MTS Helpdesk
Question:
Why does the guidance from MHRA regarding investigation of out of specification (OOS) results not allow retesting in phase 1?
Answer:
The major issue with retesting in phase 1 is that it is difficult to interpret the results and therefore it is difficult to make any meaningful conclusions about the quality of the material under test. If a sample is retested, by which we mean that the full analytical procedure is performed on the test material, and the result is within the specification, there are now two results: the original OOS and the in-specification retest. There is no scientifically or statistically valid reason why you can choose the in-specification result over the OOS result.
Since the entire method has been performed on the additional portion of the sample taken, it is not possible to definitively prove any hypothesis about what might have happened in the initial analysis to cause the OOS result. This is why hypothesis testing that involves re-measurement, where you are able to definitively isolate an error in the original analysis is acceptable, but retesting is not.
In addition, the results that are generated from retesting are valid data for the batch in question and final reporting requirements need to be considered. In remeasurement, any data collected is only for the purposes of the investigation to prove or disprove a hypothesis and as such are not required to be retained with the batch as final reportable results.
Retesting is performed in phase 2 of the investigation at a statistically valid level, most accurately referred to as ‘additional’ testing’. The extra replicates prepared from the original sample provides increased statistical confidence in the result obtained and allows comparison of the retests against the original. This enables more meaningful conclusions to be reached regarding the whether the original analysis is valid or not.
It is worth noting that even in this case, the original result cannot be invalidated unless the underlying assignable cause is identified. If the retests are all within specification and it is clear that the original result is different, and not just out of specification due to the variability of the method, then you could reasonably conclude that an unknown issue occurred during the original testing. Without invalidating the original OOS result, the QP could decide to release the batch if all other data was acceptable. In the MHRA and FDA guidelines, in this instance it is recommended to err on the side of caution.
On some occasions, there may be evidence of a potential hypothesis which cannot be investigated without a retest. For example, if it is suspected that an error occurred during a weighing of the sample, this is very likely the first step in the method and to investigate it would necessitate following the full procedure, and therefore it would be a retest. In this case, it would probably be best to accept that the hypothesis cannot be proven in phase 1 and thus move the investigation to phase 2, where additional testing is performed. Depending on the strength of the evidence of the weighing error, it may be possible to invalidate the original OOS result in this situation.
Another interesting potential example where hypothesis testing would involve retesting would be in the case of a suspected injection error for a chromatography analysis, where the sample is injected ‘as is’, without any sample preparation. In this case, the hypothesis test of performing a reinjection would be able to prove the hypothesis and may be considered. However, the nature of the results that would be generated would need to be carefully considered.
In these less straightforward cases, a sound, scientific approach is recommended where all decisions are justified in a thorough documentation of the investigation.
In conclusion, the reason why retesting is not performed in phase 1 of an OOS investigation is that on most occasions, it does not provide meaningful results and therefore does not actually assist in the investigation.
Submit a Question to the MTS Helpdesk
Do you have a question for our expert relating to the topic of analytical chemistry? Use our enquiry form to send it to us.
Related content from our Resources Library
A full discussion of OOS/OOT investigations is provided in our course, ‘Investigating Out of Specification (OOS) Results‘.
Visit our courses page for a full list of available courses and the schedule of currently available course dates.