Me Din A Apr 2009

Handling OOS and OOT Results: FDA’s Current Expectations Prepared by

Precision Consultants, Inc. Carmen Medina, M.P.H., Former FDA Investigator

2000 by b Precision P i i Consultants, C lt t Inc. I All rights i ht reserved. d Reproduction R d ti in i partt or in i whole h l without ith t written itt permission i i is i strictly t i tl prohibited. hibit d Precision Consultants, Inc. is a registered name.

Optimize current OOS and OOT Investigation Procedures

– – – –

Judge Wolin’s Ruling (aka Barr Decision) Current OOS Investigation Requirements Out-of-Trend Issues Annual Product Review: QC Profiles

OOS—Out-of-Specification OOT—Out-of-Trend 2

Judge g Wolin’s Rulingg Some of the items in the Barr decision and areas relevant are: Retesting The decision to retest “must be based on the objectives of the testing and sound scientific judgment. judgment Rational retesting should be done by a second analyst.” -Judge Judge Alfred Wolin

Judge g Wolin’s Rulingg Retesting

There should be a retest procedure defining the number of tests. Must not try t retest to t t a product d t into i t compliance li (averaging, etc.) If there are grounds for invalidating an original result and replacing it with the retest value, the original g data must be kept p along with an explanation of what and why. 4

Judge g Wolin’s Rulingg Averaging Results

May be a valid strategy, depending upon the sample and how the technique is used. Averaging A i - in i certain t i situationsit ti can provide a truer picture of reality, but when used inappropriately inappropriately, can hide important variations.

5

Judge g Wolin’s Rulingg Resampling Should rarely occur! Will occur when the original sample was not truly representative of the batch or there was a documented/traceable lab error in its preparation. Sound scientific justification must be employed if resampling is to occur. 6

Failure Investigations DOCUMENTATION OF OOS

• Phase ase I:: Who, W o, How ow (Lab) ( ab) • Phase II: Who, How (QA/QC Mgr/Supervisor)

7

Fail re Investigations Failure In estigations PHASE II - FORMAL INVESTIGATION

• Ruled out laboratoryy error • QA/QC ppersonnel with lab staff • Standardized form (UIR, DR, etc.)

8

Fail re Investigations Failure In estigations LABORATORY ERROR OR ANALYST ERROR (LABORATORY -- INFORMAL INVESTIGATION)

• Incorrect standard • Calculation error • Instrument malfunction 9

Fail re Investigations Failure In estigations NON-PROCESS ERROR (INFORMAL INVESTIGATION)

• Human or mechanical malfunctions (e.g., (e g an uncalibrated meter; incorrect amount of raw material;; wrongg ingredient) g )

10

Fail re Investigations Failure In estigations PHASE 1-LABORATORY INVESTIGATION OR INFORMAL INVESTIGATION

• This information would get generated by the OOS observer, onto a standardized form. • The results would determine if a formal investigation is required. required

11

Fail re Investigations Failure In estigations An OOS may be invalidated via explanation an/or retesting. Always y begin g byy eliminatingg the followingg possible p errors: •

Operator

•

Laboratory/Analyst

•

Process

•

Method

•

Non process Non-process 12

Fail re Investigations Failure In estigations If an OOS cannot be demonstrably identified: • Review batch records or device history file to identify any trends related to OOS results. • Revisit the process validation to identify sources of variability related to OOS results • Define Failure 13

Fail re Investigations Failure In estigations • Review of full set of test data for the batch to identify relationship between parameters that may suggest a cause of error. Start with high and low OOS results. • Revisit i i pilot il batch b h data. d Design i off experiment i

14

Fail re Investigations Failure In estigations THE SOP:

• O Out-of-Specification t f S ifi ti SOP should h ld contain t i directions on handling and properly disposing of an OOS laboratory or QC test results. results

15

Fail re Investigations Failure In estigations • What is the difference between an out-of-specification result (OOS) and a “failure”? • When does an OOS not require an investigation?

16

Investigation of Out-of-Specification Laboratory Test Results – General

• cGMP training on the handling of OOS and atypical results must be part of an analyst training program prior t commencing to i workk • Test results that are invalidated byy an investigation g are not used to determine the final result for the test, but these results must be retained in the test documentation

17

What Do Yo You In Investigate? estigate? • OOS test results for any material or process controlled under specifications • OOS laboratory test results obtained during the t t for test f disposition di iti off in-process, i fi l andd final stability testing of raw materials, in-process materials,, finished components p and finished kits • OOS laboratoryy test results obtained duringg validation studies 18

T pes of Errors Types • Clerical Error: – An error that can be attributed to an incorrect calculation or transcription error

• Determinate i Error: – A known or identifiable error

• Indeterminate Error: – An error that cannot be identified 19

T pes of Errors Types • Laboratory Error: – An error that occurs in the process of testing, which can be ascribed to an analyst error error, an equipment error or an unplanned test interruption

