Preventing and Managing a Stockout

Part of Clinical Trial Assistant (CTA) Training: Supply Forecasting and Stockouts.

About 4 minutes. Free, in your language, no account needed.

Why this matters

A stockout means a participant cannot receive treatment on the day the protocol says they should. That is a protocol deviation, it may be a participant safety issue depending on the product, and in a blinded study it can compromise the data.

It is also nearly always preventable.

Prevention

Set a trigger level and act on it without debate. When cover falls below lead time plus buffer, a resupply is requested. Not discussed, requested.

Keep buffer proportional to uncertainty rather than uniform. A site in a country with unpredictable customs clearance needs more buffer than one supplied domestically. A site enrolling erratically needs more than one enrolling steadily.

Watch expiry as carefully as quantity. Rotating stock so the shortest-dated units are used first is standard practice and is frequently not done, which produces waste and a false sense of supply.

When a stockout is imminent

Say so immediately and precisely: which site, which participants, which visit dates, and how many days short.

The specificity matters because the mitigations available depend on it. One participant three days short is a scheduling conversation. Four participants two weeks short is a supply escalation.

The mitigations, roughly in order of preference:

Expedite resupply. Often possible at cost, and the cost is almost always less than a lost visit.

Reschedule within the protocol window. Many protocols allow a visit window of several days. Moving inside the window is not a deviation. Moving outside it is, and needs recording.

Transfer between sites. Possible in some studies and forbidden in others. It requires sponsor authorization and full documentation of the transfer, because the custody chain must remain unbroken. Never arrange this informally between sites.

Accept the deviation and document it. Sometimes the right answer. A documented missed dose with a clear record is far better than an undocumented workaround.

What not to do

Do not dispense expired product. Do not dispense quarantined product while a disposition is open. Do not move product between sites without authorization. Do not unblind to make the supply work.

Each of these has happened, each was done by someone trying to help a participant, and each converted a supply problem into a data or compliance problem that outlived the study.

Afterwards

Every stockout has a cause, and it is usually a forecasting or lead-time assumption that was wrong. Find it and change the number. A stockout that recurs at the same site is an unexamined assumption.

A worked escalation

Site 305 will be two kits short for two participants whose visits fall on 14 and 16 October. You find this on 2 October.

The message. "Site 305, two participants short. Participant 3051 dosing 14 October, 3052 dosing 16 October. Current stock covers neither. Resupply requested 30 September, standard lead time to this country is 19 to 31 days, so it will not arrive in time. Options: expedite at cost, or move both visits inside the protocol's plus-or-minus-three-day window to the 17th and 19th, which the resupply would likely make."

That message contains the problem, the arithmetic, and two viable options. It gets a decision the same day.

What it does not contain. Any suggestion of borrowing from site 306, because inter-site transfer requires sponsor authorization and full custody documentation, and proposing it casually invites somebody to arrange it informally.

After the event

Site 305 ran short because the resupply trigger used a 19-day lead time and the real distribution runs to

  1. That is one number in one spreadsheet, and changing it prevents the next occurrence.

Most stockouts trace to an assumption like that: a lead time taken from the contract rather than from experience, an enrollment rate from the plan rather than the actuals, a buffer set uniformly across countries with very different customs behavior.

Find the number, change it, and say which one you changed. A stockout with no identified cause will recur, and the second one is much harder to explain.

Key takeaways