Out Of Specification Investigation Phase II (MHRA)
Conducted when the phase I investigations did not reveal an assignable laboratory error. Phase II investigations are driven by written and approved instructions against hypothesis. Prior to further testing a manufacturing investigation should be started to determine whether there was a possible manufacturing root cause.
If not already notified the contract giver/MAH/QP (in accordance with the responsibilities in the TA) should be notified along with production and QA if a manufacturing site.
It is important when considering performing additional testing that it is performed using a predefined retesting plan to include retests performed by an analyst other than the one who performed the original test. A second analyst performing a retest should be at least as experienced and qualified in the method as the original analyst.
If the investigation determines analyst error all analysis using the same technique performed by the concerned analyst should be reviewed.
Phase II Flow Chart
Hypothesis/Investigative Testing –
Is testing performed to help confirm or discount a possible root cause i.e what might have happened that can be tested:- for example it may include further testing regarding sample filtration, sonication /extraction; and potential equipment failures etc. Multiple hypothesis can be explored.
Re-Test –
Performing the test over again using material from the original sample composite, if it has not been compromised and/or is still available. If not, a new sample will be used.
Re-sample –
A new sample from the original container where possible, required in the event of insufficient material remaining from original sample composite or proven issue with original sample integrity.
Most probable cause –
Scientifically justified determination that the result appears to be laboratory error.
Phase II Investigation – Unknown Cause / No Assignable Cause
Hypothesis Testing (Applicable to Phase Ia and Phase II):
Should be started as part of Phase Ia and continue into Phase II if no assignable cause found.
Description of the testing should be written, and then approved by QA/Contract Giver/QA equivalent prior to initiating investigational testing. The requirements of investigational testing listed below:
The description must fully document
•The hypothesis to the test the root cause being investigated.
•What samples will be tested.
•The exact execution of the testing.
•How the data will be evaluated
This Hypothesis testing may continue from the re-measurement of the original preparations.
Investigational testing may not be used to replace an original suspect analytical results. It may only be used to confirm or discount a probable cause.
If no assignable cause that could explain the results can be identified during the manufacturing investigation or the assay failure investigation retesting may be considered. Part of the investigation may involve retesting a portion of the original sample.
Retesting:
•Performed on the original sample not a different sample.
•Can be a 2nd aliquot from the same sample that was the source of the original failure.
•If insufficient quantity of the original sample remains to perform all further testing then the procedure for obtaining a resample must be discussed and agreed by QA/Contract Giver/QA equivalent. The process of obtaining the resample should be recorded within the laboratory investigation.
•The decision to retest should be based on sound scientific judgement. The test plan must be approved before re testing occurs. The minimum number of retests should be documented within the procedure and be based upon scientifically sound principles. Any statistical review with regards to %RSD and repeatability should relate to the values obtained during method validation (accuracy, precision, and intermediate precision). The number of retests should be statistically valid; papers have suggested 5, 7, or 9.
•The retests should be performed by a different analyst where possible. The second analyst should be at least as experienced and qualified in the method as the original analyst.
Averaging:
•The validity of averaging depends upon the sample and its purpose. Using averages can provide more accurate results. For example, in the case of microbiological assays, the use of averages because of the innate variability of the microbiological test system. The kinetic scan of individual wells, or endotoxin data from a number of consecutive measurements, or with HPLC consecutive replicate injections from the same preparation (the determination is considered one test and one result), however, unexpected variation in replicate determinations should trigger investigation and documentation requirements.
•Averaging cannot be used in cases when testing is intended to measure variability within the product, such as powder blend/mixture uniformity or dosage form content uniformity.
•Reliance on averaging has the disadvantage of hiding variability among individual test results. For this reason, all individual test results should normally be reported as separate values. Where averaging of separate tests is appropriately specified by the test method, a single averaged result can be reported as the final test result. In some cases, a statistical treatment of the variability of results is reported. For example, in a test for dosage form content uniformity, the standard deviation (or relative standard deviation) is reported with the individual unit dose test results.
•In the context of additional testing performed during an OOS investigation, averaging the result (s) of the original test that prompted the investigation and additional retest or resample results obtained during the OOS investigation is not appropriate because it hides variability among the individual results. Relying on averages of such data can be particularly misleading when some of the results are OOS and others are within specifications. It is critical that the laboratory provide all individual results for evaluation and consideration by Quality Assurance (Contract Giver/QP).
•All test results should conform to specifications (Note: a batch must be formulated with the intent to provide not less than 100 percent of the labelled or established amount of the active ingredient.
•Averaging must be specified by the test method.
•Consideration of the 95% Confidence Limits (CL 95% ) of the mean would show the variability when averaging is used.
Averaging continued: Consideration of using 95% Confidence Limits (CL 95% ) of the mean would show the variability when averaging is used.
The confidence interval is calculated from the formula:
CL= sample mean ± t 95% sample standard deviation/ √ n
–Where t is a value obtained from tables
–Where n is the sample size
–Table:
| n | t 95% |
| 2 | 12.71 |
| 3 | 4.30 |
| 4 | 3.18 |
| 5 | 2.78 |
| 6 | 2.57 |
| 7 | 2.45 |
| 10 | 2.26 |
| 20 | 2.09 |
| 120 | 1.98 |
| ∞ | 1.96 |
Re-sampling:
Should rarely occur, If insufficient quantity of the original sample remains to perform all further testing then the procedure for obtaining a resample must be discussed and agreed by QA/Contract Giver/QA equivalent. The process of obtaining the resample should be recorded within the laboratory investigation. Re-sampling should be performed by the same qualified methods that were used for the initial sample. However, if the investigation determines that the initial sampling method was in error, a new accurate sampling method shall be developed, qualified and documented. It involves the collecting a new sample from the batch. Will occur when the original sample was not truly representative of the batch or there was a documented/traceable lab error in its preparation. Evidence indicates that the sample is compromised or invalid. Sound scientific justification must be employed if re-sampling is to occur.
Outlier test:
An outlier may result from a deviation from prescribed test methods, or it may be the result of variability in the sample. It should never be assumed that the reason for an outlier is error in the testing procedure, rather than inherent variability in the sample being tested. Statistical analysis for Outlier test results can be as part of the investigation and analysis. However for validated chemical tests with relatively small variance and that the sample was considered homogeneous it cannot be used to justify the rejection of data.
• While OOS guidance is not directly intended for bioassay analysis, it can be used as a starting point for the investigation. Compendia such as the BP; PhEur and USP, provide guidance on outliers for these types of analysis