CLSI updates method evaluation standard on user verification of precision and estimation of bias

The Clinical and Laboratory Standards Institute (CLSI) has published an updated standard—User Verification of Precision and Estimation of Bias; Approved Guideline—Third Edition (EP15-A3)—which describes the estimation of imprecision and of bias for clinical laboratory quantitative measurement procedures using a protocol that can be completed within five days.

Included in this edition of EP15 are guidelines for duration, experimental designs, materials, data analysis summarization, and interpretation—techniques that are adaptable for a wide range of analytes and device complexity. A balance is created in this document between the complexity of design and formulae, and the simplicity of operation. The protocol is designed to be completed within five working days based on a uniform experimental design yielding estimates of imprecision and bias.

In this edition of EP15, the user is intended to verify the manufacturer’s claim for precision, and estimate bias, because there is unlikely to be a bias claim to verify. Most manufacturers follow CLSI document EP05-A3—Evaluation of Precision of Quantitative Measurement Procedures; Approved Guideline—Third Edition to establish precision claims, and these claims are relatively easily verifiable using the approach prescribed in EP15. The document development committee chose to keep the number of days in the experiment at five, and to increase the number of replicates per day to five, in order to obtain more reliable estimates of repeatability and within-laboratory imprecision. The most complicated calculations were replaced by tables to make calculations easier and to reduce the chances of mathematical errors.

Two principal goals underlay the development of EP15. One was to develop a testing protocol that is suitable for use in the large clinical laboratory, yet simple enough to be applicable in the point-of-care or physician’s office laboratory. The second was to develop a protocol that is sufficiently rigorous to provide statistically valid conclusions for verification studies. The bias is assessed by a recovery experiment. Instead of manual worksheets, calculations may also be readily performed with method evaluation software or generic spreadsheet software.

This document is primarily intended for use when an established measurement procedure is initially set up in the laboratory. It may also be used to verify performance after corrective action following a failed proficiency testing event. Read sample pages.