Moderator Note: send responses directly to E-mail: bkirsch@home.com
Greetings all,
I have been participating in PDA's PharmSciTech Discussion
Group for quite some time now and know that it is an excellent forum
for exchange of ideas on a variety of pharmaceutical topics. I have
found that the audience is one of the most sophisticated of any
on-line Scientific Newsgroup. It is for this reason that I would
like to bring your expertise to bear on a particularly important and
timely issue and ask a special favor of some of you.
I am planning to make a presentation at the AAPS Workshop on
Stability Practices in the Pharmaceutical Industry scheduled for
March 29-30, 1999 in Crystal City, VA. The talk is entitled:
Handling of Out-of- Specification and Other Unexpected Results - An
Industry Perspective. I will be joined during this session by C.
Russ Rutledge the CDER/FDA author of the new draft Guidance for
Investigation of Out-of-Specification (OOS) Test Results for
Pharmaceutical Production , who will be giving the FDA Perspective
as well as Dr. Peter Cooney, also from CDER/FDA, who will be
speaking about Microbiological Issues in Stability Testing. It
should be a very interesting session.
Since I am only one person with one set of opinions, I would like
to place a SURVEY onto this discussion
group. The Questionnaire is designed to deal comprehensively with
OOS issues and poll a cross-section of pharmaceutical personnel who
are involved, in one way or another, in the Pharmaceutical Stability
process. The survey immediately
follows this message. Please complete it and return it directly to me (NOT THE PDA MODERATOR) by no later than January 20, 1999. This was a suggestion made by the PDA moderator and has his/her blessing. It is preferable that the entire process be carried out electronically, however I will take the
completed surveys in any form (mail, fax, etc.).
For those concerned about confidentiality, I would ask that you
complete the survey and either e-mail it to me in an
anonymous fashion, FAX it to me or snail-mail it to me at addresses
which are provided below. If you choose not to
identify yourself, I would appreciate it if you can identify your
company affiliation and whether your involved with
proprietary, generic or OTC drug products. Your particular dept.
would also be helpful. If you elect to remain completely anonymous
(I would prefer at least a company), please fill out the survey
anyway and just indicate propriet.., generic or OTC manufacturer as
your ID.
Also, Please coordinate with others in your dept. to assure that I
receive only ONE completed survey which is
representative of your dept. practices. If they're obtained
anonymously, more than one survey from a single dept. can seriously
bias the results and compromise the survey.
I apologize for the length of the survey, but OOS and Lab
Investigation issues are complex, controversial and I needed to
cover as much of this important subject matter as I can. The more
information I obtain, the more informative and effective the outcome
will be for all participants. Once tabulated, I will publish the
results of this survey to the entire Discussion group as well as
presenting it at the Workshop in March. If you need to know more
about this Workshop, just let me know or call AAPS at 703-548-3000.
I would really appreciate your cooperation in this effort. I know
your all very busy, but I believe the results of
this survey will begin to answer many questions about just how we
collectively go about handling OOS and other
unexpected stability results.
Thank you all very much or your cooperation!!
Bob Kirsch, Ph.D.
608 W. Nichols Rd.
Arlington Heights, IL 60004-1322
E-mail: bkirsch@home.com
Fax: 847-818-8714
Phone: 847-818-8633
____________________________________________________________________________________
Survey of Industry Stability Representatives:
OOS Data Handling/Lab Investigations
Name(Optional):
Position:
Date:
Dept.:
Company:
1. Are you in any way affiliated with a laboratory, which carries
out testing for stability studies?
Yes (if the answer is no, please do not proceed further)
2. What is your role in the stability process (Briefly describe)?
Stability
Admin.: Lab
Management :
Lab Analyst (chemist or pharmacist): QA/RA
:
Auditor
:
Other (please explain)___________
3. Does your laboratory support Product Development (preclinical,
clinical or generic) or Marketed Product stability studies or both?
(Consider line-extensions, new packaging component(s), new container
sizes as part of Marketed Products).
Product Development:
Marketed Product:
Both:
4. Does your laboratory have an SOP for Investigation of
Out-of-Specification (OOS) Stability Results?
5. Does the SOP deal only with OOS results (as opposed to other
unexpected, but passing results)?
