c Given recent advances in the development of quantitative standards, particularly WHO international standards, efforts to better understand the commutability of reference materials have been made. Existing approaches in evaluating commutability include prediction intervals and correspondence analysis; however, the results obtained from existing approaches may be ambiguous. We have developed a "deviation-from-ideal" (DFI) approach to evaluate commutability of standards and applied it to the assessment of Epstein-Bar virus (EBV) load testing in four quantitative PCR assays, treating digital PCR as a reference assay. We then discuss advantages and limitations of the DFI approach as well as experimental design to best evaluate the commutability of an assay in practice.
Quantification of viral load is integral to clinical care, particularly among immunocompromised patients (1-4). Increases in viral load may trigger preemptive therapy to prevent or treat systemic viral disease. It may also be used to monitor therapeutic response and to determine treatment endpoints. A wide variety of assays are used for these determinations (few FDA approved), and the field continues to be hindered by a lack of agreement among the results of these tests (5-8). In turn, this lack of agreement (particularly between laboratories) prevents both the development of standardized therapeutic breakpoints and the portability of patient results among institutions. Numerous factors have been shown to contribute to the variability of results (9), one of which is a lack of standardized calibration standards. The availability of WHO international quantitative standards for cytomegalovirus (CMV) and, more recently, Epstein-Barr virus (EBV) should help the process of developing such standardization; however, numerous issues remain. These issues include the development of reliable secondary standards, which are traceable to and accurately representative of the WHO standard material. As has been shown for CMV secondary standards, such trueness cannot be assumed (10).Another issue of concern is the commutability of reference materials, which focuses on whether standards behave like patient samples and are consistent across different assays. Commutability has been defined as "the equivalence of the mathematical relationships among the results of different measurement procedures for a reference material and for representative samples of the type intended to be measured" (11). First defined and accepted in the field of clinical chemistry, more recent work in clinical molecular virology demonstrated its importance to interassay agreement, with commutable standards improving and noncommutable standards diminishing quantitative agreement (12-15). Currently, commutability is typically assessed via prediction interval approaches (14) or by multivariate techniques such as correspondence analysis (16), both based on evaluating whether a reference material belongs to the same distribution obtained from measures of human specimens using two or more assays. Predicti...