RealTime after that, and I like both; they are built for the study level and the invoiceable procedure, and getting down to the granular how-to of a visit became their "EDC" products, with the coding and the cost that implies. Not an EDC either, and not the source. A run sheet for each visit: every step in order, who is covered for it, how our site does it, where, and the protocol paragraph one click away so the person running the visit can read it instead of trusting it.

The site layer is a recipe. You write how your site does an OCT once, and the next study that has an OCT starts with it attached. A complete novice, given the recipe and the citation, has a high likelihood of running the visit the way our best coordinator would have. That is the whole bet.

The first recipe I ever saw was a how-to taped to one of our imaging machines. It walked a coordinator through getting the camera booted and ready so the photographer could walk in and go straight to the settings and the naming convention instead of starting from a cold machine. It was great until the tape tore during clinic, the sheet hit the floor, and it went out with that night's trash.

Who this is for

Before I left the practice I was in conversations with the state about a registered apprenticeship for coordinators. The idea was simple. If the process is documented well enough, you can bring in a high-school graduate, teach a, then b, then c, and they can execute a clinical trial visit without four years of college first, with a front-row seat to the research, a decent wage, and a ladder they can climb on the job or back through school. Duke and Washington University now run registered coordinator apprenticeships. It can be done.

I want to be careful about what that claim is. It is not that a coordinator's job can be done by anyone. It is that the coordinator should be the author of the process rather than its only copy. The author gets to go on vacation. The only copy does not.

That is why the first study is free, why there is no PHI in it, and why it prints: the sheet has to be something you can hand to whoever is covering, on paper, on the day. If you run a small site and you have been the only copy for a while, I would like to hear how you do it today. It reaches me, not a queue: support@crcrunsheet.com.

Chris Riebling coordinated retina trials at Wagner Kapoor Research Institute for over seven years and now runs Carbonic Software LLC. This note is his own account.