The Clinical and Laboratory Standards Institute (CLSI) has published a new document called How to Construct and Interpret an Error Grid for Diagnostic Assays; Approved Guideline. Document EP27-A provides guidance on constructing error grids, locating the error grid zones, and estimating proportions of differences in results. It contains illustrated examples of these concepts.
An error grid demonstrates the relationship between results obtained by one quantitative test to those obtained by a second one, while considering the diagnostic or therapeutic implications of the magnitude of the difference between the two results. Error grids inform users about the performance required to prevent potential patient harm. Once constructed, error grids can be populated with data from a measurement procedure comparison experiment. Then, one can calculate the proportion of data in each error grid zone as well as confidence intervals.
The focus of this document is to improve error grid methodologies by keeping the patient in mind.
Says EP27 Chairholder S. Nandagopalan, PhD: “The error grid approach is a simple, nonparametric graphical tool that evaluates measurement error in terms of its clinical consequence, or potential for patient harm. It is widely used for evaluating the performance of blood glucose monitoring systems, and it is hoped that EP27 will promote the use of this approach for other diagnostic quantitative assays as well.” This document is intended for use by developers of measurement procedures, including laboratory-developed tests, and by clinical laboratories, and can be purchased through the CLSI website. Preview sample pages.
