ML20043B594

Provides Rept of Unsatisfactory Performance Testing,Per 10CFR26,App A.Error Caused by Olympus Analyzer Which Allowed Same Barcode to Be Assigned to Two Different Samples. Smithkline Taken Action to Prevent Recurrence of Scan Error

ML20043B594
Person / Time
Site:	Farley
Issue date:	05/25/1990
From:	Hairston W ALABAMA POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9005300258
Download: ML20043B594 (6)

Text

_.. _ . _

• , *' ~ Altbama Power Company 40 Irworness Center Parkway

• Post Office Box 1295 BirminDham, Alabama 35201 Telephone 205 868-5581 W. G. Haltston, til Senior Vice President Nuclear Operations AlabamaIbwer May 25, 1990 the sourhorn ekctresniem 10CFR26 Docket Nos. 50-348 50-364 U. S. Nuclear Regulatory Commission Attna Document Control Desk Vashington, D. C. 20555 Gentlemen:

Joseph M. Farley Nuclear Plant Report of Unsatisfactory Performance Testing 100FR26 Appendix A paragraph 2.8 requires that blind performance test specimens be submitted to both an onsite test facility and an HHS-certified laboratory. On May l', 1990, a blind sample of urine was processed through the Farley Nuclear Plant screening facility and was determined to be presumptive positive for cannabinoids. The blind sample was then shipped on May 2, 1990 to SmithKline Beecham Clinical Laboratories for confirmatory testing. On May 7, 1990, a false negative determination was reported by Alabama Power Company's Medical Review-Officer based on the testing results supplied by SmithKline.

Alabama Power Company subsequently conducted an investigation of this error in conjunction with SmithKline. The cause of the error was determined to be an Olympus analyzer which allowed the same barcode to be assigned to two different samples. All previous Olympus data for all clients were reviewed by SmithKline and two similar errors were found.

In both of these cases, data were incorrectly transferred but the incorrect data transfer did not result in false positive or false negative patient results. All performance samples submitted by Alabama Power company to SmithKline' prior to this sample have been properly reported. SmithKline has taken action to prevent recurrence of the barcode scan error.

Enclosed as Attachments I and II are the findings and corrective actions taken. This letter is considered to satisfy the reporting requirements of 10CFR26 Appendix A paragraph 2.8.

-nn - $r pn awcx **y g, gd: p y is/go i ,

i

i

,. l i

• i

! U. S. Nuclear Regulatory Commission 3

Attn Document Control Desk Page 2 .

I If you have any questions, please advise. j Respectfully submitted, ALABAMA POVER COMPANY l

1 l

th. $ Y V. G. Hairston, III j i

VGH,III/JARamV1739  ;

1 Attachments cci Mr. S. D. Ebneter i Mr. S. T. Hoffman Mr. G. F. Maxwell I l

4 l

\

I.

t t

l t ..

i o

t F '

i.

, RITACIMNT I s g, . . -

IMV-14-90 fl0N 15:32 Sr.BLATLANTA FAX NO. 4049B67153 P.02

( ,

'.. SmothKime Beecham ClinicalLaboratories May 11, 1990 Tot Paul 31sdak FROM William Shaw, ph.D.

Forensic Toxicology Director 80832C7: False Negative Results On Blind Drine 9 ample Submitted By.The-Alabama Power company on May 3, 1990 a blind urine sample, accession $0241350 from the Alabama power company was submitted to this laboratory. The sample contained THC at a concentration which should have been reported as a positive result.

Mowever, this laboratory reported all test results on this sample as-megative. The following actions were taken to resolve this error.

1. The sample was rotested with the screening test and was found to be positive. The sample was then tested for THC by the OC/MS.

The result was confirmed positive'for THC by OC/MS.- An amended report was sont indicating the-altered test result for THC.

2. The MIDA management group Dr. Shaw, pat ten-Dev, Ben Wells.

Valencia Rhodes, and several of--the Olympus operators met-to review Olympus operating procedures. The Laboratory Director, Dr. Nabb, an expert in barcode technology and automation was called to the site to saamine the situation.

3. The appropriate SmithKline Beecham corporate managers were notified of the problem immediately.

4. A service engineer from the Olympus Corporation was called in to examine the problem. A copy of his examination is attached.

8. A memo to the staff (copy attached) defining the problem and the observed actions to be taken to prevent a recurrence-was sent out on May 10.

6. All previous Olympus' data for all. clients were reviewed.

Similar errors were found to have occurred only twice.- In both of those cases, data were incorrectly transferred btit ths~

incorrect data transfer did not result in false positive or negative patient results.

