B-143 Using Big Data Analytics to Measure the Multi-year Consistency of Sigma Metrics for Vitros Oncology Assays

J Miller
DOI: https://doi.org/10.1093/clinchem/hvae106.504
IF: 12.114
2024-10-27
Clinical Chemistry
Abstract:BackgroundSigma metrics assess lab quality by incorporating precision, accuracy and total error allowed (TEa). Sigma metrics report quality on a 0-6 scale, with 6 representing World-Class quality. Due to logistical constraints, Sigma metrics are often calculated with limitations such as a small number of analyzers, single reagent lot or short duration. QuidelOrtho developed a novel approach using big data analytics for the objective, automated calculation of Sigma metrics across thousands of Vitros analyzers running in real lab conditions. This approach was previously used to report the Sigma metric performance for 115 assays using data from 2022. Of particular interest is the consistency of quality levels for Vitros oncology assays assessed in the study. The aim of this analysis was to examine Sigma metrics for Vitros oncology assays and compare Sigma metrics to data from 2022.MethodsThe methods follow the previous approach using the same Python algorithm to extract worldwide QC data from Jan. to Nov. 2023 (inclusive). A QC fluid lot from the study period was selected for each analyte to align with the fluid level reported in the 2022 study, which was based on medical relevance. Metrics were calculated for each set of QC data that originated from a single analyzer, reagent lot and calibration curve; the median Sigma metric is reported in the Results. Proficiency testing limits were utilized as TEa limits where available, with preference to CLIA 2024 limits.ResultsThe following assays reported Sigma metrics as >6 using both 2022 and 2023 data: alpha-fetoprotein, cancer antigen 125, carcinoembryonic antigen, prostate-specific antigen (PSA), free PSA, IgA, and IgG. Assays with higher metrics using 2023 data: cancer antigen 19-9 (5.0 vs >6) and IgM (5.3 vs >6). Assays with lower metrics using 2023 data: cancer antigen 15-3 (4.8 vs 4.4).ConclusionsVitros oncology assays report similar performance for 2022 and 2023. All Vitros oncology assays assessed report Sigma metrics of 4 or greater, indicating they achieve results within the corresponding allowable error.
medical laboratory technology
What problem does this paper attempt to address?