The Allocation System Is the Record

Part of Trial Pharmacy (PHARM) Training: Randomization and the Blind.

About 2 minutes. Free, in your language, no account needed.

Why this matters

On a randomized trial, which participant gets which kit is not a pharmacy decision. It is an allocation, made by a system, and recorded.

Your job is to execute it exactly and to record that you did.

How allocation usually works

A participant is randomized, often through an interactive system. The system assigns a treatment arm and, at each dispensing visit, a specific kit number.

The kit number is the instruction. Not the drug name, not the dose written on the box.

Two kits containing identical product are not interchangeable if the system assigned one of them.

Stratification

Many trials randomize within strata, such as disease severity or prior treatment.

That means supply must be adequate in each stratum, not just overall. A pharmacy with plenty of stock can still be unable to dispense to the next participant.

Check availability by stratum when forecasting.

When the system is unavailable

There is a documented backup procedure. It usually involves a manual allocation from a sealed list, with a defined approval and a requirement to reconcile in the system afterwards.

Follow it exactly and record everything. Do not improvise an allocation and plan to tidy it later, because the system will not accept a retrospective entry that contradicts its own sequence.

In practice

The system assigns kit 0207. The shelf holds 0207 and 0209, both the same product and dose. 0207 is at the back.

You use 0207. Reaching for the front box because the contents look identical breaks the allocation record and may unblind, since kit numbers often encode arm in ways not visible to you.

What this looks like when it goes wrong

A pharmacy substitutes a kit during a busy clinic and documents the substitution carefully.

The documentation does not repair it. The randomization record now says the participant received one kit and the accountability says another, and the sponsor must decide whether that participant's data is attributable.

Careful documentation of a broken allocation is still a broken allocation.

Key takeaways