One contaminated unit in a media fill fails the aseptic process simulation under EU GMP Annex 1, where the target is zero growth.
The investigation looks at the process, the people and the environment. For the environment, the first evidence is the viable monitoring from the fill itself: active air samples, settle plates, contact plates, and glove and gown samples from the operators. The organism recovered from the failed unit is compared with what that monitoring found.
The HVAC and EMS record sits beside it. Differential pressures, Grade A airflow, continuous particle counts, temperature and humidity show whether the room held its qualified conditions through the fill. Stable, in-limit trends support a conclusion that the environment was not the source, though on their own they cannot settle it. A gap in those trends leaves a question the investigation has no data to close.
Three things are worth doing before the fill. Review the HVAC trends for the preceding days: a differential drifting down or a filter near its change point is easier to deal with then. Confirm that monitoring is logging, controller clocks are synchronized and backups ran. And agree with QA which HVAC events during the fill count as an interruption, while nobody yet knows the result.
We write the validation strategy alongside the design, so the monitoring behind this record is specified from the start.
Industries




