Regression-Adjusted Real-Time Quality Control.

Area of Science:

• Clinical Laboratory Science
• Medical Diagnostics
• Quality Management Systems

Background:

• Patient-based real-time quality control (PBRTQC) is increasingly adopted in clinical laboratories.
• Concerns exist regarding PBRTQC's performance and applicability for certain analytes.
• An extended method, regression-adjusted real-time quality control (RARTQC), is proposed to enhance real-time quality control protocols.

Purpose of the Study:

• To introduce and evaluate the regression-adjusted real-time quality control (RARTQC) framework.
• To compare the performance of RARTQC against patient-based real-time quality control (PBRTQC).
• To assess the effectiveness of RARTQC in detecting analytical errors.

Main Methods:

• Implemented RARTQC with an added regression adjustment step before statistical process control algorithms.
• Utilized patient test results for 4 analytes from Zhongshan Hospital in 2019.
• Introduced constant, random, and proportional analytical errors to compare PBRTQC and RARTQC performance using false alarm rates and error detection charts.

Main Results:

• RARTQC demonstrated superior performance compared to PBRTQC.
• RARTQC improved the trimmed average number of patients affected before detection (tANPed) by approximately 50% for constant and proportional errors.
• The study assessed error detection capabilities and false alarm rates for both protocols.

Conclusions:

• The regression adjustment in RARTQC effectively removes autocorrelation and enhances data transformation.
• RARTQC offers a powerful framework for advancing real-time quality control research.
• The RARTQC method shows significant improvements in detecting analytical errors, reducing patient impact.