Standard: 2003 NELAC
Link to relevant standard
The following comments and concerns are base on actual practices observed in laboratories based on possible interpretations of the NELAC standard.
Section 18.104.22.168 begins with the statement "When an initial instrument calibration is not performed on the day of analysis, the validity of the initial calibration shall be verified prior to sample analyses by continuing instrument calibration verification with each analytical batch. The following items are essential elements of continuing instrument calibration verification:". This is a forward looking statement meaning that the pass/fail status of the CCV standard being run is evaluated only in light of it’s impact on the samples which follow the CCV standard.
Section 22.214.171.124 e) reads "If the continuing instrument calibration verification results obtained are outside established acceptance criteria, corrective actions must be performed. If routine corrective action procedures fail to produce a second consecutive (immediate) calibration verification within acceptance criteria, then either the laboratory has to demonstrate acceptable performance after corrective action with two consecutive calibration verifications, or a new initial instrument calibration must be performed."
The corrective action language in the standard only address what is necessary to proceed with analysis without recalibration. I referred to this evaluation as being "forward looking". There is no interpretation given regarding any additional considerations, or limitation on corrective actions for nonconforming CCV events where they occur in the middle or the end of a sequence that requires acceptable bracketing CCVs such as in GC analysis without the use of internal standards.
The following practices have been observed in NELAC accredited laboratories:
• A laboratory routinely will set up two consecutive CCVs during an automated sequence. If the first CCV passes, the laboratory will not evaluate the second. However if the first CCV fails and the second one passes the laboratory will report all preceding and trailing samples as being bracketed by an acceptable CCV.
• In an "attended" continuous sequence it has also been observed that a laboratory will perform instrument maintenance such as changing an inlet liner etc. in between CCVs and again treat the second passing CCV as the acceptable bracketing CCV for the preceding samples.
Comment: these practices constitute the "priming" of an instrument before running a CCV, the treatment of QC samples differently from the associated samples, and the "cherry picking" of QC that passes over QC that fails. However, there is no language in the standard that clarifies the CCV evaluation regarding their potential impact on the preceding samples.
In the above cases the laboratory has argued that the NELAC standard allows for this.
It is requested that clarification be provided regarding the impact on a failing CCV on the preceding samples during a continuing sequence where acceptable bracketing CCVs are required.
TNI FINAL RESPONSE:
(Quality Systems Expert Committee/NELAP Board, 12-19-08)
Running a second CCV in a sequence is not the intention of the standard. The practice of running two CCVs routinely would require that the laboratory evaluate each of them on every occasion. There must be a form of corrective action (i.e., instrument maintenance) prior to the second CCV being evaluated. Since no corrective action is being taken between the two CCVs, the laboratory is failing the requirement in 126.96.36.199 e of performing routine corrective action (unless it can be documented that something occurred in the first CCV, such as poor sample introduction, that did not occur in the second CCV), and cannot use the second CCV to alleviate the failing of the first.