Commutability of the first world health organization international standard for human cytomegalovirus

Abstract

Quantitative detection of cytomegalovirus (CMV) DNA has become a standard part of care for many groups of immunocompromised patients; recent development of the firstWHOinternational standard for human CMV DNA has raised hopes of reducing interlaboratory variability of results. Commutability of reference material has been shown to be necessary if such material is to reduce variability among laboratories. Here we evaluated the commutability of theWHOstandard using 10 different real-time quantitative CMV PCR assays run by eight different laboratories. Test panels, including aliquots of 50 patient samples (40 positive samples and 10 negative samples) and lyophilized CMV standard, were run, with each testing center using its own quantitative calibrators, reagents, and nucleic acid extraction methods. Commutability was assessed both on a pairwise basis and over the entire group of assays, using linear regression and correspondence analyses. Commutability of theWHOmaterial differed among the tests that were evaluated, and these differences appeared to vary depending on the method of statistical analysis used and the cohort of assays included in the analysis. Depending on the methodology used, theWHOmaterial showed poor or absent commutability with up to 50% of assays. Determination of commutability may require a multifaceted approach; the lack of commutability seen when using theWHOstandard with several of the assays here suggests that further work is needed to bring us toward true consensus.