OOS (Out-of-Specification) - FAQ

General description

1. Why do I need to conduct an investigation if an OOS test result is obtained?
2. What is the purpose of an OOS investigation?
3. Who should investigate OOS?
4. Which parts of the CGMP regulations apply to laboratory operations?
5. Which section of the CGMP regulations specifies that products failing to meet established standards and other relevant quality control criteria will be rejected?
6. How do I handle an OOT result?
7. Can I ignore an OOS investigation if the rejection of a batch is based on an OOS result?
8. Why is it necessary to carry out an OOS investigation if rejection of the batch is based on an OOS result?
9. Which sections of the CGMP regulations require that written records of the investigation are made, including conclusions of the investigation and follow-up?
10. How do I carry out a meaningful OOS investigation?
11. What should the analyst do, whenever possible, before discarding the test solution?
12. When do I need to conduct a complete failure investigation?
13. Who is responsible for the first course of action?
14. Who has the primary responsibility for ensuring accurate laboratory test results?
15. What should the analyst do during the testing process?
16. Which section of the CGMP regulations requires that the analyst should ensure that only those instruments meeting established specifications are used?
17. Which section of the CGMP regulations requires that the analyst should ensure that only properly calibrated instruments are used?
18. What is the usual time requirement during the LAL-endotoxin kinetic determination?
19. What system suitability requirements may be considered?
20. Why could the optical density change calculation method generate OOS results?
21. If signal drift is a problem, then which LAL-endotoxin kinetic curves are most likely to be affected?
22. What should be done if unexpected results are obtained and no obvious explanation exists?
23. Can you list some of the obvious errors that could invalidate results?
24. What should be done in the case of obvious errors?
25. Is it a good idea to continue with an analysis to see what results can be obtained when an obvious error is known, and that the analysis will be invalidated at a later time for an assignable cause?
26. What should the supervisor's responsibilities be in the case of OOS results?
27. What should be included in the re-examination of an immediate assessment?
28. What steps should be taken during the supervisor's assessment?
29. What kind of transient equipment malfunction can occur in HPLC and how can it be resolved?
30. What kind of transient equipment malfunction can occur in LAL-endotoxin kinetic testing and how can it be resolved?
31. How frequently is it reasonable to experience a lab error?
32. What should be done with OOS results if clear evidence of laboratory error exists?
33. What should be done with OOS results when evidence of laboratory error remains unclear?
34. How should I select the procedure for an OOS failure investigation?
35. If varying test results are obtained, and when the evidence of laboratory error remains unclear, then what may be the cause of the OOS problem?
36. Who should conduct an investigation and what departments should be included?
37. What practices are used in the laboratory phase of an OOS investigation?
38. What conditions indicate retesting?
39. Who should do the retesting?
40. Which section of the CGMP regulations requires the establishment of specifications, standards, sampling plans, test procedures, and other laboratory control mechanisms?
41. How many times can I repeat the retests?
42. What should be done if test results obtained after the predetermined number of retests are still unsatisfactory?
43. If I have 2 results, one original OOS result, and one of the retest results which one should be included in the batch?
44. How should I document my retest results?
45. How does resampling differ from retesting?
46. When is resampling of the batch indicated?
47. Based on the above problems, can I use a new sampling method?
48. What determines validity for the averaging of test data?
49. Can you give examples when averaging can be used?
50. Can you give examples when averaging cannot be used?
51. What conditions are required for the use of averaging?
52. Do the CGMP regulations require that the acceptance and/or rejection levels be based on statistically valid quality control criteria?
53. How can an outlier point generate OOS results for LAL-endotoxin testing?
54. What corrective action can you take in case of outliers?

1. Why do I need to conduct an investigation if an OOS test result is obtained?
   An investigation is required by FDA regulations.
2. What is the purpose of an OOS investigation?
   To determine the cause of the OOS result.
3. Who should investigate OOS?
   Both pharmaceutical manufacturing companies and contract laboratories.
4. Which parts of the CGMP regulations apply to laboratory operations?
   Part 211, subparts I (Laboratory Controls) and J (Records and Reports)
5. Which section of the CGMP regulations specifies that products failing to meet established standards and other relevant quality control criteria will be rejected?
   Section 211.165(f).
6. How do I handle an OOT result?
   Results that are out-of-trend (OOT) can be handled similarly to OOS investigations.
7. Can I ignore an OOS investigation if the rejection of a batch is based on an OOS result?
   No, batch rejection does not negate the need to perform the investigation!
8. Why is it necessary to carry out an OOS investigation if the rejection of a batch is based on an OOS result?
   To determine if the result is associated with other batches of the same drug product or other products.
