Received on October 3, 2006
Accepted on January 15, 2007

Automation and Analytical Techniques

Evaluation of Assigned Value Uncertainty for Complex Calibrator Value Assignment Processes: A Prealbumin Example

¹ Department of Clinical Chemistry Development, Beckman Coulter Inc., CA
² Math/Stats Group, Beckman Coulter, Inc., CA

^* To whom correspondence should be addressed. E-mail: jsmiddleton{at}beckman.com.

Background: Errors of calibrator-assigned values lead to errors in the testing of patient samples. The ability to estimate the uncertainties of calibrator-assigned values and other variables minimizes errors in testing processes. International Organization of Standardization guidelines provide simple equations for the estimation of calibrator uncertainty with simple value-assignment processes, but other methods are needed to estimate uncertainty in complex processes.

Methods: We estimated the assigned-value uncertainty with a Monte Carlo computer simulation of a complex value assignment process, based on a formalized description of the process, with measurement parameters estimated experimentally. This method was applied to study uncertainty of a multilevel calibrator value assignment for a prealbumin immunoassay.

Results: The simulation results showed that the component of the uncertainty added by the process of value transfer from the reference material CRM 470 to the calibrator is smaller than that of the reference material itself (<0.8% vs 3.7%). Varying the process parameters in the simulation model allowed for optimizing the process, while keeping the added uncertainty small. The patient result uncertainty caused by the calibrator uncertainty was also found to be small.

Conclusions: This method of estimating uncertainty is a powerful tool that allows for estimation of calibrator uncertainty for optimization of various value assignment processes, with a reduced number of measurements and reagent costs, while satisfying the requirements to uncertainty. The new method expands and augments existing methods to allow estimation of uncertainty in complex processes.