One unit in an aseptic process simulation shows growth. What happens next?

Why

An aseptic process simulation, or media fill, replaces product with growth medium and runs the filling process with its normal interventions. Annex 1 says the target is zero growth and that any contaminated unit results in a failed simulation. The required actions are an investigation into the most probable root cause, corrective measures, a prompt review of aseptic production records since the last successful simulation with a risk assessment of possible sterility breaches, inclusion of unreleased batches in the scope, quarantine of product made on the line since the failure, and normally a minimum of three consecutive successful repeats.

This is stricter than the FDA's 2004 aseptic guidance, which allows one contaminated unit in runs of 5,000 units or more to trigger an investigation and considers two the cause for revalidation, while still calling any contaminated unit objectionable. A company supplying both markets in practice works to the stricter rule. The investigation uses the organism's species identity to point at a source.

Skin flora suggests an operator intervention, a spore former suggests the environment or a sterilization gap, and a water organism suggests utilities or a container wash. The time the unit was filled, from video or line records, links it to a specific intervention. The limit is that a simulation estimates the process's capability with a few thousand units, so passing does not prove every commercial unit sterile.

How to work it through

1. Confirm the growth is real, protect the unit and identify the organism to species level.
2. Quarantine product made on the line and stop further aseptic runs pending assessment.
3. Link the contaminated unit to its fill time and the interventions performed then.
4. Compare the organism with environmental, personnel and utility isolates using the same identification method.
5. Review records since the last successful simulation and assess the sterility risk to each batch, including released ones.
6. Implement corrective actions, then run consecutive repeat simulations, normally at least three, before resuming.

Where the answer stops holding

Matching species does not prove the source, because skin organisms are common, so strain-level typing helps where it is available. Repeat simulations show the line is back in control but say little about batches already filled. Rules differ between Annex 1 and FDA guidance, and the stricter one usually applies for a multi-market line.