Example 2:  Bad Assay, Good Control

Figure 1.  An excerpt from the Summary Page of the Assay Report shows the Statistical Quality Control Evaluation: "INSPECT ASSAY".  (More information on this table and Method Controls (MC) and Subpop. Distributions (SP) can be found in the Assay Report's Summary Page in the Operator’s Guide and the Glossary.)

Identify any flagged QC parameters (outliers).

Figure 2.  The outlier table from the Summary Page of the Assay Report shows the flagged parameters in red. (More information about the outlier table can be found on the Assay Report's Summary Page in the Operator’s Guide.)

Figure 3.  An example of the Statistical Tables from the Assay Report shows the statistical comparison results for the assay parameters.  Note that the flagged parameters on the statistical tables (which appear in red) are also listed on the outlier table on the Summary Page (Figure 2).

Figure 4.  The modified Levey-Jennings chart shows the 4.7ng/ml standard responses (adjusted and normalized) for the last thirty assays.  Note that both responses are flagged for assay number 32 and are listed in the outlier table (Figure 2.).  (More information about adjusted responses and normalized responses can be found in the Glossary.  More information about the charts can be found in the Assay Report's Graphs section in the Operator’s Guide.)

To identify the failed component, match the flagged QC parameters with the flagged parameters under each component listed in the top section of the StatLIA troubleshooting chart (see below).

Figure 5.  The illustration shows the process involved in troubleshooting an assay flagged by StatLIA.  In this example, the Std-Adj ABS: 4.7ng/ml   (Standard Adjusted Response) and Std-%B/St6: 4.7ng/ml (Standard Normalized Response) on the report is matched to the flagged parameters on the chart (red circle) in step A.  Note that the absence of any other flagged parameter confirms that this column is the only one that matches.  The failed component is determined by reading the component header from this column in step B, which in this case is the Standards (4.7ng/ml).

The parameters in the bottom section of the chart, such as control specimen concentrations, ED20, and minimum detectable concentration, are used to gauge the impact of the failed reagent or incubation step on the unknown concentrations.

Figure 6.  The illustration shows the final step in the troubleshooting process.  The gauging parameters, in this case the Parallel Probability, ED20 (Estimated Dose at 20% Binding) and Min DC (Minimum Detectable Dose) are flagged.  These flagged parameters indicate that the standard curve is not correct (at the low concentration end) and that the unknown results are not accurate.