9. Which sections of the CGMP regulations require that written records of the investigation are made, including conclusions of the investigation and follow-up?
   Section 211.192.
10. How do I carry out a meaningful OOS investigation?
    A meaningful OOS investigation should be: thorough, timely, unbiased, well-documented, and scientifically defensible.
11. What should the analyst do, whenever possible, before discarding the test solution?
    An initial assessment of the accuracy of the laboratory's data should be conducted. Why? In this way, any hypothesis inferring laboratory error or instrument malfunction may be tested using the same solutions.
12. When do I need to conduct a complete failure investigation?
    If the initial assessment indicates that no errors were made in the analytical process used to obtain the data, a complete failure investigation should follow.
13. Who is responsible for the first course of action?
    The analyst.
14. Who has the primary responsibility for ensuring accurate laboratory test results?
    The analyst.
15. What should the analyst do during the testing process?
    The analyst should be aware of the potential problems that can occur during the testing process, and should watch for problems that create OOS results.
16. Which section of the CGMP regulations requires that the analyst should ensure that only those instruments meeting established specifications are used?
    211.160(b)(4).
17. Which section of the CGMP regulations requires that the analyst should ensure that only properly calibrated instruments are used?
    211.160(b)(4).
18. What is the usual time requirement during the LAL-endotoxin kinetic determination?
    One hour or longer.
19. What system suitability requirements may be considered?
    Drift, noise, temperature variability, photometric absorbance variability.
20. Why could the optical density change calculation method generate OOS results?
    There is no safe method to avoid signal drift being utilized as LAL-endotoxin reaction signal.
21. If signal drift is a problem, then which LAL-endotoxin kinetic curves are most likely to be affected?
    Turbidimetric determinations at the lowest endotoxin quantification levels.
22. What should be done if unexpected results are obtained and no obvious explanation exists?
    Test preparations should be retained, the analyst should inform the supervisor, and an accuracy assessment should be started immediately.
23. Can you list some of the obvious errors that could invalidate results?
    Spilling of a sample, incomplete transfer of sample, obvious pipetting errors, bubble formation in microplate wells, dirty lab equipment.
24. What should be done in the case of obvious errors?
    The analyst should immediately document what happened.
25. Is it a good idea to continue with an analysis to see what results can be obtained when an obvious error is known, and that the analysis will be invalidated at a later time for an assignable cause?
    No, analysts should not knowingly continue any analysis they expect to invalidate.
26. What should the supervisor's responsibilities be in the case of OOS results?
    A supervisor's assessment should be objective and timely, and free from any preconceived assumptions as to the cause of OOS results. Data should be assessed promptly to ascertain if the results may be attributed to laboratory error, or whether the results could indicate problems in the manufacturing process.
27. What should be included in the re-examination of an immediate assessment?
    Re-examination of the actual solutions, test units, glassware, and lab equipment used in the original measurements. This will allow more credibility to be given to laboratory error theories.
28. What steps should be taken during the supervisor's assessment?
    (1) Review training results of the analyst to determine the analyst's competency to carry out the analysis, (2) Review the test procedure and associated SOPs for correctness, (3) Discuss the test method with the analyst to confirm analyst knowledge of, and performance of, the correct procedure, (4) Examine raw data obtained in the analysis, including all of the analytical (measurement) signals generated during analysis, such as chromatograms and spectra, and identify anomalous or suspect information, (5) Confirm the performance of the instruments (PQ), (6) Determine that appropriate reference standards, solvents, reagents, and other solutions were used and they meet quality control specifications, (7) Evaluate the performance of the testing method to ensure that it is performing to the expected standard, based on method validation data, (8) Document and preserve evidence of this assessment.
29. What kind of transient equipment malfunction can occur in HPLC and how can it be resolved?
    An auto-injector could inject bubbles during an operation; therefore, repeated re-injections will provide strong evidence that the problem can be attributed to the instrument, rather than the sample or its preparation.
30. What kind of transient equipment malfunction can occur in LAL-endotoxin kinetic testing and how can it be resolved?
    Photometric measurement failure could generate a radically different signal, which can radically distort the OD limits, which in turn would provide a false reaction time for the endotoxin calculation; consequently, removal of the outlier OD reading can provide strong evidence that the problem should be attributed to the instrument, rather than the sample or its preparation.
31. How frequently is it reasonable to experience a lab error?
    Laboratory error should be relatively rare. Frequent laboratory errors could mean one or more of the following: (1) inadequate training of the analysts, (2) poorly maintained equipment, (3) improperly calibrated equipment, (4) careless work.
