Number of patient samples affected before error detection: Strategic implications for internal quality control and...

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Quality Control

Background:

• Internal quality control (IQC) infrequently performed risks missed errors.
• Patient-based quality control (PBQC) is often perceived to require high laboratory volumes for effective error detection.

Purpose of the Study:

• To evaluate the performance of quality control (QC) practices using the "average number of patient samples affected before error detection" (ANPed) metric.
• To compare the relative effectiveness of various IQC and PBQC settings in error detection.

Main Methods:

• Simulations were conducted using routine sodium and aspartate aminotransferase (AST) assays.
• The ANPed metric was employed to assess error detection capabilities across different QC scenarios.

Main Results:

• Higher ANPed values were observed with fewer IQC samples per run or more patient samples between IQC runs.
• PBQC models for sodium and AST demonstrated lower ANPed than IQC performed every 100 samples, barring small systematic errors in AST.

Conclusions:

• The ANPed metric effectively illustrates the impact of IQC frequency and the number of IQC levels.
• PBQC can be superior to IQC, even in low-volume laboratories.
• ANPed enables the design of tailored QC strategies aligned with specific risk tolerance.