Vendor-Recommended Protocol Is Not Adequate To Verify Analyzer Performance After Probe Replacement

To verify performance after probe replacement on Siemens ADVIA® 1800 analyzer, the vendor recommends running 2 levels of quality control (QC) for each analyte; however, this approach provides only a snapshot of analyzer performance. Additional experiments were conducted to determine if the vendor-recommended protocol adequately verifies analyzer performance. Serum from 20 random patients was mixed, centrifuged at 1000 g for 10 min, and supernatant prepared (PPS). Intra-assay precision was assessed by analyzing 20 replicates of 2 levels of QC, and PPS for the following indicator assays (total allowable error [TEa] in parentheses): alanine aminotransferase (20%), calcium (6%), creatinine (15%), immunoglobulin G (25%), and sodium (5 mEq/L). Inter-assay precision over past 30 days was calculated from the values (90 – 114) of 2 levels of QC. Violation of 13s QC rule is most commonly used for detecting critical error (error > TEa). We used TEa/4 and TEa/3 as the acceptability criteria for intra- and inter-assay precision. Inter-assay precision was also compared to the precision claim by the vendor. Bias between the mean of QC values from intra- and inter- assay precision was calculated and TEa/4 used as the acceptability criterion. Intra-assay precision for both levels of QC for all analytes was acceptable. Intra-assay precision (5.48%) for only alanine aminotransferase in the PPS was unacceptable, indicating that patient matrix may impact precision for certain analytes. Inter-assay precision (2.24%) for level 1 QC for calcium was unacceptable. For all analytes, inter- assay precision compared to vendor’s claim was found to be significantly higher (32% - 405%). The bias was unacceptable for level 3 QC for sodium (-2.11). Increased imprecision and/or bias may be observed if either probe replacement or alignment was not appropriate; or the verification protocol was not adequate. This study demonstrates the need to conduct experiments beyond vendor-recommended protocol to ensure production of clinically acceptable results.