Improving the Bias of Comparator Methods in Analytical Performance Assessments Through Recalibration

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Metrology

Background:

• Systematic differences (bias) between laboratory analyzers impact analytical performance studies.
• The choice of comparator method is critical due to inter-analyzer bias.
• Assessing retrospective recalibration feasibility is essential for accurate laboratory measurements.

Purpose of the Study:

• To evaluate the feasibility of retrospective recalibration to minimize bias in laboratory measurement results.
• To assess the effectiveness of using higher-order methods or materials for recalibration.
• To reduce the impact of comparator method selection in analytical performance studies.

Main Methods:

• Retrospective analysis of existing data from continuous and blood glucose monitoring systems.
• Comparison of subject samples with a higher-order method or standard reference material (SRM).
• Linear regression analysis for recalibration equations applied to complete data sets, assessing bias before and after recalibration.

Main Results:

• Recalibration reduced bias between devices from -3.6% to +0.6% and +11.0% to +0.3% in two data sets.
• Recalibration using higher-order materials reduced bias from -3.1% to -0.1% and -4.3% to -2.7% in respective data sets.
• Demonstrated a significant decrease in inter-device bias post-recalibration.

Conclusions:

• Retrospective recalibration effectively decreases bias between laboratory analyzers.
• This method can mitigate the influence of the chosen comparator method in performance studies.
• Prospective study designs are recommended for verification of the recalibration procedure.