Question 20 of 20 Advanced Troubleshooting
A sterility test shows growth. When can the result be invalidated?
Short answer
Only when conclusive, documented evidence shows the growth came from the test itself, not the product. If the evidence is inconclusive, the batch fails. The investigation must cover both laboratory and manufacturing, using species identification and comparison with environmental and personnel isolates.
Why
The FDA's aseptic processing guidance says that when growth is seen, the lot should be considered nonsterile and investigated. An initial positive is invalid only when growth can be unequivocally ascribed to laboratory error, and when evidence is inconclusive the batch should be rejected. The reason for this strictness is the test's weak sensitivity.
The guidance gives an example: a 10,000-unit lot with 0.1 percent contamination tested with 20 units has a 98 percent chance of passing. A positive is therefore rare, valuable evidence. The investigation identifies the organism to species level and compares it with isolates from the testing lab, the production environment, personnel and product bioburden, using the same identification method.
It reviews lab deviations, controls and the test environment. The guidance warns that if production facilities are much better controlled than the test lab, a real product contamination may be wrongly blamed on the lab. If the organism is rare in the lab, product contamination is more likely.
If it is found in both the lab and production, that can still mean product contamination. Using an isolator for sterility testing reduces false positives and makes the argument cleaner. Deviations that compromise a test should invalidate it immediately, before incubation, not after growth appears.
The whole manufacturing process should also be investigated, since sterility problems often extend beyond one batch.
How to work it through
- Treat the lot as nonsterile and quarantine it and any related lots.
- Identify the organism to species level, ideally with a genotypic method.
- Compare it with lab, production, personnel and bioburden isolates using the same method.
- Review test controls, negative controls, environment and analyst records for documented failures.
- Review the batch's manufacturing and aseptic records and environmental data.
- Invalidate only on conclusive evidence of a test fault, otherwise reject the batch and extend the investigation.
An example
One of 20 units in a sterility test grows Micrococcus luteus on day 9. The test was run in a grade B room, not an isolator. Lab monitoring that day shows Micrococcus luteus on the analyst's glove print, and the negative control was clean.
The production grade A and B data for the batch show no Micrococcus, though one grade B air sample was over alert. Matching species in the lab is suggestive, but strain typing is not available and the production excursion is unexplained. The evidence is not conclusive, so QA rejects the batch, opens a manufacturing investigation and moves sterility testing into an isolator.
Where the answer stops holding
Species matching is weak evidence for common skin organisms, so strain typing strengthens a conclusion. Rules and wording differ slightly between pharmacopoeias and regions, so the site procedure must reflect the applicable texts. A negative sterility test never proves a batch sterile, which is why process controls matter more.
A strong answer covers
- Lot considered nonsterile on growth
- Invalid only on unequivocal evidence of laboratory error
- Inconclusive evidence means rejection
- Species identification compared with lab and production isolates
- Low test sensitivity explains the strict rule
Common mistakes
- Invalidating a positive because the organism is a common lab contaminant.
- Retesting and releasing on a passing retest without conclusive evidence of lab error.
- Investigating only the laboratory and not the manufacturing process.
The interviewer might ask next
How would moving sterility testing into an isolator change the way you investigate a positive?