School of Medicine’s Kroll Authors CLSI Revised Guideline

Professor of pathology and laboratory medicine at BUSM Martin Kroll, MD, recently authored the Clinical and Laboratory Standards Institute (CLSI) publication guideline, EP24-A2, Assessment of the Diagnostic Accuracy of Laboratory Tests Using Receiver Operating Characteristic Curves; Approved Guideline—Second Edition.

This document provides a protocol for evaluating the accuracy of a test to discriminate between two subclasses of subjects when there is some clinically relevant reason to separate them. In addition to the use of receiver operating characteristic (ROC) curves and the comparison of two curves, the document emphasizes the importance of defining the question, selecting the sample group, and determining the “true” clinical state.

“This guideline explains how to evaluate the usefulness of clinical laboratory tests using ROC curves. In the current resource-constrained environment, laboratories will need to use robust tools to assess the usefulness of tests, whether they are performed in-house or sent out to a reference laboratory,” explains Kroll, MD, who also serves as Chief, Department of Laboratory Medicine, Boston Medical Center and Chairholder of the document development committee. “EP24 describes how to select the data, create ROC curves, and interpret the results. This version of the guideline explains both qualitative and quantitative interpretations of ROC curves and provides an in-depth discussion of the quantitative calculations.”