6. If there is a QC Release Unit (Lab) at your site?
7. If so, is your SOP the same as that used by the QC lab? Or is
there an SOP dedicated to and customized for the stability testing
area (i.e. separate and distinct)? Please briefly explain. -
8. Does the SOP require that a checklist of items and activities be
verified by analyst and/or supervisor during the informal lab
investigation (e.g. correct standard conc., sample weights, dilution
volumes, calculations, etc.)?
9. Does you laboratory distinguish and document out-of-trend (OOT)
results from acceptable stability results?
10. Does your laboratory or administrative function use batch
stability results to "trend" stability performance of drug
substance and/or drug product batches over time?
11. Is a statistical procedure used to determine whether one or more
results are OOT? If so what is the procedure (briefly explain)?
12. Does your SOP deals with OOT results as well as OOS results?
13. If so, are OOT results treated the same way (i.e. same
procedure) as OOS results?
14. Are Alert (or warning) limits placed on any of the tests
performed for stability purposes (Assay, Degradation Products,
Optical Rotation, etc.)?
15. If Alert Limits are used, is the procedure for dealing with an
OOA (Out-of-Alert) result included in the same SOP as for OOS
results?
16. What action is generally taken if an OOA result is obtained?
- Increase sampling/testing frequency and closely
monitor.
- Simply monitor next interval more closely and defer
Investigation/decision until then.
- Statistically treat OOA with balance of historic batch
data to determine if within trend.
- If OOA within existing trend, modify Alert limits.
- Other, please briefly
explain________________________________________________
17. Are the investigation/documentation procedure(s) for handling
OOA results the same as for OOS results?
18. Is there a separate procedure or SOP for handling OOA results?
19. If more than one batch/sample is analyzed within the same run as
the OOS sample, are they also investigated even though they may have
yielded passing results?
20. If an OOS result is encountered in your lab, is anyone outside
the lab informed immediately upon discovery by the
analyst/supervisor (i.e. before lab investigation has taken place)?
If so, what function(s)?
21. What function(s) are notified when a formal investigation is
required (i.e. going outside the lab, e.g., R&D, QA, RA, Reg.
Compliance, etc.)?
22. At what point in the OOS process does a Formal Investigation
commence (extra-lab activities)?:
- Immediately upon analyst observation of the OOS?
- Following laboratory conformation of the OOS?
- Following informal lab investigation and prior to or
during Retest?
- Following completion of all lab activities?
- Other times? Please explain:
22. Does immediate notification outside the lab occur irrespective
of whether the sample is part of a batch currently involved in a
preclinical or clinical study?
23. Is the batch record examination part of the lab investigative
responsibilities or more formal (external) investigation?
24. Generally, how many days are permitted for the overall lab
investigation (from OOS discovery to issuance of final report)?
______ Days
25. Are all test solutions, flasks, pipettes, vials,
instrumentation, etc., preserved once an OOS result is reported to
lab supervision (and lab investigation is underway)?
26. If so, at what point are the OOS-related materials released for
cleaning and usage elsewhere in the lab?
27. If retesting is necessary, are the retest samples taken from the
same "composite" of ground tablet powder, mixed capsule contents,
pooled vial liquid contents, etc., as was the original sample(s)?
28. If retesting is necessary, is the retest material taken from the
original collection of intact dosage units used in the original
analyses?
29. Does the lab originally receive at least 2-times (2X) the normal
amount of sample required to perform all stability testing and
retesting?
30. Does the informal lab investigation prior to retesting include
use of the sample composite or does it start with the first
treatment in the sample preparation procedure (i.e. solvent
extracted powder, stock solution, etc.)? Please briefly explain
31. Do you use a separate Retest Protocol in addition to or in-lieu
of the SOP for each instance in which Retesting is called for?
32. For a given type of test (e.g., potency Assay, deg. products)
will the # of retests and retest criteria vary from one situation
to the next (i.e. based on product, study type or interval)? ?
33. Are # of retests and retesting criteria always the same
irrespective of type of test or analyte concentration/potency?
34. What is the number of retests performed if a lab investigation
is inconclusive, that is, yields no determinant error or assignable
cause of the OOS? (If this varies, what is the most common number)?
- _________ Retests.
35. Is management approval required to initiate Retesting of
samples? What is the highest level required? .