32. What should be done with OOS results if clear evidence of laboratory error exists?
    The OOS test results should be invalidated.
33. What should be done with OOS results when evidence of laboratory error remains unclear?
    A failure investigation should be conducted to determine what caused the unexpected result.
34. How should I select the procedure for an OOS failure investigation?
    A predefined procedure should be applied.
35. If varying test results are obtained, and when the evidence of laboratory error remains unclear, then what may be the cause of the OOS problem?
    The OOS problem may be due to either the manufacturing process or the sampling process.
36. Who should conduct an investigation and what departments should be included?
    The QC unit should conduct the investigation, which should involve all other departments that could be implicated.
37. What practices are used in the laboratory phase of an OOS investigation?
    (1) Retesting a portion of the original sample, (2) testing a specimen from the collection of a new sample from the batch, (3) resampling test data, (4) using outlier testing.
38. What conditions indicate retesting?
    Situations where retesting is indicated include investigating testing instrument malfunction or identifying possible sample handling integrity problems, for example, a suspected dilution error.
39. Who should do the retesting?
    An analyst other than the one who performed the original test should perform retesting.
40. Which section of the CGMP regulations requires the establishment of specifications, standards, sampling plans, test procedures, and other laboratory control mechanisms?
    Section 211.160. (Note: USP 23, under General Notes and Requirements, p. 9, states: testing must be in accordance with "predetermined guidelines or sampling strategies.")
41. How many times can I repeat the retests?
    The number of retests to be performed on a sample should be based on company specifications contained in the relevant SOP.
42. What should be done if test results obtained after the predetermined number of retests are still unsatisfactory?
    If, at this point, the results are unsatisfactory, the batch is suspect and must be rejected or held pending further investigation (§211.165(f)).
43. If I have 2 results, one original OOS result, and one of the retest results which one should be included in the batch?
    In the case of a clearly identified laboratory error, the retest results would substitute for the original test results. If no laboratory or statistical errors are identified in the first test, then both test results should be reported and considered in batch release decisions.
44. How should I document my retest results?
    In all cases the original test result should be retained, an explanation recorded. The record should be initialed and dated by the persons involved, and include a discussion of the error with supervisory comments.
45. How does resampling differ from retesting?
    While retesting refers to analysis of the original sample, resampling involves analyzing a specimen from the collection of a new sample from the batch. (Note: testing should be done in accordance with predetermined procedures and sampling strategies (§211.165(c).)
46. When is resampling of the batch indicated?
    A resampling of the batch should be conducted if the investigation shows that the original sample was not representative of the batch. (Note: in some cases, when all data have been examined (e.g., widely varied results obtained from several aliquots of the original composite/after determining there was no performance error made in the analysis), it may be concluded that the original sample was improperly prepared and was therefore not representative of the batch (§211.160(b)(3)).
47. Based on the above problems, can I use a new sampling method?
    No, resampling should be performed by the same qualified, validated methods that were used for the initial sample. (Note: if the investigation determines that the sampling method was in error, a new sampling method must be developed, qualified, and documented (§211.160 and 165(c)).
48. What determines validity for the averaging of test data?
    The validity of averaging depends upon the sample and its purpose. (Note: if the sample can be assumed to be homogeneous, i.e., an individual sample preparation designed to be homogeneous.) Using averages can provide more accurate results. For example, in the case of microbiological assays, the USP prefers the use of averages because of the innate variability of the microbiological test system.
49. Can you give examples when averaging can be used?
    Kinetic scan of individual wells, or endotoxin data from a number of consecutive measurements (note: the determination is considered one test and one result), however, unexpected variation in replicate determinations should trigger investigation and documentation requirements (21 CFR 211.192).
50. Can you give examples when 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.
51. What conditions are required for the use of averaging?
    (1) All test results should confirm to specifications (Note: a batch must be formulated with the intent to provide not less than 100 percent of the labeled or established amount of the active ingredient (21 CFR 211.101(a)), (2) Averaging must be specified by the test method.
52. Do the CGMP regulations require that the acceptance and/or rejection levels be based on statistically valid quality control criteria?
    Yes, according to Section 211.165(d) of the CGMP, statistically valid quality control criteria should be used for appropriate acceptance and/or rejection levels.
53. How can an outlier point generate OOS results for LAL-endotoxin testing?
    If OD limit calculations are used, an outlier point can generate an OD limit at the point of instrumental malfunction or noise generated.
54. What corrective action can you take in case of outliers?
    Instrument malfunctions can occur due to power line surge, sudden lamp failure, sudden detector failure, high frequency transient noise, power failure.