A theoretical basis for clinically relevant proficiency testing evaluation limits. Sensitivity analysis of the effect of inherent test variability on acceptable method error.

Abstract

The types and performance characteristics of the present rules for acceptance and rejection of quantitative results by proficiency testing agencies are reviewed. The rates of false acceptance and rejection by peer group evaluation are high when compared with explicit medical usefulness criteria. Accuracy-based target values and fixed limits provide the technologic basis for process control of interlaboratory accuracy, and their efficacy contrasts with that of peer limits. However, uniform criteria for result accuracy are also required. A clinical statistical model predicting the effect of method error on clinical diagnostic accuracy is proposed. A criterion is proposed for result accuracy defined in terms of uniform clinical benefit. The criterion specifies a fixed error limit for each type of application of a medical test that is appropriate for the inherent variability of the test application. Recent College of American Pathologists Survey experience with clinical fixed limits for allowable analytic error calculated by the model is presented.