ML20067E218
| ML20067E218 | |
| Person / Time | |
|---|---|
| Site: | Farley  |
| Issue date: | 02/06/1991 |
| From: | Hairston W ALABAMA POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9102140105 | |
| Download: ML20067E218 (8) | |
Text
,
f
- i 4
- Alabama Power Compary
-40 inverr'ess Center Parkway
- Post Othee Dox 1295 Birmingham, Alabama 35201 Telephone 205 868 5501 i W. G. Hairston, lll -
~ Senior Vice President
- l Nuclear Operations AlabamaPower the southem electoc system February 6,1991 Docket Nos. 348 50-364-i J
U.S. Nuclear Regulatory Commission Attn:
Document Control Desk Washington,-DC 20555'
. Gentlemen:
Joseph M. Farley Nuclear Plant Reoort of Unsatisfactory Performance Testino 10 CFR 26 Appendix A, paragraph 2.8 requires:that blind performance test-specimens be submitted to each HHS certified laboratory performing drug screens for the licensee.: On January 8, 1991 two presumptive positive samples were forwarded from-the Southern Nuclear Operating Company screening facility. to SmithKline Beecham Clinical Laboratories-(SBCL) for confirmation testing. One sample-was a known blind positive and the other was-an actual sample. On January 15, 1991 both sam)les were reported to be negative by the Medical Review Officer. Since tie blind positive sample had'not been reported to Southern Nuclear as a positive test, an investigation was begun.
The-subsequent investigation determined that an error had occurred on the-part of SBCL, The blind positive sample and an actual sample were assigned consecutive log numbers by SBCL on January 9, 1991.
Both of these two samples tested positive.in the initial screening for.
(
amphetamines. An admiaistrative mistake was then made during the confirmatory testing using a gas chromatograph-mass spectrometer. The results of the tests of these consecutively numbered samples were reversed. On January 18, 1991,.. as a result of the discovery that a mistake was made, both samples were retested by SBCL and the actual results were made known. The Medical Review Officer had originally dispositioned the erroloyee sample as ~ negative when it was received by him
.from SBCL due to prescribed medications the individual was taking-which the Medical Review Officer believed would generate a. positive test result.
- Therefore, no personnel sanctions had been imposed against any individual.
'SBCL has taken. action to prevent recurrence of this kind of error in the future.
i gg21;gggyfa88gge i i P
2
s.34 - -.
=. -
U.' S. Nuclear Regulatory Commission.
Page 2:
This is _the second-report of unsatisfactoryEperformanceftesting -against
'SBCL1byia-Southern Company utility, (Ref: Alabama Power Company to U. S.
l
. Nuclear Regulatory Commission letter of May 25,-1990), JLs a. result of -
this second unsatisfactory performance, the: Corporate. Quality Services -
organization of Southern Nuclear 0perating Company will be auditing SBCL in the near future..
Enclosed 'are the findings and corrective actions' taken by SBCL as reported
.from William Shaw, Ph.D.,.
of SBCL.
This letter is considered to satisfy-the reporting requirements of 10 CFR 20 Appendix A,. paragraph 2.8.
If you have any questions, please advise.
Respectfully submitted, Alabama Power Company (tl.
W. G. Hairston, III WGh,It.1/JMG
~
Enciosure
- cc: Mr. S. D. Ebneter
-Mr.
S'. T. Hoffman; Mr. G; F. Maxwell:-
L i
i i
e SmithKhne Beecham ClinicalLacoratorie=
January 29,.1991 TO:-
' Paul'Bizjak cc:~W. Mercer, Ph.D.
~
.FROM:
William Shaw, Ph.D.
p[
SUBJECT:
Amended Amphetamine Results On Southern Company Samples Two samples from the-Southern Company, C910008 and C910020, were received-into -this laboratory for drug testing on 1/9/91 about 3:00_ am.
1These' samples,were assigned SBCL log numbers 090195G and-090196G, respectively.'.Both samples were aliquotted for screening and' tested.
positiveffor amphetamines.
Since.the screening tests were positive'for amphetamines,.both samples were pulled from temporary storage'and-aliquots were removedfor confirmation.
The aliquots-for-GC/MS confirmation were. poured'on-1/9/91 by the aliquotter and' transferred to
-. technologist'A'who performed the extractions and loaded _the autosampler of)the GC/MS. ;The contents of the vials were injected by'the autosampler and reviewed by technologist B, checked'by, technologist C, j
tend. certified by a-' certifying. scientist.
Results were then released.
1
-In the-GC/MS run on 1/9/91, the employee sample 090195G was determined
-to be positive with-an-amphetamine value of 2702 ng/ml urine.
The results for sample- 090196G which-weolater learned was'the blind sample,
. wore-negative for-amphetamine.
All quality control samples, standards, and~ blind samples were acceptable.
The retention time and ion ratio values of-the amphetamine on sample 090195G were all acceptable.
>o so. ccm. wm u. u.ic2. i cc w a2.n
Memo - Shaw
- 1/29/91
-Page 2 Less than one week after reporting these results, we were notified by the Southern' Company that sample 090196G was a blind positive sample while=SBCL had reported the results as negative.
Based on these results, we repeated the screening tests and also repeated the GC/MS tests on each of the samples on 1/16 and 1/17/91.
