ML20077S579
| ML20077S579 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 12/21/1994 |
| From: | Willette R External (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML20077S571 | List: |
| References | |
| NUDOCS 9501240156 | |
| Download: ML20077S579 (2) | |
Text
.,
INVESTIGATION REPORT on Blind Performance Test Samples Obiective:
The licensee shall investigate any unsatisfactory performance testing result.
A record shall be made of the investigative findings and the corrective action taken by the laboratory. The licensee shall send the document to the NRC as a report of the unsatisfactory performance testing incident within 30 days.
References:
1.
10 CFR 26 Appendix A, Subpart 2.8(e)(4) 2.
FR 53, 11970 (1988), Subparts 3.19(b)(2) & (5)
Observation:
Two blind quality control samples containing a combination of morphine, morphine-3-glucuronide, codeine and 6-acetylmorphine were analyzed by the SmithKline Beecham Clinical Laboratories-Norristown. One of the samples was reported incorrectly as negative and the second sample was reported positive for morphine and 6-acetylmorphine. An unannounced laboratory site visit was conducted by Dr. Edward J. Cone on December 8, 1994.
Findina:
The sample reported incorrectly as negative for morphine was prepared by the Seabrook site and received by the laboratory on October 7, 1994. The sample was subjected to the initial immunoassay test on October 8, 1994. The sample gave a positive response in the Emit opiate assay, with a value that was 1.553 times the cutoff calibrator (300 ng/mL of morphine) value. The sample was then submitted to the GC/MS confirmation assay on October 8, 1994. The sample was analyzed for codeine and morphine, obtaining concentrations of 97 and 273 ng/mL, respectively. The control samples in the run tested appropriately within expected ranges and the run was accepted. Since the morphine result was below the cutoff of 300 ng/mL for morphine, the sample was reported as negative.
An identical sample was submitted by the Seabrook site, which was received on August 20, 1994, and two other identical samples were also received on August 27, 1994 and October 25, 1994 by the laboratory, which were submitted by dif-ferent clients. One of these samples was submitted under DOT testing rules.
The three samples gave similar screening values of 1.669, 1.490, and 1.667, respectively. The August 20th Seabrook sample gave analytical results of 169 ng/mL of codeine, 317 ng/mL of morphine, and 31 ng/mL of 6-acety1 morphine, and was reported as positive. The August 27th DOT sample gave analytical results of 93 ng/mL of codeine, 346 ng/mL of morphine, and 36 ng/mL of 6-acetylmor-phine, and was reported as positive. The October 25th sample gave analytical results of 132 ng/mL of codeine and 235 ng/mL of morphine, and was reported as negative.
The average reference values obtained on this sample lot was 112 ng/mL for codeine, 522 ng/mL for total morphine (119 ng/mL of free morphine plus ca. 500 ng/mL of morphine glucuronide), and 61 ng/mL of 6-acetylmorphine.
9501240156 950111 PDR ADDCK 05000443 p
PDR j
l
~
Invesdgadon Report Page 2 The current results are similar to those found for three identical samples anglyzed in March-May 1994 by this laboratory. Morphine quantitation by the laboratory was substantially lower than found by the reference laboratories.
As previously noted, the results suggest incomplete hydrolysis of morphine glucuronide. However, the laboratory includes high and low control samples that contain a mixture of morphine and morphine glucuronide to monitor the efficiency of the hydrolysis process. The recovery of morphine in these l
samples was within the expected range of efficiency. Several other steps in the sample extraction process could also account for the low recoveries of l
morphine including addition of too little enzyme solution to effect hydrolysis in individual samples, incorrect pH adjustment which could alter the effici-ency of hydrolysis or extraction recovery, and variable efficiency of the solid phase columns for removal of morphine from solution. These latter vari-ables would apply to single samples rather than to entire betches.
Conse-quently, control samples would test appropriately in most cases even when individual specimens did not.
This type of performance is the most difficult to detect and to correct since it is random in nature and often dependent upon the analyst rather than the specific assay procedure.
1 1
Recommendation:
l These results represent a disturbing trend suggesting that the laboratory's
[
confirmation assay for morphine may have produced inconsistent results over l
the course of May, 1994 to the present time. Morphine samples from the same lot have consistently tested low, some sufficiently low to be reported nega-l tive. The August 20th Seabrook sample that was reported as positive was placed in frozen storage by the laboratory as part of its routine procedure.
Therefore, it is recommended that the laboratory reanalyze the August 20th Seabrook sample and submit an appropriate aliquot to an outside reference laboratory for independent analysis.
It is suggested that the aliquots be sent to an HHS-certified laboratory for reanalysis.
Further, it is recom-mended that the laboratory be requested to evaluate its current opiate hydrolysis procedure and extraction process with samples containing known standard amounts of morphine and morphine glucuronide to determine recovery efficiency and precision for each step in the process.
Identification of the step which produces erratic morphine recovery would allow the laboratory to adjust the procedure to produce a robust assay that produces consistent morphine quantitation results.
The two Seabrook samples reviewed in this inspection are identified by the SBCL accession numbers 031332B (reported negative) and 976221A (reported positive).
It is our judgement that the results reported here do not constitute unsatisfactory performance under the NRC and HHS regulations. Laboratories j
are permitted an overall 10% false negative rate, but no more than 50% fa]se i
negatives for any single analyte. Through 1994, the laboratory has maintained an overall score of 98% (166/170) and a score of 71% for morphine (5/7). The other two false negative results were for cannabinoids, which was also 71%
(5/7).
Prepared for:
North Atlantic Energy Service Corporation by:
h Date:
AdM 2b 8f7
/
Duo Research Inc.