The Effect of Uncertainty Components Such as Recalibration on the Performance of Quality Control Charts
P Winkel, Nien F. Zhang
Daily QC measurements of biochemical quantities were recorded during 4 to 5 months while methods and analyser showed no signs of malfunctioning. The time series of QC values were divided into subseries according to reagents or electrolyte diluent lot and (within diluent subseries) disposable electrode used. ANOVA was used to examine if the mean level changed significantly between subseries. All time series, as well as reagents and diluent subseries were examined for autocorrelation. The X-chart and the EWMAST (in autocorrelated series) or EWMA chart were applied to each time series and each reagents and diluent subseries and the number of values falling outside the 3 SD control limits were noted. Results: The mean levels changed significantly due to diluent lot changes, replacement of disposable electrodes and recalibrations following reagents lot changes. These changes caused spurious autocorrelation as evidenced by the ACF plot. In 42% of all reagents subseries a significant autocorrelation could also be demonstrated, however; 5.64% and 29.1% of all time series values fell outside the control limits of the X-charts and the EWMA or EWMAST charts respectively. These percentages were reduced to 0.44 and 0.7 when separate control charts were calculated following recalibrations and changes of diluent lot. Conclusions: The mean level may change due to recalibrations and change of electrode diluent lot that causes an excessive number of false alarms unless new control charts are calculated subsequent to these events.
autocorrelation, control charts, EWMA charts, EWMAST charts, quality control, statistics
and Zhang, N.
The Effect of Uncertainty Components Such as Recalibration on the Performance of Quality Control Charts, Clinical Chemistry
(Accessed June 2, 2023)