Contents Objective Definitions and terminology General investigation principles Investigation stages Initial Laboratory Investigation FullScale OOS Investigation Additional Laboratory Testing ID: 142167
Download Presentation The PPT/PDF document "Out of Specification and Atypical Test R..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Out of Specification and Atypical Test ResultsSlide2
Contents
Objective
Definitions and terminology
General investigation principles
Investigation stages
Initial Laboratory Investigation
Full-Scale OOS Investigation
Additional Laboratory Testing
Review of Production
Importance of corrective and preventive actionsSlide3
Process overviewSlide4
Cause for an investigation
Out of Specification arises:
Investigation
Analytical (lab) error?
Root cause assigned?
Corrective and Preventive Action?
Production error?Slide5
References
FDA Guidance for Industry : Investigating Out-of-Specification (OOS)Test Results for Pharmaceutical Production, October 2006 : Contains Nonbinding Recommendations
USP– NF General Chapter <1010> Analytical Data – Interpretation and treatment
ICH Q7A Note for Guidance on Good Manufacturing Practice for Active Pharmaceutical Ingredients, November 2000
EU Guidelines to GMP Part I Chapter 6 Quality Control Section 6.32.Slide6
Definitions and Terminology (1)
Out of Specification (OOS) Result
Reportable result that falls outside established specifications or acceptance criteria.
Do not include situations where Control test is not continued due to violations of system suitability tests (SST) limits
Identification of gross operator-errors during a run: e.g. wrong instrument settings
Such deviation cases should be tracked in order to apply CAPA if appropriate
Reportable Result
Result that must be compared to a particular acceptance criterion.
May be “an average value, an individual measurement, or something else”, as defined in the control test [USP]Slide7
Definitions and Terminology (2)
Established Specifications or Acceptance criteria:
Approved specifications or acceptance criteria (e.g. limits) with respect to the final result (reportable result) in order to evaluate the quality of a product or sample.
May be official (part of a regulatory file) or internal (e.g. additional or alternative control tests)
These do not apply to:
Acceptance criteria during the course of analysis (before the reportable result) such as system suitability tests (SST), warning limits, in-process control limits for process adjustments
Atypical Test Result or Out of Trend / expectation, suspect, …
Result within specification
Result different from those usually obtained or expected
Same general investigation principles as OOS results
Site Quality Control Manager must determine extent and depth of any
investigationSlide8
Definitions and Terminology (3)
Re-analysis
Analysis of retained sample preparations
Repeating some part of the testing, e.g.
re-measurement
re-injection
redilution of sample and standard preparations
Further extraction
Fresh analysis
Performing the testing as normal (New execution of the control test)
As described in the control test
Only allowed if initial results could be invalidatedSlide9
Definitions and Terminology (4)
Retest :
New preparation and analysis of a
portion
of the
original sample
Part of a full-scale OOS investigation
Using a predefined procedure/plan with a
MAXIMUM number
of retests,
Used to verify/ reject the possibility of laboratory errors
Control Sample :
A sample of material that has previously been tested and approved or well characterised.
Re-sampling :Collecting and analyzing a new sample from the product
Must have a documented rationale for re-sampling, (e.g. original sample not representative or compromised/contaminated)Used for confirmation of product failureSlide10
Presentation Scope
Applicable to QC laboratories, Manufacturing sites and Affiliates
All GMP relevant laboratory testing including inprocess controls
Where approved specification or acceptance limits are established.
