Statistical process control tools for monitoring ELISA quality control results

Introduction: Quality Control (QC) is responsible for evaluating product quality, including performance, as described in its technical documentation.In ELISA assays, QC faces challenges such as variability in optical density (OD), well-to-well inconsistencies, and differences in plate coating.Statistical process control (SPC) is a tool that demonstrates random and common variations in production, identifying outof-control points and trends and enabling corrective actions to maintain expected product quality while reducing costs.The use of SPC in the analysis of controls and serum panels, combined with standardized analytical tools, is essential to distinguish true assay deviations from methodological errors.Objectives: To apply SPC using ELISA QC results as a model to drive continuous improvement implementation.Methodology: OD values of 40 plates of two ELISA's batches incubated with positive controls (PC), negative controls (NC), negative samples (NS), positive samples (PS) and without serum (WS) were evaluated.Descriptive and dispersion analyses, which consisted of calculating the mean, variance, standard deviation, and coefficient of variation (CV) were conducted.In addition, normality and the presence of outliers were assessed using the Anderson Darling/Shapiro Wilk tests and Grubbs' test, respectively.The performance of the plates was subsequently monitored through X-bar control charts (CC).Results: CV shown high consistency, with most groups ranging from 4.5-12% and within the 30% acceptability threshold indicating good technical capability.Two groups exceeded this limit (37.9% and 115.4%), representing isolated variability that did not compromise overall reliability.Control chart analyses confirmed the process is under control, with mean and acceptable limits generally maintained and only occasional Shewhart rule violations identified.PC showed isolated out of control points in batch 1. WS and NC displayed only sporadic deviations, with most OD measurements within control limits.Grubbs' test identified a small number of outliers, mainly in the NC and WS.Their recurrence suggests a potential source of variability, although overall assay consistency remained unaffected. Conclusion:CC indicated the process is under control and outliers' identification was crucial to understand possible variation sources and to avoid extreme values compromising the results' interpretation.Both tools provide information to support continuous improvement, such as operational adjustments to ensure reliability and analytical process reproducibility, thereby supporting product quality and reliability.