Within-Individual Mean Corpuscular Volume Variation

Abstract

We read with interest “Biological Variations of Hematologic Parameters Determined by UniCel DxH 800 Hematology Analyzer.”.1 Most of the estimates of biologic variation of the common analytes reported by Zhang et al1 were close to previously reported literature values. We are puzzled, however, about the broad intraindividual variation for mean corpuscular volume (MCV; 1.12%). Although compilations of biologic variation published before 2000 indicated intraindividual MCV coefficients of variation (CVs) could exceed 1%, 2 recent studies determined the CV was much lower, either 0.18%2 or 0.34%.3 As a result, the Zhang et al reference change values for MCV are much higher than our intuitive reference change values (perhaps, 2 fL) when we examine sequential patient MCVs measured by analyzers that are much older than the analyzer used by Zhang et al.The larger intraindividual MCV CVs reported before 2000 were likely associated with the use of more-imprecise instrumentation. Today, our analytic variations in MCV measurements are miniscule.4 Reasons for a tripling or quadrupling of the estimated intraindividual CV might include the incorporation of some outlying data, analyzer noise, or even some unusual individual-specific variation. We are wondering if Zhang et al would reexamine their MCV data and correlate any available quality control results with those data. Could this variation have been due to some sporadic instrument or individual source?