710 likes | 2.63k Views
OOS (Out of Specification). The failure of a batch or any of its components to meet any of its specificationsOOS indicates non-compliance with specifications. Outliers. Statistically invalid individual test resultCan be eliminated. OOS ? General Principles. How should we handle OOS results?OO
E N D
1. Out of Specification (OOS) Test Results ????? ?????
? ??
2003? 8? 25?
2. OOS (Out of Specification) The failure of a batch or any of its components to meet any of its specifications
OOS indicates non-compliance with specifications
3. Outliers Statistically invalid individual test result
Can be eliminated
4. OOS – General Principles How should we handle OOS results?
OOS cannot be ignored or discounted without basis
When does testing stop (accept the result)
First test failure?
Second test failure? Tenth test failure?
5. Barr Decision (1993) cGMP regulations are vague about exactly how to handle OOS
1993? Barr Laboratories, Inc
United States District Court for the District of New Jersey
Jurdge Wolin made it clear
what is acceptable and
what is not acceptable in handling OOS
6. What does the cGMP Regulation say about Handling OOS Results?
7. cGMP Part 211.192 The failure of a batch or any of its components to meet any of its specifications shall be thoroughly investigated…
The investigation shall extend to other batches of the same drug product and other drug products that may have been associated with the specific failure of discrepancy…
8. cGMP Part 211.194 A written record of the investigation shall include the conclusions and follow-up…
A complete record of all data secured in the course of each test…
Complete records shall be maintained of all stability testing performed…
9. Barr Decision Failure investigation is central
Failure investigation
Immediate
Timely follow-up
Appropriate resolution
Documented
10. FDA: Interpretation of investigation results Batch acceptance or rejection must be based on scientific justification
Overall body of data must be considered in judging the batch
11. Essence of Barr Labs Decision Include all suspect results that fall outside established specifications
Drug testing
Drug companies must
Validate their processes and
Ensure the quality of the batch for release
Out-of-Specification
Better word than “failure” for non-passing results
12. Essence of Barr Labs Decision (continued) Failure investigation: Elements include
Reason for investigation
Summation of process sequences
List of other batches and results
All comments and signatures
13. Key Points: Essence of Barr Labs Decision Release of batch
Testing must show satisfactory conformance to specifications, including ID/Strength of each active ingredient
Context and history of product and batches inform the final conclusion
14. Averaging Assay results should never be averaged because averaging hides individual variability:
e.g. 89, 89, 92 (x=90)
Individual content uniformity tests should not be averaged to obtain passing value
Microbiology averaging is acceptable due to biological variability
15. Averaging (continued) Relying on the average of OOS results and in-spec results is usually misleading and unacceptable
Should not be used to hide variation in individual test results
As a general rule, avoid averaging
Preferred only when it is originally designed as part of test
16. Averaging of Results: FDA Caution with product release at either end of limits
17. Avoid “Testing into Compliance” Retesting: Additional test should be for the same sample, only after a failure investigation is underway
“Testing into compliance” is not acceptable
Retest only by predetermined written SOP
18. Retesting In accordance with SOPs
Re-test done by a second analyst
Number of retests predetermined
Results substitute for original
Investigation ? Remove all bias ? Retest
If the retest is pass, it is pass.
19. Resampling: Controversial Cannot be relied upon when testing and retesting have failed
Appropriate for USP content uniformity and dissolution testing and for limited circumstances when initial sample is unrepresentative
Never do resampling
20. Resampling Never do resampling
Justification for resampling
Not representative of batch
Sample preparation error
Sample consumed in analysis
Only for homogeneous material
21. Reprocessing/Reworking Reprocessing
Repeat of the validated process
Reworking
Repeat of the unvalidated process
22. Reprocessing/Reworking (continued) cGMP is vague in allowing reprocessing and reworking
cGMP allows reworking if the batches conform to all established standards, specifications and characteristics at CFR211.115
23. OOS Investigations: FDA Expectations Written SOPs
Responsibility
Methodology – Handling of OOS
Results
Documentation
24. Is the First Result Real? Are the first results true and accurate?
Validated methods?
Changes in validated methods?
Qualified and calibrated laboratory equipment?
Correct instrumentation parameters?
Correct sample and standard weights?
Reference standards dried properly?
25. Lab Error Evaluation System suitability before and during run?
Incomplete
Dissolution of active?
Dilution, shaking, extraction times?
Proper
Glassware/solution storage?
Trained personnel?
Documentation problems?
Transcription errors?