36. Does your SOP/protocol allow for combining retest results with
lab investigation reanalyses (i.e. vial reinjection, reanalysis of
stock or diluted solution, analysis of redilution of stock, etc.) in
order to reach a final decision on an OOS result?
37. Does your SOP/Protocol permit averaging of retest results if
they are all pass test limits?
38. Are all values reported on stability report or just the average
value? .
39. Does your SOP/Protocol permit any circumstance where the OOS and
retest results may be averaged (e.g. all within the same statistical
population)?
40. Does your SOP/Protocol permit averaging of retest and original
results if they are all within the test acceptance limits (for OOT
or OOA only)?
41. Does your SOP/Protocol require a Retest result variability
criterion (i.e. all values within a specified RSD or confidence
interval (CI)) in addition to passing the test*s specified limit(s)?
43. Does your SOP/protocol permit averaging of passing retest
results with one or more original OOS results if they have all been
shown statistically to be part of the same population (i.e. using a
95% CI, standard deviation or RSD)?
44. Does the SOP allow the calculation of a Confidence Interval (CI)
around the mean of original and/or retest values? If so, must the
entire CI be within the specified range of the test?
45. If the CI is calculated around the average of (retest and
original values) and is entirely within the specified limits of the
test, could one value be OOS and the test still pass according to
the SOP?
46. Does the SOP procedure require:
- that a second analyst carry out the entire set of
retests?
- that both original and second analyst carry out the
retests?
- Depending on circumstance of the OOS, one or the other
of the above could apply?
47. What action is taken if the OOS result(s) is/are confirmed by
retest/resample results, i.e. the OOS is real? ? Please briefly
explain :
48. Is the response time to outside authorities different if sample
is from clinical supply in man or preclinical supply in animals?
(Development Stability)?
49. Does the SOP call for evaluation of previous method performance
(e.g. erratic results, bias, high variability, temp. /light/heat
sensitivity, etc.) to rationalize OOS result?
50. Can Outlier testing be used under any circumstances in which a
chemical test is performed in your lab? Yes ® No ®
51. Can an Outlier test be used to make a decision on original
and/or retesting results?
- without any other supporting analytical evidence.?
- along with a preponderance of other supporting
analytical evidence?
- If the composite sample from which the original and/or
retest samples were drawn is highly homogeneous
(e.g. well-ground tablets, well-mixed capsule contents,
pooled aqueous solutions, etc.), so that an outlier is more
easily detected?
- more than one of the above situations is true?
- Any other requirements for use of Outlier testing?
Please briefly explain:
52. If an Outlier test is used, what actual statistical procedure is
performed (e.g., Dixon, Q-test, etc.)
Do Stability dept. personnel or other biometrics or statistical
specialists outside the dept. perform the procedure or
consult on its results?
53. If an Outlier test is used to evaluate an OOS result, are all
results (original and retest) reported irrespective of the outcome
of the Outlier analysis?
54. Under what circumstances can Resampling be performed (i.e.,
obtaining another analytical sample from outside of the original
collection brought to the laboratory)?
- Original sample is adulterated?
- Original sample is exhausted?
- Original sample is suspect, however no definitive
evidence can be given?
- There is no provision in the SOP for Resampling.
- Any other reason? (Please briefly explain)
55. Does the SOP/protocol spell out how the resample material will
be taken?
56. Is the resample procedure described in a separate SOP (e.g.,
QA)?
57. Is there a deadline for completion of the final Investigation
Report? If so, how long?
58. If there is an OOS Investigation deadline, is it fixed or can it
vary with sample type or other circumstance (clinical supply,
preclinical supply, marketed product, etc.)?
59. What functions other than analyst and supervisor must Approve
(sign) the Final Investigation Report?
60. Has at least one of your investigations conducted using the SOP
been reviewed by FDA inspector(s)?
61. If so, was it found to be acceptable?
62. Was it found to be acceptable with comments?
63. If inspector comments were obtained based on review of an
investigation, could you share them with us? Please briefly
describe the most significant two items.
END
THANK YOU VERY MUCH FOR PARTICIPATING!!
Please return directly to:
BOB KIRSCH
608 W. Nichols Rd.
Arlington Heights, IL 60004-1322
E-mail: bkirsch@home.com
Fax: (847)818-8714
Phone: (847)818-8633
|