Performance evaluation of the MBio Diagnostics point-of-care CD4 counter

The measurement of the absolute CD4 T-cell count is critical in the initial evaluation and staging of HIV-infected persons, yet access to this technology remains limited in many low resource settings where disease burden is highest. Here we evaluate the performance of a prototype point-of-care device (POC) to quantify CD4 T cells from MBio Diagnostics, Inc. Whole blood samples, both venous and capillary (ringer stick), were collected from known HIV-infected participants at the University of California, San Diego Antiviral Research Center, and tested using the MBio system and conventional flow cytometry. A total of 94 venipuncture and 52 capillary samples were processed and statistical analyses included comparison to flow cytometry results. For the venipuncture samples, Bland-Altman analysis resulted in a mean bias of -10 cells/mu L (-23 to +3 cells/mu L, 95% CI), and limits of agreement (LOA) of -132 and +112 cells/mu L. For the capillary samples, Bland-Altman resulted in a mean bias of -4 cells/mu L (-31 to +23 cells/mu L, 95% CI), and LOA of -195 and +186 cells/mu L. For the San Diego study cohort, the prototype MBio system showed negligible quantitative bias relative to flow cytometry. Higher variability was observed in the capillary samples relative to venipuncture, but system precision for both capillary and venipuncture samples was good. There was also close agreement between results from the same participant when tested with two different systems, different operators and different locations. This preliminary evaluation suggests that the MBio CD4 device holds promise as a POC system for quantitation of CD4 T cells in limited-resource settings. (C) 2012 Elsevier B.V. All rights reserved.