26. Qualification Installation Qualification (IQ)
Operational Qualification (OQ)
Performance Qualification (PQ)
Maintenance Procedures (MP)
27. Installation Qualification (IQ) Verify receipt of software and hardware product
Identify all component of software and hardware product
Connect
Fluid
Electrical power
Communications
Document
28. Operational Qualification (OQ) Verify equipment performance (each component, in case of HPLC)
Pump
Flow rate accuracy test
Gradient proportioning valve test
Auto sampler
Injection accuracy test
Heater/cooler test
Column heater test
29. Operational Qualification (OQ, continued) Detector
Wavelength accuracy test
Linearity test
Software
Internal or external standard and unknown samples
Actual vs. expected output testing
30. Performance Qualification (PQ) Documented verification of system (hardware & software) performance
Verify total system control integrity
Analytical system
Reproducibility test
Peak area
Peak height
Retention time
Injection linearity test (5, 10, 20, 50 ?l)
31. Maintenance Procedures (MP) Periodic procedures
To minimize the risk of losing raw data and analytical results
Inspection/replacement of normal wear and maintenance items
File back-up and recovery
Data archival and retrieval
Security
Lan administration
32. System Suitability Test H/W, S/W and analysis checks to provide assurance of system integrity
Before, during and following analysis of unknown samples
System precision
Separation parameters
System suitability testing alone is not sufficient, qualification is still required
33. Precision (Repeatability) Retention time (RT)
Area
Height
% RSD ? 1% for N ? 5
34. Separation Parameters Tailing factor (T)
T = W/2f
W = width of the peak determined at 5% from the baseline of the peak height
f = Distance between peak maximum and peak front at W
T ? 2
35. Separation Parameters (continued) Theoretical plate number (N)
N = 16 (tR/tW)2
tR: Retention time of the analyte
tW: Peak width measured at the baseline
36. Analytical Methods Validation or verification
Finished dose form
API
Intermediates
Ruggedness and robustness
Transferability
Person-to-person
Equipment-to-equipment
Site-to-site
37. Chromatography Chromatography
HPLC
GC
TLC
Inspection
Column (Medium)
Mobile phase (pH, buffer, etc.)
Expiration date (change with time)
Detector (wavelength)
Flow rate
Injection volume
Cleaning and regeneration
38. Chromatography Inspection Identification of equipment and materials
Status
Reagent mfg by
Mfg date & expiration date
Cautionary statements on label
Expiration date
RH (relative humidity)
Light
Temperature
Cleaning:
Lab, table, equipment
39. Chromatography: Impurities 0.1% individual impurities
Identify known impurities
1% total
Impurity profile
40. HPLC Consistency Peak areas
Retention times
Unusual peak shapes
Poor chromatography
Unexplained peaks
Shoulders
Poor separation
Qualification/System suitability
41. UV/Vis Spectrophotometers No filters (UV region 200 to 380 nm)
Internal calibration only
Old-outdated standard spectra
Generally poor quality spectra
No baseline
Point readings
Scanning
42. IR Spectrophotometers Reference standard missing
Polystyrene film
Old standard spectra library
Poor quality spectra
Follow SOP for “fingerprint” scan
Poor peak comparison for fingerprint region
Fast vs slow scan speed (follow specification)
Varying scan speeds - problem
43. Balances Equipment not identified
Not checked or calibrated each day of use
Linearity test
Zero plus 2 weight points (3 points)
Weights missing
No periodic checks by manufacturer
Faulty room design
Air flow directly onto balance
Table stability
44. pH Meters Buffer
No in-house manufacturing record for buffer
One point calibration?
Linearity
Lot number missing
Temperature compensating check?
45. Stability Test Temperature recording
Controlled temperature vs monitored
Actual product packaging
Method validation – missing?
Stability indicating methods
Impurities
Protocol discrepancies
Missing time points
Late time points
46. OOS Investigations: FDA Focus Pre-established protocol for investigating OOS results
Summary of specific investigation steps:
Evaluation of other batches
Conclusion and follow-up
Action necessary to prevent similar recurrences
47. OOS Investigations: FDA Focus (continued) Timely investigation must be performed within 30 business days
Investigate before retest
Well documented
Scientifically sound
48. Investigations If the investigation has an attributable cause, then corrective action must be instituted
Cause: Incomplete capsule dissolution
Corrective action: “A 15 minute sonication step will be added to the assay prep procedure”
49. Expectation of FDA OOS results will be generated
Comprehensive, honest approach to investigation of OOS results
Evaluation using scientifically valid principles
Learn from the experience
Permanent solution to the problem
50. Documentation Document reasons for investigation
Report the set of events that may have caused the problem
Report results (actual or probable cause)
Review of other batches and products
Document corrective action
51. Warning Letters In FY 1997, about 30% of GMP Warning Letters sent to firms referenced deficient OOS failure investigations…
Failure to conduct an investigation
Inadequate follow-up
52. Inspectional Observations on FDA 483s “Laboratory Investigations for OOS Results”, SOP# xxx does not specify the number of retests which can be performed.
For lot# xxxx, the product was retested 6 times by 3 different analysts
53. Inspectional Observations on FDA 483s (continued) 22 out of 37 failure investigation reports reviewed did not include corrective actions.
There is no procedure in place to assure that corrective actions are actually made
For report # xxxx, 14 months had elapsed before the corrective action was actually implemented
54. Inspectional Observations on FDA 483s (continued) The laboratory does not have procedures in place for tracking and trending laboratory investigations.
The investigations do not include
The corrective actions necessary, nor do they include a review of batches with similar OOS results, or other products affected
55. Inspectional Observations on FDA 483s (continued) The firm does not conduct failure investigations for every product failure
During the failure investigation, it was determined that the analytical method used was not stability indicating, and therefore the analytical results were not valid.
No follow-up, changes or corrective actions were instituted regarding the analytical method.
56. FDA Out of Specifications Guidance Draft issued in September 1998