Out of specification top  five questions and answers pharma

Can we average an OOS result with passing retest results to get a passing average?

A: Absolutely not. Both FDA and MHRA guidelines strictly prohibit averaging an OOS result with passing retests to “hide” the failure or bring the batch into compliance. Averaging hides the inherent variability of the product. If a retest is performed because a lab error is suspected but not conclusively proven, all individual data points must be reported and evaluated. A single OOS result can be a true reflection of a non-homogeneous batch.

Q2: What is the difference between an OOS (Out of Specification) and an OOT (Out of Trend) result?

A:OOS: A result that falls outside the legally registered, pre-determined acceptance criteria or specification limits (e.g., an assay specification is 95.0%–105.0%, and the result is 94.2%). 

OOT: A result that is still within the official specification limits but falls outside the historical statistical expectation or trend of that product line or stability study (e.g., historical data for an assay is always between 99.0%–101.0%, and a new batch comes in at 96.0%). It is within specification but indicates potential process drift.

Q3: When is “Resampling” permitted during an OOS investigation?

A: Resampling (collecting a completely new sample from the manufacturing line) is only permitted if the original sample was physically compromised, contaminated, or if the original sampling plan was proven to be non-representative or flawed. It cannot be used simply because the original sample gave an OOS result.

Q4: If an OOS occurs during an In-Process Control (IPC) test, do we have to follow the OOS guidelines

A: It depends on the purpose of the test. If the IPC test is used for a hard batch release decision, is part of the regulatory dossier, or is used for batch calculation, the OOS guidelines apply.

However, if the IPC is a routine monitoring tool designed to trigger real-time equipment or system adjustments (like checking pH to adjust an acid/base addition to a tank to achieve an endpoint), it is handled as a process deviation or adjustment rather than a formal OOS. 

Q5: Can a batch be released if the OOS investigation is inconclusive?

A: No. If a thorough investigation is conducted and no clear laboratory or manufacturing error can be identified as the root cause, the original OOS result must be considered valid. The batch cannot be released to the market and must be rejected. The Quality Unit must always err on the side of caution to protect patient safety.