A method for optimization and validation of moving average as continuous analytical quality control instrument...

Area of Science:

• Clinical Chemistry
• Analytical Chemistry
• Laboratory Medicine

Background:

• Moving average (MA) methods are used for quality control in laboratory diagnostics.
• Accurate bias detection is crucial for reliable creatinine measurements.
• Existing methods may lack realistic validation for MA bias detection.

Purpose of the Study:

• To develop and validate a moving average (MA) optimization method for creatinine bias detection.
• To achieve reproducible and fast bias detection with minimal false alarms.
• To identify an optimal MA procedure for continuous analytical quality control.

Main Methods:

• Investigated multiple MA procedures using power function analysis and simulation.
• Simulated MA bias detection using reported consecutive results for creatinine.
• Selected an optimal MA procedure based on detectable bias within scheduled quality control (QC) measures.

Main Results:

• Increasing batch sizes and stringent truncation limits improved MA bias detection probabilities but could delay detection.
• The optimal MA procedure enabled reproducible detection of a 20% creatinine bias within 23-145 results.
• A 40% creatinine bias was detected within 22-60 results using the optimal MA procedure.

Conclusions:

• Simulating MA bias detection provides a more realistic assessment than power function analysis alone.
• The developed method is valuable for implementing MA as a continuous analytical quality control instrument.
• This approach enhances the reliability and efficiency of quality control in creatinine testing.