Application Process: Program-Specific PT Sample Grading Criteria

Wisconsin has several specific rules for grading reference sample results, which are described below:

• The "50% Rule" (multiple concentration levels or matrices)

  For each analyte or analyte group, laboratories need to pass 50% of the concentration levels and matrices attempted in a particular study to receive an evaluation of "Acceptable".

• WP Metals Many WP studies for metals contain multiple PT samples at different concentration levels. If you analyze both concentration levels for a PT study, you must pass greaster than 50% of the concentration levels (or both of them) to receive an evaluation of "Pass".


• PCBs Many WP studies for PCBs offer two matrices (PCBs in water, PCBs in oil) and two concentration levels per matrix.
  If lab X analyzes either ONLY the water PTs or ONLY the oil PTs, it must pass BOTH concentration levels if both are reported (the 50% rule).
  If lab X analyzes BOTH the water PTs AND the oil PTs, it must pass 3 of the 4 combined concentration levels if both concentration levels are reported for each matrix (the 50% rule).
  If lab X analyzes BOTH the water PTs AND the oil PTs, it must pass both matrices if only a single concentration level is reported for each matrix (the 50% rule). (the 50% rule).
  




• The "80% Rule" (multiple analytes)

  Laboratories need to pass 80% of the analytes present in (spiked into) a multi-analyte (GC, GC/MS, HPLC) reference sample to receive an evaluation of "Acceptable".

• False positives (reporting a result for an analyte which was not spiked) count as an analyte failure towards 80%.


• False negatives (e.g. "<" X ) where the true value is greater than X count as a failure towards 80%.

If you report "< (LOD)" for any analyte present at a concentration greater than (LOD), your result will be scored as a failure..
• Unreported analytes(for analytes present in the sample) count as a fail towards 80%.
• Unspiked analytes ARE NOT CONSIDERED in the accounting towards 80%.
• Undetected analytes (e.g. "<" X ) where the true value is LESS than X are NOT CONSIDERED in the accounting towards 80%.


• Multiple concentration levels for a given analyte(or analyte group).
  

  If multiple concentration levels are available for an analyte/analyte group within a partiular study --and the laboratory chooses to report results for both concentration levels-- the laboratory must submit acceptable results on greater than 50% of the concentration levels for the cumulative results of the study to be considered acceptable. This is called the “50% Rule.”

  Example: ACME Laboratories chooses to report results for arsenic in both the APG "WP" level sample, and also for the "APG+" concentration level. ACME receives "Acceptable" results on the "WP" level and "Not Acceptabler" results on the "APG+" level.
  ACME Labs will receive a score of "F" (Fail) for this PT study. This result may have implications on either annual renewal, or enforcement action related to PT sample performance.

• Reporting results from multiple methods for a single PT sample.

  For non-drinking water (WS) PT samples, when a laboratory employs multiple methods to analyze a reference sample or reference sample group, at least ONE of the results must fall within acceptance criteria.

  Example: ACME Laboratories chooses to report results for cadmium under category 08 (Metals I) by ICP method 200.7 (Not Acceptable), by ICP-MS method 200.8 (Not Acceptable) and also by graphite furnace AA by method 200.9 (Acceptable). ACME Labs will receive a score of "P" (Pass) for this PT study in our system. These results are acceptable for certification renewal of cadmium in test category 08.



• Multiple analytes and multiple levels.

  When a reference sample study for a multi-analyte group includes multiple ampules with each one containing a different concentration level, the results reported for each level are first graded separately using the 80% rule. Once the acceptability of each level has been determined, the 50% Rule is applied to determine the overall acceptability of the study. The acceptability of the study is not determined by the average acceptability rate of the multiple levels.

  Example: Consider the case of a laboratory that submits results for a two-level VOC reference sample study. If 92% of the analytes in one study level are reported correctly, but only 70% of the analytes in the other level are reported correctly, the overall study result is unacceptable because only 50% of the individual levels satisfy the 80% rule. The 81% average acceptability rate between the two levels does not influence the acceptability of the study.