MTS Helpdesk

Question:

What results should be reported on the Certificate of Analysis for an inconclusive OOS phase 2 investigation where the retest results are acceptable?

Answer:

This question relates to the process of investigation, as defined by guidance from the UK MHRA and the US FDA, which is followed when out of specification (OOS) results are obtained during QC testing of pharmaceuticals. The investigation consists of several phases. In phase 1, a laboratory investigation is performed, the aim of which is to try to identify whether the OOS result is due to an error in the testing.

If no assignable cause is identified in the phase 1 laboratory investigation, then a phase 2 investigation is commenced. This consists of an investigation into the manufacturing process to identify whether there could be a reason for the OOS there, and also additional laboratory testing may be performed. The purpose of the additional testing is to retest the material at a statistically relevant level of replication to provide a better estimate for the true value of the test material and thus aid in decision making. The number of replicates in additional testing is typically 5, 6 or 7.

If the results from the additional testing are also OOS, then this would imply that the test material does not meet the required specification and the OOS is confirmed. The mean of all of the results, including the original testing result, is calculated and is the reportable value for the test material.

There are three possible outcomes where the additional testing results are in specification:

The original OOS result is an outlier when compared to the retest results and the assignable cause is known.

In this case, the cause of the original OOS result is known, and therefore the OOS result can be invalidated. The reportable value for the test material is the average of the additional testing results. It is worth noting that the assignable cause needs to be proven by the additional testing. In most cases where this outcome occurs, evidence of the assignable cause is identified in the phase 1 laboratory investigation, but it is not possible to prove it there without generation of retesting data.

The original OOS result is an outlier when compared to the retest results and the assignable cause is unknown.

The fact that the original OOS result is an outlier when compared to the retest results implies that it is not part of the same population and therefore there is a clear difference between the two sets of data. It seems most likely that some unidentified issue occurred during the original testing but without proof of an assignable cause it is not possible to invalidate the original OOS result. The reportable values for the test material are both the OOS result and the retest results, the mean of the retests represents the most probable quality of the test material.

The original result is not an outlier when compared to the retest results and the assignable cause is unknown.

If the original OOS is not an outlier when compared to the additional testing results, it implies that they are all part of the same population, and that there is no reason to differentiate between the OOS result and the retests. The attribute of the test material that is being tested is close to the specification limit. There is no justification to exclude the original OOS result and thus the reportable value is the mean of all of the test results, including the original OOS result. To fully assess whether the test material is in specification, it is recommended that the 95% confidence interval of the mean is calculated. If the full confidence interval is within the specification limits, then the test material is in specification and potentially releasable, but probably out of trend (OOT).

In all of these outcomes, the retesting results are in specification and therefore theoretically acceptable but the action in terms of what is the reportable value and therefore what should be reported on the Certificate of Analysis is different.

The outcome described above where the original OOS result is an outlier when compared to the retest results and the assignable cause is unknown, is an inconclusive investigation and it is clear in both guidance documents that it is expected that all results should be reported on the certificate of analysis:

MHRA:

‘If no laboratory or calculation errors are identified in the Phase I and Phase II there is no scientific basis for invalidating initial OOS results in favour of passing retest results. All test results, both passing and suspect, should be reported (in all QC documents and any Certificates of Analysis) and all data has to be considered in batch release decisions.’

FDA:

‘If no laboratory or calculation errors are identified in the first test, there is no scientific basis for invalidating initial OOS results in favor of passing retest results. All test results, both passing and suspect, should be reported’¹⁰ and considered in batch release decisions.

¹⁰In other words, all data are reported in, for example, quality control reports, batch records, Certificates of Analysis, in accordance with §§ 211.188 and 211.192.’

In practice, it is often difficult to include both the OOS original result and the multiple individual additional testing results on the Certificate of Analysis. Although it is a regulatory expectation to do so, many companies take the decision to only include the mean of the retests as the reportable value and accept the compliance risk.

References

US FDA Guidance for Industry, ‘Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production’, Revision 1, May 2022

UK MHRA, ‘Out of Specification & Out of Trend Investigations’, Oct 2017

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.