Explanation of Repeat Testing and Retesting Utilized During Microbiological Out-of-Specification Laboratory Investigations
Introduction
Repeat testing and Retesting can be two separate events that may occur during a microbiological Out-of-Specification (OOS) laboratory investigation.
During a laboratory investigation, confusion may arise as to the difference between these terms and their overall purpose. This document provides a more detailed explanation of these differences and the individual importance of these tools during a microbiological OOS laboratory investigation. In addition, this guidance will also briefly clarify similarities and differences of these definitions as compared to analytical OOS laboratory investigations.
Microbiological repeat testing may be defined as additional laboratory testing performed to replace original invalid data when a laboratory assignable cause has been identified.
Microbiological retesting may be defined as additional laboratory testing performed to corroborate the original OOS test result when a laboratory assignable cause has not
been established. This guidance clarifies these definitions, explains the intent of this testing, and contrasts these definitions to how they are utilized for analytical OOS laboratory investigations.
Recommendations & Rationale for Recommendations
Repeat Testing
Microbiological repeat testing is only to be performed when the OOS investigation has invalidated the original test result by the establishment of an assignable cause not related to the quality of the sample being tested. An assignable cause is identified through a laboratory investigation.
The repeat test result is intended to replace the original invalidated OOS test result. The original test result and the repeat test result need to be reported in the Laboratory Investigation Report (LIR) with a descriptive rationale that details the conditions that allow this result replacement.
The repeat/retest protocol portion of the LIR shall be approved by lab supervision prior to any repeat testing and this portion shall be used to document the repeat testing and the results obtained.
In the case of sterility testing, if an assignable cause associated with laboratory error is identified, an invalid sterility test may only be repeated once.
Microbiological repeat testing is very similar to repeat testing conducted during analytical laboratory investigations. If an assignable cause is clearly identified, the initial OOS test result is invalidated and the original testing is repeated to generate a valid result. Unlike microbiological testing, some analytical test methods consist of multiple analyses per test sample. In such cases, a single repeat test of only one analysis may be acceptable to replace a single initial OOS result. It is important to recognize that this option is not available in microbiological repeat testing because multiple analyses per test sample are not performed.
Retesting
Microbiological retesting is an investigational tool that may be used as part of the Investigational Measurements Protocol. The number of retests to be performed on a sample should be specified in advance and should prescribe a point when to end testing and begin evaluating the product/material.
According to the FDA OOS Draft Guidance, repeat testing until a passing result is obtained (i.e., testing into compliance) is objectionable under the cGMPs.
If the compendia do not define retest criteria, the decision to retest shall be made by Q.A. management and must be based on a sound scientific rationale. A retesting protocol should be created and approved by laboratory supervision prior to any retesting.
The retest protocol must be based on the specific problem identified, the history of the product, the method, the batch/lot used, and any applicable compendial requirements.
The retest protocol must also delineate the number of retests to be performed. The number of retests to be performed may vary, but this number should be based upon sound scientific reasoning. When sufficient sample is available, retesting must be executed using the same sample set that was the source of the original OOS test result, unless there is scientific rationale for not using this sample.
The retest plan must also include a control lot to be used to verify the accuracy of the analyses and the acceptance criteria for the control lot must be specified.
Microbial limit retests may use up to a maximum amount of 25 grams for a solid sample or 25 ml for a liquid sample.
The value of retesting during a microbiological OOS investigation is fairly limited in scope. Retesting may be performed only to corroborate or confirm the original OOS test result. If the retest results do confirm the initial OOS test result, this data can be used to support the case that the initial OOS test result is valid and not due to laboratory contamination. A confirmed OOS test result will cause the rejection of the test article (unless approved for reprocessing).
If the retest results do not confirm the initial OOS test result, the data can be used to only help support the claim that the initial OOS test result is invalid and attributable to laboratory contamination. Retest results alone cannot be used to invalidate the initial OOS test result. Only a clearly identified assignable cause can invalidate the initial OOS test result.
The FDA OOS Draft Guidance also states that if no laboratory errors are identified in the initial test, then there is no scientific basis for invalidating the initial OOS test result in favour of the passing retest result.
Microbiological retesting is significantly different to the retesting conducted during analytical method laboratory investigations.
First, analytical method retesting requires a minimum of five retests for formulated products (e.g., in process and finished product samples) and a minimum of three retests for other samples types (e.g., raw materials, API).
Second, analytical laboratory investigation retest results can be utilized to overcome the initial OOS test result after evaluation by Q.A. management. For example, if the retest results for a raw material, intermediate, or API sample are within specification, then the mean of the retest results shall be reported as the final valid result.
Furthermore, for formulated products, if the retest results are within specification and the original OOS result is outside three standard deviations of the mean of the retest results, then the average of the retest results shall be reported as the final valid result.
Conversely, this type of approach is not applicable for microbiological OOS laboratory investigations. Microbiological retest results are only used to corroborate or confirm the initial OOS test result. This is because the microbial population present in the test sample may not be uniformly distributed or the microbial population within the test sample may die off over an extended period of time (i.e., between when testing and retesting occurs).
In summary, repeat testing, for both microbiological and analytical method OOS laboratory investigations, is only to be performed when the OOS investigation has invalidated the original test result by the establishment of an assignable cause. The repeat test result is intended to replace the original invalidated OOS test result.
Alternatively, retesting is used primarily as an investigational tool if no readily apparent assignable cause has been determined during the initial investigation. Microbiological OOS retesting differs significantly from analytical method OOS retesting with regards to testing requirements and data interpretation.