These results contradicted the 1/9/91 results.
An additional GC/MS run on 1/18/91 demonstrated that sample 090195G was negative and sample 090196G was positive for amphetamine with a value of 2748 ng/ml.
These results also-contradicted the initial GC/MS run on 1/9/91.
Furthermore, the similarity of quantitative results 2748 ng/ml for 090196G on the 1/18/91 run and 2702 ng/ml for sample 090195G on 1/9/91 led to the strong suspicion of a sample mixup between sample 090195G and sample 090196G.
We received information from Paul Bizjak that the employee results had been reported to the company as a negative.
The employee results tested positive on the screening test because the employee was on the drugs phentermine and tenuate, drugs related to amphetamine.
This -information was obtained from Mr. Bizjak from information obtained in the Southern Company drug screening program.
Our GC/MS procedure:is not subject to-interference from phentermine or tenuate.
-The use of phentermine by the employee explains why the screening test
-was positive since phentermine does give a false positive screening
-result with the EMIT test but not with the GC/MS test.
- - =.
o,
+
i Memo - Shaw' l
1/29/91
)
Page.3' A summary of all tests results on the two samples is give in Table 1.
All of the data indicate that the screening tests were done correctly.
Both samples were reactive with EMIT.
The employee sample contained no amphetamine or methamphetamine but contained amphetamine-like compounds that gave a positive EMIT result.
Such i
reactions are extremely common and I emphasize that this finding was not an error.
The-employee sample should have tested negative by the confirmation test but was comehow switched with the blind sample that differed by!only one digit of the log (accession) number.
The sample switch had to.have been made from the time aliquots were taken for
~
GC/MS to the time:the extracts of the samples was loaded on the
.autosampler of the GC/MS',
The mixup could have occurred at the time
,the sample was aliquotted for GC/MS, during-extraction, or when the vials containing extracts were transferred to the GC/MS.
I reviewed each of the procedures for sample identification during each of these steps:with each of the persons involved and found that all procedures ~for identification were followed.
Barcode t
identification is used to remove aliquots for GC/F, so that I think it is unlikely (but still possible) that an error occorred at this step.
f
..,_t Memo - Shaw
~
1/29/91
- Page 4 i
I think that the error probably occurred during the extraction when manually written ~ identification numbers are written on tubes and vials.
The two identification (log) numbers differed by only a' single digit (5 vs 6) in a seven digit code.
Furthermore, a handwritten 6-at
- the end of the identification code may appear similar to a handwritten letter G.
I had to stop several times when reading these numbers to prevent an error in writing this report.- The Technologist A who i
performed the, extraction is rated as a very good technologist.
- However, she stated that she was interrupted in the middle of the run by the need to attend a safety meeting.
Her concentration or normal precautions may have been disrupted by the meeting.
The most-likely cause of the error is a human error in which two 1
identification codes (log numbers) differing in a single digit out of neven digits were-r.isread.
I have held a meeting with the entire a
staff in which all sample identification steps have been reviewed.'
Strict compliance to all sample identification steps has been re-emphasized to all staff.
I have also asked Dave O' Bryan, Ph.D., a SBCL Vice-President in our corporate office to evaluate the-feasibility of adding a system ~ capability for barcode identification
- labels that match our log numbers and which would be placed on-the vials containing. extracts for positive samp'le verfication by the barcode readers of the autosamplers of the GC/MS systems.
Mem'o - Shaw
- 1/29/91 Page 5 This is the first error due to probable human error in over 100,000 samples processed.
Furthermore, the error was detected and corrected within a week of its occurrence, demonstrating the fundamental soundness of the SBCL drug testing program.
The accuracy rate for this laboratory for NIDA/NRC samples is 99.999%.
However, we are not satisfied with this performance end will use this opportunity to make further improvements in our sample identification procedures so that we can achieve 100% accuracy.
I have completed meetings scheduled with all staff to review all aspects of sample identification and many good ideas have been generated which will be evaluated over the next several weeks.
The critical nature of this testing has been re-emphasized to the staff and several enhancements for sample identification will be evaluated.
A review of all sample identification procedures that are being ovaluated should be completed by March 15, 1991 and improved procedures will be implemented before that date if increased reliability is evident.
Employees who make identification errors are subject to disciplinary action.
However, disciplinary action was inappropriate in this case because the person making the error could not be definitely established.
A full report on all identification enhancements will be forwarded to you within 30 days.
.e
- . vt ;.
..," W ;
~
y
'"I i
Table 1 Summary Of Test Results On Samples'For' Amphetamine: Testing From The. Southern Company.-
-Initial
'Second-1/9/91' 1/16/91.
1/17/91 1/18/91' Sample..
EMIT.
EMIT.
.GC/MS
'GC/MS, GC/MS
.GC/MS l
Identification Screen.
. Screen Amp.
Amp.
Amp.
Amp.
C-910008 (090195G)
. Amp. -
Amp..
2702 Not Employee Sample.
Positive;.
' Positive ng/ml Done Negative Negative-
. j Containing Tenuate i
and Phentermine C-910020 (090196G)
Amp Amp.
2718 Not 2748 Blind Positive Positive
. Positive Negative
.ng/ml-Done
.ng/ml for Amphetamine i
4 f
e
. ~ _
., ~,,...
., ~ _. _ -
g, v.