Investigate an OOS Result
A finished-product assay comes back below specification. Walk it from the bench through the laboratory investigation, hypothesis testing, the full-scale investigation, and batch disposition — and see whether your record would survive an inspection.
What a scenario is not
A scenario is a judgment exercise, not an assessment and not a compliance verdict. Finishing it evidences that you worked through it — nothing more — and the dimensions it scores are SPEQ’s, not any regulator’s.
A test result failed. What you do next decides whether you find the truth about the batch — or bury it. Some of the tempting shortcuts are exactly what regulators watch for.
An out-of-specification (OOS) result must be investigated before it can be discounted. A result is only invalidated on a documented, assignable laboratory cause; retesting follows a pre-defined scientifically sound plan, never "test until it passes"; and if the lab investigation finds no cause, the investigation widens to manufacturing. Each choice here carries forward.
The OOS procedure is where data integrity is proven or lost. The Barr ruling and FDA’s OOS guidance were written around the shortcuts below; an inspector will pull exactly this thread.
You are a QC analyst. A finished-product assay reads 92.1% of label — the specification is 95.0–105.0%. The result is confirmed as reported by the instrument; it is not a data-entry slip.
Nothing here is a trick. Choose what a controlled laboratory would actually do — the situation will change based on your choice.
The assay is out of specification. No one else has seen the result yet.
This is a judgment exercise, not a quiz: your choices change the situation. It is scored across rigor · integrity · rootCause · disposition — the same dimensions an inspector effectively weighs when they follow the thread of a real deviation. There is no compliance verdict here; the debriefs explain consequences.
Finishing this exercise is recorded. What it evidences — and the things it does not — is set out against the published standard.
What this evidences →