Three Positive Units in a 5,000-Unit Media Fill, Confirmed After Twelve Batches Were Released

What the data say: any positive unit in an aseptic process simulation requires a documented investigation, and the pattern here is not random. Three positives, all at head 4, plus a Grade A settle plate at the limit and an air-velocity reading at the very bottom of the acceptance band, point to a location-specific loss of first-air protection rather than a one-off handling error. Two different organisms (a skin commensal and an environmental Micrococcus) indicate that more than one route may be involved, or that a single compromised surface is shedding.

Investigation before any conclusion:

1. Verify the simulation itself: growth-promotion testing of the medium, incubation conditions, vial integrity, and handling of positives, to exclude false positives.
2. Microbial characterisation: identify the isolates to strain level and compare with isolates from head 4 swabs and with the operator's glove and gown flora.
3. Disassemble head 4 - nozzles, valve, tubing, filling needles - inspect for damage or product residue, and swab for bioburden.
4. Review airflow visualisation at head 4 and re-measure velocity with a calibrated instrument; the 0.36 m/s reading needs explanation, not acceptance.
5. Review video of the jam clearance and all 40 interventions, and assess whether the intervention procedure was followed.
6. Review operator qualification status, gowning records and the glove/gown counts, which are elevated. Only after that can the result be closed as a defined root cause rather than an unexplained failure.

Treat them as potentially affected until proven otherwise, and act before the investigation closes, because patient exposure is ongoing.

1. Quarantine any remaining inventory of those batches immediately and block further release or shipment; issue a formal hold notice.
2. Assemble the batch records, environmental monitoring, personnel monitoring, intervention logs and sterility test data for each of the 12 lots. Where the same line, shift, head or operators are implicated, weight the risk accordingly.
3. Perform a health-hazard assessment. The sterility test samples a small number of units and cannot assure a lot; the media fill is the stronger signal. If the investigation supports a loss of control at the time of filling, a recall or field action should be assessed rather than avoided.
4. Apply the local regulatory requirements for notifying the authority, and involve the qualified person or authorised person early; do not let the notification deadline be discovered late.
5. Communicate with the clinical customer about the lot in use, in coordination with the quality head.

Do not argue that 'the batches passed their release tests' as the basis for doing nothing. The failure is in the process control, and released status does not answer the question of whether the process was in control when those vials were filled.

Preconditions, in order:

1. Root cause established and the corrective action implemented and verified - engineering repair of the head-4 airflow or valve defect, with velocity, airflow visualisation and particle data confirming restoration of Grade A conditions.
2. The critical zone requalified, including airflow visualisation at each head with the line running and with the doors in the normal operating position.
3. The intervention procedure revised so that a jam clearance does not require opening the critical-zone doors, or does so with a validated tool and technique; operators retrained and requalified for gowning and interventions.
4. A cleaning and disinfection programme for head 4 with bioburden verification, and resolution of the elevated glove and gown counts (gowning qualification, glove change frequency).
5. Repeat aseptic process simulations after the corrective actions, to the number required by your procedures and the applicable guidance, with zero positives and acceptable environmental monitoring before commercial filling.
6. A documented change control, a completed deviation and CAPA file with effectiveness checks, and quality approval to restart. Notify the authority if required.

Five months to the next scheduled requalification is not a reason to postpone any of this; the schedule is a minimum, not a ceiling, and the line is not in a qualified state until these are met.