Keeping a Regulatory Position Current

Part of Regulatory Affairs and the Jurisdiction Map: Building and Defending a Global Position.

About 6 minutes. Free, in your language, no account needed.

Why this matters

Everything in this course describes work that was correct on the day it was done. The distinguishing question for a regulatory function is what happens on every subsequent day, because a regulatory position is not a deliverable. It is a state that must be maintained, and it degrades without any event to mark the degradation.

A trial authorized correctly, amended correctly and reported correctly can still end up non-compliant, simply because the framework moved and nobody was watching for it.

The three things that change underneath you

The regulation changes. An instrument is amended or superseded. The trial that was compliant under the old text may not be under the new, and transitional provisions determine which applies.

The trial changes. Countries are added, sites are added, protocols are amended, vendors change, and each change re-asks questions that were answered for a configuration that no longer exists.

The reference documents change. The Investigator's Brochure is updated, which changes what counts as expected, which changes what must be reported on an expedited clock. This one is particularly easy to miss, because the update is a safety activity and its regulatory consequence is downstream.

Note what the second and third have in common with the first: none of them generates an alert. They are all things that happen, correctly, elsewhere.

What a maintenance system actually looks like

A calendar of dated obligations. Every periodic report, every continuing review anniversary, every authorization expiry, every commitment made to an authority with a date attached. Each entry has an owner. This is unglamorous and it is the single highest-value artifact a regulatory function maintains.

A re-verification cycle on the jurisdiction map, prioritized by where activity actually is rather than alphabetically.

A change trigger wired into the events that matter: a protocol amendment triggers the classification matrix, an Investigator's Brochure update triggers a review of what is now expected, a new country triggers a full jurisdiction verification.

An alarm on absence. This is the part almost everyone lacks. Most regulatory failures are not wrong actions but absent ones, and absence generates no event. Something must look for the thing that did not happen: the report not submitted, the row not re-verified, the commitment not met.

Each of the first three tells you about work you are doing. Only the fourth tells you about work you are not.

Handover, which is where positions die

Regulatory positions are most often lost at handover, and the mechanism is consistent.

The outgoing person hands over what is active: submissions in flight, questions open, deadlines approaching. All of it visible, all of it correctly transferred.

What does not transfer is the reasoning. Why this country was classified as it was. Which claims were verified and which were assumed. What was raised with an authority and left unresolved. Which commitment in a three year old letter is still live.

The incoming person inherits a working system and no idea which parts of it are load-bearing. They will not discover this until something breaks, and when it breaks the answer will be that it was known once.

The countermeasure is to write the reasoning down as you go rather than at handover, because at handover there is no time and the reasoning is what gets dropped.

In practice

A regulatory lead takes over a program and spends her first three weeks doing no submissions at all.

She builds three things. A calendar of every dated obligation across nine countries, with an owner against each, which surfaces two overdue items and eleven with no owner. A verification status for each country's map row with the date it was last checked, which shows four rows unverified for over three years in countries carrying active sites. And a list of every commitment made to an authority in the program's history with its current status, which finds three live commitments nobody was tracking, one of which had not been met.

Her manager asks why nothing has been submitted.

The answer is that the program had three live problems that nobody could see, and the mechanism that should have surfaced them did not exist. Building it took three weeks. Discovering those three problems at inspection instead would have cost considerably more, and the third one, an unmet commitment, would have been found by the authority that was owed it.

What this looks like when it goes wrong

A sponsor runs a study for four years without incident. Every submission made, every report on time.

An Investigator's Brochure update in year two removed an event from the reference safety information, because the accumulating data no longer supported it as expected.

That change made the event unexpected from the date the new version took effect. Six events of that type occurred afterwards and none was reported on an expedited clock, because the safety team's assessment rule had been set when the previous version was current and nobody re-derived it.

The Investigator's Brochure update was correct. The original assessment rule was correct. Nothing was done wrong at either point. The failure sits entirely in the space between them, where a change in one document should have triggered a review in another function and there was no mechanism to make that happen.

That is the characteristic shape of a regulatory failure in a mature organization. Not an error, but a missing connection between two correct things.

Key takeaways