Note:
Compendial
tests
Where interpretation and handling of test results are described, those procedures are to be followed.Slide11
Principles of Investigation (1)
All OOS or atypical test results must
be thoroughly investigated and documented
Includes identification of root cause
OOS or Atypical test results are
only
allowed to be
invalidated
if a
laboratory error can be assigned
Demonstrated or indicated with high probability and/or a product failure definitely excluded
Supporting information and/or data
Timely initiation
(e.g. 2 business days) and completion (e.g. 20 business days)Otherwise written rationale
Sterile products: a parallel investigation in the manufacturing area must be initiated immediatelyA risk assessment and notification in case of confirmed OOS for product already on the market or for a risk that could affect other batches or products on the marketSlide12
Principles of Investigation (2)
Consider:
Problem has occurred previously (historical data)
Possibility that other batches or products are affected
Corrective actions must be defined as a conclusion of the investigation and followed through to prevent reoccurence
A system must be in place for the tracking of OOS results
Investigation phases
Initial Laboratory Investigation
Full-Scale OOS Investigation
Additional Laboratory Testing
Review of ProductionSlide13
Responsibilities (1)
Analyst
Achieves accurate analytical results
Should not use systems that do not meet system suitability requirements (SST); or in such case during analysis identifies data collected during suspect time period (e.g. reference standard injection runs in a chromatographic system)
Should not continue test in case of obvious error (spilling of a
sample solution, incorrect dilution etc.)
Immediately reports the OOS to his/her supervisor
Retains the original test solutions, materials and source data
Participates in the investigationSlide14
Responsibilities (2)
Supervisor / QC Manager
Conducts an objective and timely investigation
Informs Site Quality Manager (
immediately
in the case of a sterile product)
Decides upon investigation of atypical results
Site Quality Manager
Ensures specific site procedures and systems exist
Ensures appropriate training given
Makes batch related decisions when an OOS or atypical result is confirmed
Informs all relevant departments
Assuring that corrective actions are taken to prevent reoccurrence
For sterile products, promptly initiates parallel investigation in laboratory AND manufacturing area.Slide15
Initial Laboratory Investigation
OOS/OOT results identified by Analyst. Supervisor informed immediately
INITIAL LABORATORY INVESTIGATION
Assess whether the test procedure was followed correctly
(calculation; sampling; sample preparation; equipment functionality& qualification; training etc)
Laboratory Error Identified?
YES
NO
See slide 18
(Lab error IS identified)
Full scale OOS investigation required
(See slide 20 )Slide16
Initial Laboratory Investigation:
examples (1)
Method discussed between analyst and supervisor : confirm analyst knowledge of and performance of the correct procedure
Verification of calculations
Review of samples for correct labelling and identification
Verification of correct laboratory test methods
Examination for proper documentation
Review of chromatograms, spectra, data and calculations
Review of reagents, media, diluents, controls and standards
Examination of the instruments and laboratory systems (qualification, calibration, maintenance)
Review other samples run concomitantly (if any)Slide17
Initial Laboratory Investigation:
examples (2)
Verification of analyst’s training history
Inspection of prepared sample
Reanalysis/Redilution of test samples (if stable)
Supervisor fully document and preserve records of the Initial Assessment
Everything except new sample preparationSlide18
Laboratory Error
is
Identified
Invalidate original test results
(Initial Lab investigation- slide 15)
Laboratory error IS identified
Consider Implication on other analytical tests/products
CAPA plan
Fresh analysis
New Evaluation
Conclusion
Evaluation if positive
Implement CAPA action plan and follow throughSlide19
Categorising Errors
Error type:
Description:
Transcription Error
Wrong interpretation or calculation
Identifiable Analytical
Error
Non-adherence to control test instructions
(wrong materials, dilutions, instrument settings etc.)
Equipment malfunctions
Anticipated OOS
E.g. stability, accelerated conditions or beyond shelf-life
Consider Actions for improvement of processes/ prevention of errors – e.g. training analysts; instrument re-qualification, update to procedures.Slide20
Full scale OOS Investigation
Additional Laboratory testing
Laboratory error is NOT identified
(from slide
15):
Full Scale OOS investigation
Review of Production (slide 22)
Full Laboratory investigation
Follow Company protocols on:
Re-testing
Re-sampling
Areas to review will include:
Batch Dossier Evaluation
Equipment: Validation, Qualification, Calibration
Training of Personnel
Risk Management Evaluation
OOS Confirmed?
(See slide 21)
Deviation explaining OOS confirmed?
NO
YES
Batch Failure Investigation
(See slide 21)Slide21
OOS Confirmed?
Batch Failure Investigation
(Additional
Laboratory
testing, slide 20)
OOS Confirmed?
Lab error identified?
Evaluate rejection of batch
Risk Assessment
Quality Alert Notification
Impact on Other Products?
