Assessment of split-sample proficiency testing for cholesterol by use of a computer simulation model.

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Biostatistics

Background:

• The National Cholesterol Education Program set performance goals for cholesterol measurement in 1992.
• Accurate cholesterol measurement is vital for patient care and disease management.
• Assessing laboratory performance requires robust methods to account for measurement errors.

Purpose of the Study:

• To develop and validate a computer model for assessing clinical laboratory cholesterol measurement performance.
• To evaluate the impact of various factors on the accuracy of laboratory performance classification.
• To ensure compliance with established cholesterol measurement standards.

Main Methods:

• Utilized Monte Carlo simulation techniques to model cholesterol measurements with systematic and random errors.
• Employed linear regression analysis on split-sample data from clinical and reference laboratories.
• Quantified the influence of laboratory bias, imprecision, reference laboratory error, sample size, and decision limits.

Main Results:

• The model accurately classifies laboratory performance as acceptable or deficient based on predefined decision limits.
• Laboratory bias and imprecision significantly impact performance classification accuracy.
• Reference laboratory imprecision, number of split samples, and chosen decision limits critically affect classification reliability.

Conclusions:

• Computer modeling provides a powerful tool for assessing cholesterol measurement performance.
• Careful selection of decision limits and adequate sample sizes are essential for accurate laboratory performance classification.
• The developed model aids in ensuring reliable cholesterol testing and adherence to national standards.