A-196 Comparative Analysis of the Impact of Total Allowable Error on the Six Sigma Metrics of Serology Assays

Abstract Background In recent years, Six Sigma Metrics has become one of the most efficient tools to monitor the analytical performance of clinical laboratory systems, since it unifies Total Allowable Error (TEa), random and systematic errors in a single index. In infectious serology testing, there is no universally accepted gold standard or reference, which complicates the establishment of performance specifications. Defining Six Sigma Metrics becomes difficult, as they rely on the total allowable error (TEa). To address this challenge, we evaluated three different approaches to estimating TEa for serology tests: (i) Internal Quality Control (IQC) performance, (ii) proficiency testing, and (iii) the cut-off value, also referred to as the allowable loss of sensitivity. This comparative study examines how each method for estimating TEa affects the Six Sigma Metrics of five serology tests (Anti-HBE, Anti-HBS, Anti-HBC, HBEAg, and HBSAg) acrosss two Roche Cobas Pro e801 modules. Methods This analysis was performed between May and August 2024. To estimate TEa based on: (i) IQC performance, the pooled coefficient of variation (CV%) was calculated and multiplied by three; for (ii) proficiency testing, TEa was determined from the median (CV%) of the peer group over the past two years; and when estimating (iii) TEa from the cut-off, the distance between the mean of the positive control and the assay*s cut-off was divided by the mean of the positive control ((cut-off – positive mean)/positive mean). A three-month dataset was compiled, assessing internal quality control to evaluate imprecision (expressed as CV%) and proficiency testing to quantify inaccuracy (as bias%) for each analyte across two modules of the Cobas Pro e801. Six Sigma metrics were then calculated for each QC level and equipment using the formula Sigma = (TEa – bias%)/CV%. Results The Six Sigma values calculated from TEa based on IQC ranged from 2.14 to 6.02, with 55.5% between 3 and 4 Sigma. For TEa based on Proficiency, Six Sigma values ranged from 1.95 to 14.93, with 55.5% below 3 Sigma. For TEa based on the cut-off, Six Sigma values ranged from 3.84 to 23.67, with 88.9% above 6 Sigma. Since Non-Reactant levels of Anti-HBS exhibited random error equal to zero (CV% = 0), Sigma Metrics could not be calculated. Conclusion TEa has a significant impact on the Six Sigma Metrics under similar conditions. For infectious serology tests, the methodology used to estimate TEa plays a crucial role in effective monitoring of analytical performance. In our laboratory, TEa based on IQC yielded moderate Six Sigma values, which were consistent across the board and allowed more efficient QC monitoring. Proficiency-based TEa showed substantial variation between Non-Reactive and Reactive Sigmas, with a larger proportion falling below 3 Sigma. This inconsistency among Sigma values makes QC monitoring more complex and demanding. In contrast, Cut-off-based TEa resulted in the highest proportion of values above 6 Sigma, though this was due to an inflated TEa value. Although this method yielded the best Sigma values, QC monitoring is less efficient, making it harder to detect performance changes.