IRT Supply Settings: The Hidden Levers That Make or Break Trial Supply

Interactive Response Technology (IRT — also called RTSM) is the system that randomizes patients and manages drug supply across a trial: assigning kits, tracking inventory, and triggering shipments to sites. Most sponsors treat IRT configuration as a systems task delegated to the vendor. That's a mistake. Buried in the IRT's supply settings are a handful of parameters — buffer levels, resupply triggers, prediction windows, pooling rules — that quietly determine whether a trial ships tons of overage it will destroy, or gambles with stock-outs that pause dosing. IRT setup is supply strategy, expressed in configuration.

What do the core supply settings actually control?

Four families of settings do most of the work:

• Trigger and buffer levels define when a site gets resupplied and how much safety stock it holds — set them high and every site becomes a small warehouse of expiring product; set them low and a weekend enrollment surge empties the shelf.
• Predictive resupply windows tell the system how far ahead to look at scheduled visits when calculating demand — the difference between reactive shipping and anticipatory shipping.
• "Do not ship" and expiry rules govern whether kits close to expiry can still be assigned — the master valve on both waste and risk.
• Pooling and kit-sharing logic determines whether inventory is fenced per study arm, per protocol, or shared across a program — often the single largest overage lever available to a sponsor running multiple studies on one product.

Why does misconfiguration cost so much?

Because IRT settings multiply across every site, every month, for the life of the trial. Industry experience is consistent: clinical trials routinely manufacture far more drug than patients ever take — overage rates well beyond half of produced supply are commonly reported — and generic, over-cautious IRT defaults are a principal driver. Every unnecessary buffer kit is manufacturing cost, comparator cost (brutal when the comparator is an expensive commercial biologic), cold-chain shipping cost, and eventually destruction cost with its own documentation burden.

The opposite failure is worse: a stock-out at a dosing visit is a protocol deviation, a patient risk, and in competitive enrollment environments, a site-relationship injury that outlasts the trial.

How should settings differ across the life of a trial?

Static settings are the tell of an unmanaged system. Well-run trials tune by phase:

• Start-up favors generous initial site stocks and short prediction windows, because enrollment forecasts are guesses and site activation is lumpy.
• Steady state is where tightening pays — real enrollment and usage data replace assumptions, buffers come down, prediction windows extend, and shipment frequency finds its economic optimum.
• Close-out flips the objective entirely, throttling resupply so sites finish near-empty and returns and destruction shrink.

Expiry management runs across all phases: as lots age, "ship oldest usable first" logic and expiry-aware assignment protect patients while minimizing write-offs — provided someone actually maintains the expiry data feeding the system.

What does good IRT supply governance look like?

Three practices distinguish sponsors who control their supply from sponsors who watch it happen. First, a supply strategist in the design room when the IRT is specified — the settings should implement a modeled supply plan (simulated against enrollment scenarios), not vendor defaults. Second, a standing review cadence — monthly during enrollment — comparing forecast versus actual usage by site and adjusting triggers, buffers, and windows as reality reveals itself. Third, change discipline: setting changes mid-trial ripple into blinding, projections, and depot operations, so they flow through controlled change processes — agile, but never casual.

The IRT will faithfully execute whatever strategy it's given. The question is whether anyone gave it one.