Take all results into consideration
YES
NO
Investigation report
Document.
If yes, further evaluation required
NO
YES
See slide
18,
above.Slide22
Review of Production
Involvement required of all departments that could be implicated: Manufacturing, Process Development, Maintenance, Engineering, Sub-contractor
Timely, thorough, well-documented review
Review of: batch dossier, equipment (qualification, calibration), personnel training and risk management evaluation, etc.
Once Root cause is identified and CAPA plan implemented: OOS investigation may be terminatedSlide23
Additional Laboratory testing; retesting; Maximum number; design
Variables : number specified in advance, analyst, control samples …
Must be defined before testing
Case-by case decision, e.g. depending on
Possible cause of the OOS (analyst, equipment, sample …)
Level of confidence required
Analytical variability
“Extent” of OOSSlide24
Additional Laboratory testing:
Resampling
Resampling
to be carried out:
in accordance with predetermined procedures and sampling strategies
where it is identified that the original sample was improperly prepared and not representative of batch quality
Use same sampling method unless it is demonstrated as inadequateSlide25
Reporting Testing results: Averaging General Considerations
Mean
Estimate of the true value
Reliability increases with the number of determinations
However: may hide (unacceptable) variability
Reportable results and acceptance limits
Correlated
Evaluation of single results requires wider limits due to the broader distribution
If the mean is defined as reportable result, the variability should be checked (standard deviation or range)Slide26
Reporting Testing results:
Appropriate and Inappropriate Uses of Averaging
Appropriate
When sample assumed homogeneous and if the written and approved test method specify that the average of multiple replicates (assays) is considered one test and represents one reportable result
Inappropriate
When testing is intended to measure variability of a product (blend uniformity, content uniformity)
In case of additional testing during OOS investigation because variability is hidden
Specifically when some results are OOS and other within specifications
Provide all individual results to quality management responsible for approving or rejecting the productSlide27
Reporting Testing results:
Outlier Tests
Outlier result, Definition: A value obtained, on rare occasions, markedly different from the others in a series, with a validated method
Outlier testing: statistical procedure for identifying from an array those data that are extreme:
SOP is mandatory
(
includes
minimum number of results required, to obtain a statistically significant assessment)
Used when it is not possible to reveal the root cause of OOS/ deviation
May be used as investigative tool
Should not be used as sole criteria to invalidate dataSlide28
Concluding the Investigation
Results evaluated, batch quality determined,
release/rejection: decision
by Quality Management
OOS cause revealed
:
Suspect result invalidated and not used
OOS confirmed and is caused by factor affecting batch quality
OOS result used to evaluate quality of lot
OOS Cause not revealed
:
Inconclusive investigation
OOS not confirmed
OOS result given full consideration for batch decisionSlide29
Reporting and Documentation
OOS report should include the following:
Chronology of the investigation
Full description of the initial laboratory investigation including an analysis of all data derived from all testing
Batches or products involved
Justification for invalidating any data
Conclusion including conformity of final results and batch
disposition
Corrective and preventive actions taken to prevent the OOS
from occurring again
Uniquely numbered to permit tracking
Maintained in a controlled and retrievable mannerSlide30
OOS Investigations
Clear definition of the reportable result
Thorough, timely, well documented, scientifically justified with no preconceived assumptions
Clear and methodical escalation
Objective to assign a cause
Product failure
Laboratory (analytical) error
Corrective actions to prevent reoccurrence
Atypical results: chance to increase data quality
E.g. in stability studiesSlide31
Examples: Identification of Analytical Errors
Examples of Errors
Solutions
One injection OOS, reanalysis/reinjection ok
Instrument
failure
(
consider
maintenance,
repair
, qualification)
All injections of one sample preparation OOS,
redilution
ok
Dilution error (consider training, clarification of control test)
One sample preparation OOS, retesting 2nd analyst ok
Analyst error (consider training, clarification of control test)
Results near limit, one result just OOS; 1 out of 4 retests also OOS
Reportable result? May indicate unsuitable specification limits (too tight for the analytical variability)
All sample preparations OOS, retesting
OOS, control samples ok
Sampling error or batch failureSlide32
Thank You
Any Questions