Management of Cytomegalovirus, Epstein-Barr Virus, and HIV Viral Load Quality Control Data Using Unity Real Time.

ABSTRACT

Quality control (QC) rules (Westgard rules) are applied to viral load testing to identify runs that should be reviewed or repeated, but this requires balancing the patient safety benefits of error detection with the cost and inefficiency of false rejection. In this study, we identified the total allowable errors (TEa) from the literature and utilized a commercially available software program (Unity Real Time; Bio-Rad Laboratories) to manage QC data, assess assay performance, and provide QC decision support for both FDA-approved/cleared (Abbott cytomegalovirus [CMV] and HIV viral load) as well as laboratory-developed (Epstein-Barr virus [EBV] viral load) assays. Unity Real Time was used to calculate means, standard deviations (SDs), and coefficient of variation (CV; in percent) of negative, low-positive, and high-positive control data from 73 to 83 days of testing. Sigma values were calculated to measure the test performance relative to a TEa of 0.5 log10. The sigma value of 5.06 for EBV predicts ∼230 erroneous results per million individual patient tests (0.02% frequency), whereas sigma values of >6 for CMV (11.32) and HIV (7.66) indicate <4 erroneous results per million individual patient tests. The Unity Real Time QC Design module utilized these sigma values to recommend QC rules and provided objective evidence for loosening the laboratory's existing QC rules for run acceptability, potentially reducing false rejection rates by 10-fold for the assay with the most variation (EBV viral load). This study provides a framework for laboratories, with Unity Real Time as a tool, to evaluate assay performance relative to clinical decision points and establish optimal rules for routine monitoring of molecular viral load assay performance.