Related Experiment Videos
Use of alternative rules (other than the 1(2)s) for evaluating interlaboratory performance data
S S Ehrmeyer1, R H Laessig, K Schell
1School of Allied Health Professions, Medical Technology Program, University of Wisconsin-Madison 53706.
Clinical Chemistry
|February 1, 1988
Summary
Traditional proficiency testing (PT) rules like the 1(2)s rule are ineffective. Computer simulations show that selecting PT evaluation criteria based on the population standard deviation (SD) optimizes the detection of laboratory performance issues.
Area of Science:
- Clinical Chemistry
- Laboratory Medicine
- Analytical Chemistry
Background:
- Traditional methods for evaluating interlaboratory proficiency testing (PT) data, such as the group mean +/- 2 standard deviations (SD) and the 1(2)s rule, have demonstrated ineffectiveness.
- These established criteria often fail to accurately identify acceptable and deficient laboratory performance.
Purpose of the Study:
- To evaluate the efficiency of 244 alternative rules for analyzing PT data.
- To compare these alternatives against the traditional 1(2)s rule in identifying both good and deficient intralaboratory performance.
- To determine the optimal criteria for PT evaluations based on population characteristics.
Main Methods:
- Computer simulations were employed to generate PT data.
- 244 alternative rules, all derived from the PT population's mean and SD, were assessed.
- The efficiency of each rule was examined across a range of interlaboratory SDs (1% to 10% of the population mean).
- The rules were evaluated based on their ability to correctly classify intralaboratory performance using data from one to five PT samples analyzed concurrently.
Main Results:
- All evaluated rules demonstrated maximum efficiency within a narrow range of interlaboratory SDs.
- The effectiveness of performance criteria is significantly influenced by the population SD.
- No single rule consistently outperformed others across all tested SD ranges.
Conclusions:
- The selection of appropriate criteria for acceptable performance in PT programs is critically dependent on the specific SD of the PT population.
- Optimizing the effectiveness of PT evaluations requires tailoring the chosen rule to the observed interlaboratory variability.