Patient results and laboratory test accuracy.

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Quality Control

Background:

• Traditional quality control materials may lack commutability with patient samples, potentially failing to detect systematic errors.
• Monitoring inter-instrument bias is crucial for ensuring accurate clinical laboratory test results.

Purpose of the Study:

• To describe a method for monitoring inter-instrument bias using patient result distributions.
• To develop a bias monitoring system independent of control material commutability.

Main Methods:

• Serum calcium data from a laboratory network were analyzed.
• A reference interval was established using Hoffmann's algorithm on all available data.
• Bias was assessed by comparing observed result distributions to a zero-bias distribution derived from results outside the reference interval.

Main Results:

• Performance levels for bias were defined: minimum allowable < 1.27%, desirable < 0.85%, and optimal < 0.42%.
• Zero bias distributions were 3.92% above and 2.53% below the reference interval for calcium.
• All three participating laboratories achieved performance within the desirable allowable bias level.

Conclusions:

• Bias monitoring using patient result distributions offers a robust method for assessing systematic error between laboratory instruments.
• This approach improves laboratory quality control processes.
• Enhanced bias monitoring strengthens patient risk management by ensuring test accuracy.