VERIFICATION OF INITIAL OUT-OF-SPECIFICATION RESULTS
SCOPE
The objective of this Annex is to describe the steps to follow for the management of out-of-specification (OOS) testing results. The failure investigation may also be triggered if validation criteria defined by the method or laboratory are not met, including out-of-trend (OOT) results. To simplify the text, only OOS is used.
Management of OOS results
The management of OOS results includes:
a. Documented investigation and approval on scientific basis of the decisions taken regarding the acceptance or not of a result.
b. Determination of the cause of the OOS results.
c. Definition of the related corrections and corrective actions.
This Annex presents one approach; other approaches with different steps may be acceptable, as long as they are predefined, scientifically justified and documented.
RESPONSABILITIES
This Annex applies to results from physicochemical, microbiological and biological tests generated against specification limits given in compendial texts (e.g. Ph. Eur.) or provided by the manufacturer of the tested product.
PROCEDURE
The purpose of an OOS investigation is to determine the cause of the OOS result which could be related to the analytical laboratory, other causes or the product itself.
It must be noted that the initial result should not to be disregarded until the investigation is concluded and the assigned cause identified.
An investigation shall be thorough, timely, impartial, detailed, documented and based on scientifically sound elements, possibly considering trends in the evaluation of the causes of OOS/OOT results.
Examples of how to conduct failure investigation are given in Appendix 1 of this Annex.
Examples for managing quantitative and qualitative retesting are given in Annexes 3.2 and 3.3, respectively.
VERIFICATION OF OUT-OF SPECIFICATION RESULTS IN QUANTITATIVE TESTING
The investigation includes two phases:
Phase I: the objective is to identify an assigned cause linked to a laboratory issue or not. If no assigned cause is found, the investigation must continue with Phase II.
Phase II: describes how results from initial and repeat testing can be evaluated in order to reach a consistent conclusion on the sample.
Phase I: Identification of the assigned cause
The aim of Phase I is to investigate possible sources of error(s) in the testing process. This evaluation is carried out collaboratively by the analyst and the supervisor via a documented investigation of the possible sources of laboratory error.
The initial result is invalidated if the assigned cause is identified. Otherwise, the investigation should be continued to verify hypotheses of a possible rootcause.
Sample and standard preparations shall not be discarded during this phase, in order to use the same preparations to test hypotheses regarding laboratory error or instrument malfunction.
Phase II: Verification of OOS by retesting
If no laboratory-assigned cause is identified in Phase I of the OOS investigation, the result is confirmed by retesting.
Phase Il describes the steps to reach a conclusion about compliance of the sample, considering all valid results obtained.
The objective of Phase Il is to verify the initial OOS by retesting.
The retest is conducted as a repetition of the initial analysis (i.e. number of independent determinations according to the test method) on the initial sample (except in justified cases, e.g. resampling).
Following the principles described in Annex 2, a maximum of 6 independent determinations should be performed.
The initial result is the one that triggered the OOS investigation. The relative standard deviation (RSD) of initial and retest results (considering the values of all independent determinations) should be calculated and the consistency of the results should be checked by comparing with the corresponding T2 values in Annex 2.
If the reportable results are consistent, Phase Il can be closed and the results reported as the mean of the initial and retest results.
Where the reportable results are not consistent, the lack of consistency should be investigated using Annex 3.1 to address possible causes of inconsistency.
If a cause has been found for inconsistency, 3 scenarios are possible:
- Both initial and retesting results are affected; in such cases both results are considered not valid.
- Only the results generated in the initial testing are affected, and are therefore to be considered not valid.
- If only the results generated in Phase Il are affected (retesting) this result is not considered valid.
Reference: Evaluation & Reporting of Results – Annex 3.2: Verification of OOS results in quantitative testing
VERIFICATION OF OUT-OF-SPECIFICATION RESULTS IN QUALITATIVE TESTING
This document is an Annex of the core document “Evaluation and reporting of results”, PA/PH/OMCL (13) 113 , in its current version, and it should be used in combination with it when planning, performing and documenting the assessment of qualitative test results which are suspected to be OOS, as well as reporting these same results.
When a suspected OOS test result indicates that a sample is non-conforming, i.e. it does not meet regulatory specifications, it is necessary to decide whether a retest of that sample is appropriate.
Special care should be taken with the documentation of qualitative OOS results that are visually detected (e.g. TLC, pharmacognosy tests, appearance of the pharmaceutical dosage form).
Photographic records or documented confirmation by a second qualified analyst should be
considered.
Focus of failure investigation
The same basic principles used for failure investigation of quantitative testing results are applicable, which means that if an OOS result is due to an assignable cause (e.g. sample integrity problem, or to a traceable laboratory or operator error), then the results of the initial test may be invalidated and need not be taken into consideration in the final results.
If an OOS result is not due to a sample integrity problem, or to a laboratory or operator error, a supervisor with requisite expertise should ascertain whether the OOS result is a valid test result for reporting purposes, or whether it requires further verification via retesting.
Moreover, as this process covers qualitative tests, for which the investigation is, in general, simpler, there is no need to distinguish between Phase I and Phase II during the OOS investigation, as established for quantitative tests (Annex 3.2).
For qualitative tests, the focus of failure investigation should be further adapted, in order to help
to clarify the following situations:
In cases of non-detection of specified analytes, the focus should be on:
- influencing factors for possible analyte degradation (e.g. sample storage)
- confirmation of sample addition
- possible loss during extraction/sample preparation
- detection mechanisms of the equipment used
In cases of detection of unspecified analytes, the focus should be on possible cross-contamination or carry-over from other samples, reference standards, surfaces, glassware or
equipment.
The purpose of this stage is to determine whether the reportable result initially obtained is considered to be valid for reporting purposes, or it requires verification/confirmation via retesting.
Retest result should not be used to ignore or overrule an initial OOS result if it does not confirm the initial OOS result, unless scientifically justified.
Reference: Evaluation & Reporting of Results – Annex 3.3: Verification of OOS results in qualitative testing
Read also: