Optimization and validation of moving average quality control procedures using bias detection curves and moving...

Area of Science:

• Clinical chemistry
• Analytical chemistry
• Quality control

Background:

• Lack of practical tools for optimizing moving average (MA) settings in continuous analytical quality control.
• Limited understanding of the bias detection capabilities of applied MA procedures.

Purpose of the Study:

• To introduce bias detection curves for MA optimization.
• To introduce MA validation charts for MA validation.
• To demonstrate the application of these tools using sodium, potassium, and albumin assays.

Main Methods:

• MA optimization and bias detection simulation using historical assay data.
• Generation of bias detection curves by plotting median test results needed for detection against introduced bias.
• Development of MA validation charts illustrating the range of results required for bias detection.

Main Results:

• Bias detection curves facilitate graphical comparison and selection of optimal MA settings based on bias detection properties.
• MA validation charts provide insights into the number of assay results necessary for detecting bias with selected optimal MA procedures.

Conclusions:

• Bias detection curves and MA validation charts are valuable tools for optimizing and validating MA procedures in analytical quality control.
• These graphical tools enhance the practical application of MA for continuous monitoring and bias detection.