A Simple Mixed-Effects Modeling for the Precision of Automated Differential Cell Counters Using R

Abstract

To register their device, manufacturers of automated differential cell counters (ADCC) are required to demonstrate satisfactory precision that meets regulatory standards such as those in 21 CFR 864.5220 in the United States. Mixed-effects models can be used for the estimation of variance components from a range of factors (operators, days, reagent lots, etc.) to characterize the “within-laboratory” or “across laboratories” analytical error associated with ADCCs for a variety of samples representing a plethora of measurand intervals. However, this task can be daunting and time-consuming due to the increasing complexity of the number of blood-associated parameters reported by those devices and the iterative calculations over numerous factors. In this paper, we propose a simple-to-follow R algorithm that overcomes these challenges and provides a comprehensive and effective estimation of variance components in a regulatory-ready reporting format.