Rationale for using insensitive quality control rules for today's hematology analyzers

Abstract

Diverse approaches have been used to assure the analytical quality of automated hematology; as such, there is great variation in their error detection capabilities. We summarize the intralaboratory performance of a cohort of Sysmex XE-2100's running e-Check hematology quality control (QC). The imprecisions of a median performing (50th percentile imprecision) and more imprecise [15th percentile (15P) imprecision] Sysmex XE-2100 are compared with measures of total allowable error (regulatory and physiologically based) to obtain multiples of the usual imprecision that must be detected to prevent the hematology analyzer from producing medically unacceptable results. The resultant large multiples of the usual imprecision (s) demonstrate the need for insensitive QC rules employing very broad control ranges, control rules that have been implicitly supported by hematology analyzer manufacturers for the last several decades. For today's highly precise hematology analyzers, the following control rules are strongly advised: 13.5s, 14s and 1 4.5s rules (violated if a single control observation exceeds either its ±3.5, ±4.0 and ±4.5s limits, respectively). In order for the hematology laboratory to totally embrace expanded QC limits, manufacturers must make available their instruments' usual and poorer (e.g. the 15P performance) imprecision's. Users of hematology analyzers that require more sensitive but less specific rules to prevent the reporting of clinically erroneous data are advised to acquire more precise (and thus more dependable) instrumentation. © 2010 Blackwell Publishing Ltd.