Related Experiment Videos
Precision requirements for cost-effective operation of analytical processes
1Department of Pathology, University of Wisconsin, Madison 53792.
Clinical Chemistry
|September 1, 1990
Summary
Current analytical method performance criteria are inadequate for detecting medically significant errors. A revised standard requiring at least 4 times the standard deviation (s) to be less than the allowable total error (TEa) is proposed for improved quality control.
Area of Science:
- Clinical Chemistry
- Analytical Chemistry
- Laboratory Medicine
Background:
- Current performance criteria for analytical methods are outdated.
- The common criterion (2s < TEa) is insufficient for reliable error detection.
Purpose of the Study:
- To evaluate the adequacy of current analytical method performance criteria.
- To propose an improved criterion for quality control.
Main Methods:
- Analysis of statistical power functions for quality control procedures.
- Reverse engineering of performance criteria based on error detection sensitivity.
Main Results:
- Methods minimally meeting the 2s < TEa criterion often fail to detect medically important errors.
- Medically important errors require at least 3s to be reliably detected by current procedures.
- The 2s < TEa criterion lacks sensitivity due to inherent limitations of control procedures.
Conclusions:
- The current 2s < TEa criterion is inadequate for ensuring reliable detection of medically important errors.
- A revised criterion of at least 4s < TEa is recommended for improved analytical method performance.
- This enhanced criterion aligns with industrial process capability guidelines and improves error detection reliability.