• Non-Process Related Error: – An error that occurs during the product manufacturing f t i process, which hi h can be b ascribed ib d to t an operator error, an equipment malfunction or an unplanned interruption of the manufacturing process 20

Wh an OOS T When Testt R Result lt iis Obt Obtained i d • The analyst immediately notifies the area supervisor / designee of the occurrence of an OOS result • Initiate an investigation into the OOS result according to written procedure described to determine whether the result can be attributed to error or whether the result represents the actual condition of the material or process being evaluated 21

Wh an OOS T When Testt R Result lt iis Obt Obtained i d • Any exceptions to the timely initiation and completion of all investigations must be notified to and authorized by laboratory management • All sample preparation materials, materials such as vials or solutions, media, working standards, glassware, pipettes, etc. used in the original analysis must be retained until the test results are verified and approved. The stability of solutions must be considered when utilized for investigation purposes 22

OOS In Investigation estigation Will Include Incl de • A review of product history and any stability trends • An evaluation of the feasibility of re-analyzing th stock the t k or working ki sample l andd standard t d d preparations • The consideration / use of a second analyst in conjunction j with the first analyst y duringg retestingg 23

OOS In Investigation estigation Will Include Incl de • The consideration of using check or reference samples p • Documentation and retention of all results obtained during the investigation • A documented final decision based on all of the valid results obtained during the investigation 24

What May Be Included in the OOS I Investigation ti ti • Confirm the correct test method was used for testing • Confirm the analyst is qualified to perform the test method • Confirm calculations (if applicable) are correct • Confirm,, if applicable, pp , all negative g controls for media, diluents, and test equipment were negative 25

What May y Be Included in the OOS Investigation • Confirm environmental samples taken during testing were satisfactory • Determine if the sample was taken properly by a qualified individual • Confirm instruments used and other laboratory systems y ((where applicable) pp ) were in calibration and functioning properly 26

What May y Be Included in the OOS Investigation • Determine if other samples tested in the same time frame using the same lots of solutions and standards and testing equipment were satisfactory • Review historical data to determine if similar problems have been previously reported

27

Repeat Testing • Repeat testing can be performed only after an investigation g of failure is initiated. Repeat p testing may be done by either the original analyst or a second analyst

28

Investigation of OOS Microbiology T tR Test Results lt • The h ffollowing ll i describes d ib suggestedd procedures d to be b followed for resolution of out-of-specification microbiology test results. The same ppractices mayy be utilized,, but is not mandatory to investigate out-of-trend microbiology test results • The following items are suggested, but not limited to and should be considered and documented as ppart of the investigation in the most common microbiological testing. For the tests where a retest procedure is specified in the method or its associated documentation documentation, the method supersedes any OOS result investigation procedure 29

Investigation g of OOS Microbiology gy Test Results 1.1 12 1.2 1.3 1.4 1 1.5 1.6

Confirm the correct microbiological test method was used for testing Confirm the analyst is qualified to perform the test method pp ) are correct Confirm calculations ((if applicable) Confirm all negative controls for media, diluents, and test equipment were negative C fi growthh promotion Confirm i testing i for f all ll media di were satisfactory Confirm environmental samples taken during testing were satisfactory 30

Investigation g of OOS Microbiology gy Test Results 1.7

Determine if the sample was taken aseptically by a qualified individual 1 8 Confirm incubators, 1.8 incubators hoods and isolation chambers and other laboratory systems (where applicable) were in calibration and functioning properly 1.9 Determine if other samples tested in the same time frame using the same lots of media, diluents and testing equipment were satisfactory 1.10 Review historical data to determine if similar microbiological problems have been previously reported 31

Investigation g of OOS Microbiology gy Test Results • Repeat testing can be performed only after an investigation of failure is initiated. Repeat testing may be done by either the original analyst or a second analyst • Reporting of mean values of the microbiological i bi l i l data d t is i appropriate i t when h defined by a validated test method 32

Reporting / Documentation of OOS or Atypical Results • All ll results l must be b retained i d in i laboratory l b records. d Iff data d is i rejected, a complete description of the reason for rejection must be included in the analyst’s y records with the signatures g of both analyst and supervisor • All results from the original and repeat testing must be reported except for results that are determined to be invalid • The results of the failure investigation must be recorded in the laboratory records, and the OOS or atypical results report. Implications for previous or subsequent batches must be assessed by QA, if applicable 33

OOS Report Incl Includes des • A clear chronology of the investigation • A summary of the testing that was performed • An A analysis l i off all ll data d derived d i d from f the h testing i • A An analysis l i off any supporting ti data d t from f other th tests performed

34

OOS Report Incl Includes des • The justification for invalidating any data • The final results and the conformity of the test and/or the batch • Any corrective actions required by the outcome of the investigation

35