Westgard Rules
A multi-rule QC system combining statistical rules to detect analytical errors while minimizing false rejections in laboratory quality control.
What are Westgard Rules?
- Westgard Rules
- Westgard Rules are a set of statistical quality control rules developed by Dr. James Westgard to help laboratories detect analytical errors in their testing processes. The multi-rule approach uses a combination of rules to maximize error detection while minimizing false rejections, providing better performance than single-rule systems.
The Westgard Multi-Rule System
Warning Rule
One control exceeds the mean ± 2 standard deviations. This triggers inspection of the run using other rules but is not a rejection rule by itself.
Warning OnlyRandom Error Rejection
One control exceeds the mean ± 3 standard deviations. Indicates a random error has occurred.
Reject RunSystematic Error Detection
Two consecutive controls exceed the same mean + 2s or mean - 2s limit. Indicates systematic error (shift).
Reject RunRange Error
One control exceeds mean + 2s while another exceeds mean - 2s (range greater than 4s). Indicates random error.
Reject RunTrend Detection
Four consecutive controls exceed the same mean + 1s or mean - 1s limit. Indicates systematic error (trend).
Reject RunShift Detection
Ten consecutive controls fall on one side of the mean. Indicates systematic error (major shift).
Reject RunError Type Detection
Random Errors
Detected by: 1:3s, R:4s
- Sample handling issues
- Pipetting errors
- Air bubbles
- Electrical interference
Systematic Errors
Detected by: 2:2s, 4:1s, 10x
- Calibration drift
- Reagent degradation
- Temperature changes
- Lot-to-lot variation
LIMS Implementation
Modern LIMS and middleware systems can automatically apply Westgard rules to QC results, generate Levey-Jennings charts, and trigger alerts when rules are violated. This automation enables real-time QC monitoring and immediate response to quality issues.