A7D O D O4T II W-14-90t0H15:33 SKPLATLANTA FAX NO. 40:0BB7153 P.03 SmothKhne Beecham ClinicalLaboratories May 9, 1990 901 All MIDA Technical staff front William Shaw SUB3ECTt Palse Negative Results on Blind Urine Sample Submitted By The Southern company on May 3, a blind sample, accession 80241350 from the Alabama power Company was submitted to this laboratory. The sample conteined THC at a concentration which should have been reported as a positive resalt.

However, negative. this laboratory reported all test results on this sample as An examination of this testing error revealed that the Olympus system read two consecutive samples with different barcodes 2150 and 2151 as having the same barcode (2150). When the olympus data was transferred to the NIDA computer system, the NIDA computer system incorrectly-interpteted the results from the sample in the sup at position 21511as the final results for barcode 2150, because the OV8RLAY function was activated.

The data from sample position 2151 did not transfer over to the olympus Summary Report of the NIDA ecmputer system and therefore no raw data or results printed for barcode 2151- en the olympus Summary Report. The technologist and certifying scientist recognisod'this error for barcode 2151 but did not realise that incorrect raw data from sample 2151 had been transferred to position 2150 on the olympus Summary Report. This incorrect data transfer had not been previously recognised.

The service reasor, for this engineer thaterror is included in the report of the olympus-is attached. '

Our corporate office has been notified of this flaw. The FDA will also be notified of thia flaw. The ultinate solution to this flew will probably be a software modification.

i

, NY-14-50110N 15:33 SEBLATLANTA FAX NO. 4049B67153 P.04 l

Memo - False Negative Results 5/9/90 page 2 To prevent a recurrence of such an error, the following steps will be taken.

1. I. eave the OVERLRY function off unless a load is being repeated.

Immediately, turn off the OVERLAY function after a repeated load is completed.

2. The olympus operator must examine the olympus summary Report from the N1DA system and make sure that all results are reported 1

in the proper locations. A blank reporting field on the olympus summary Report is'an indicator of=possible barcode failure. The olympus operator should then note the missing data at the sequence number and at the top of the summary report. The raw data and the summary rep" ortThe must then be placed in the f older titled "Barcode Failures . sample rack with the repeated barcode must then be repeated.

l

3. The certifying scientist will, as an' additional check, review the olympus summary Report for any missing data. If missing data are located, the samples in the affected rack must be-repeated. The certifying scientist must also state "No Barcode Errors Detected" and' initial.

4. This same type of error may also occur with 2DP results.

Therefore, the Olympus operator must review all Olympus sheets and check to see if there are any samples with dashes across the page. If these results are detected the antire sample rack must' be repeated.

5. All technical staff needs to sign oft en the sheet on the bulletin board to indicate that they have read and understand the material in this memo. If you do not understand this issue, please contact me or your supervisors for additional training.

L l i

+ _ _ _ _ - - - - - - - - - - - - - - - ',e =vv w% www~m -- ,

I l

1 . '

fmY-14-90M3N15:33 SEBLATLANTA FAX NO. 4049BB7153 P.05

. 1

+

. i

w. ? .

(tirrser  !

OLYMPUB TVhtVATION! (John Ratter a Engineer) -

4 If there is any delay in the rack leaving the cr-11 it may read the rack.lastInbarcode this situation twice and call it the first barcode in the next  !

it will not initiate any audible '

, a arms, but the of the specimen will be correct.sequence, number will be correct and the analysis- '

the left of the barcode indicating a duplicate barcode.There will also be a flag -Al in this instance the (1) flag is printed to:the left of the i analysis number indicating the duplication of the barcode number. This flag is seen anytime the barcodes are rerun on the same day. The olympus analyser will allow the same barcode to -

be rerun as long as it is not in the same rack.. If it sees.the t same barcode twice in the same sample rack it will generate an audible alarm and print out " Broken Label". In this instance.  !

l -

L due to the delay of the rack leaving the CF-11, no audible alarm was generated.

problem: Delay in rack leaving CF-11 Most in overlay (

n Solutient correct any recognised problems which create delays /

CF-11. i L

Review worksheet for flag when not intentionally repeating in overlay mode.

• Review cautions mentioned on page 1 as alert on previous examp1'e. 'i

't The host system may overwrite the initial barcode. analysis results when operating in the overlay mode. ,

I f

D

+- -, -- -+ . . - + -

-