[Standardization of statistical procedures and evaluation scheme in external quality-control survey]

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Biostatistics

Context:

• External quality assessment (EQA) surveys in Japan lack standardized statistical processing.
• Incompatible evaluation schemes hinder inter-survey result comparison.
• Existing coding systems are limited to analytes and equipment, necessitating expansion.

Purpose:

• To propose standardized procedures for EQA surveys in Japan.
• To introduce an iterative statistical method for unbiased peer-group statistics.
• To present a unified software solution for EQA data processing.

Summary:

• Recommends systematic coding for analytical principles, manufacturers, and standard materials.
• Advocates for an iterative method to compute unbiased mean and standard deviation (SD), mitigating outlier bias.
• Suggests a common coefficient of variation (CV) evaluation scheme for fairer assessment, using within-method CV.
• Introduces SurveyMaster I & II software prototype for standardized data processing and analysis.

Impact:

• Enhances the reliability and comparability of EQA survey results across Japan.
• Improves the accuracy of statistical evaluation in laboratory quality control.
• Facilitates better standardization and reproducibility assessment for various analytes.
• Provides a foundation for unified EQA data management and analysis.