Quality control limits: Are we setting them too wide?

Area of Science:

• Clinical Laboratory Science
• Statistical Process Control

Background:

• Quality control charts (Levey Jennings Charts) rely on variation estimates.
• Clinical laboratory science (CLS) typically uses long-term variation, while other fields favor short-term variation.

Purpose of the Study:

• Compare short-term and long-term variation estimation methods.
• Evaluate the accuracy of control limits generated by each method.
• Explore the utility of adopting methods from other fields in CLS.

Main Methods:

• Literature review comparing CLS practices with other fields.
• Statistical theory to demonstrate differences between short-term and long-term variation estimates.
• Simulation studies to illustrate discrepancies in control limits.
• Analysis of laboratory data comparing both estimation methods.

Main Results:

• CLS practices significantly differ from other fields regarding variation estimation.
• Short-term variation methods are standard in other fields but rarely cited in CLS.
• Long-term variation estimates can theoretically inflate process variation.
• Analysis of 95 QC charts revealed significant differences in 93 cases, with long-term estimates often larger.

Conclusions:

• Estimates of variation derived from short-term and long-term methods can yield substantially different results.
• Long-term variation estimates in CLS are frequently larger than short-term estimates, potentially affecting quality control accuracy.