Skip to main content

The accuracy of platelet counting in thrombocytopenic blood samples distributed by the UK National External Quality Assessment Scheme for General Haematology.

De la Salle BJ, McTaggart PN, Briggs C, Harrison P, Doré CJ, Longair I, Machin SJ, Hyde K

Publication: Am J Clin Pathol (2012) • Vol 137 • Page 65-74PMID: 22180479View on PubMed

Purpose

The purpose of this paper was to determine the effects of analyzer type on platelet counts.

Conclusion

The mean platelet counts using a combination of all methods were higher than the international reference methods (IRMs). None of the instruments yielded the same counts as IRM for more than 28% of the pooled blood specimens, and significant biases were found for 17 of the 23 tested analyzers. The coefficient of variance (CV) increased with decreasing platelet counts for all analyzers. The authors state that between site variability was greater than within machine variability, but they do not discuss any potential impacts of storage or transport.

Studies (1)

Purpose

The purpose of this study was to determine the effects of analyzer type on platelet counts. Pooled, leukocyte depleted, citrate phosphate dextrose blood was stored for less than 72 h and was partially fixed with formaldehyde and gluteraldehyde in the presence of antibiotics. A total of 23 different analyzers were included in this study. Importantly, specimens were analyzed up to 7 days apart and were transported by mail or courier prior to analysis. The effects of storage and transportation were not discussed.

Summary of Findings

The three IRMs returned slightly different platelet counts, but, with the exception of 5 specimens, all platelet counts remained within the accepted 10% cut-off of each other. The mean platelet counts using all methods were higher than the IRM counts, with 16.5% of counts significantly higher than the IRM (p<0.01), and 0.3% of counts significantly lower than IRM (p<0.01). Further, none of the instruments yielded the same counts for more than 28% of the pooled blood specimens; however no significant bias in counts were found using Beckman Coulter Ac*T 5DIFF, Beckman Coulter LH 780, Horiba ABX 60, Horiba ABX 80, Sysmex SF-3000, and Sysmex XE-2100 machines. The Abbott machines most often led to significant positive bias, with significant bias observed in as few as 5 of 29 specimens (Abbott CELL-DYN 3700) or as many as 24 of 28 specimens (Abbott CELL-DYN 3200). The CV increased with decreasing platelet counts for all analyzers. The authors state that between site variability was greater than within machine variability, but they do not discuss any potential impacts of storage or transport.
  • Biospecimens

    FluidBlood
  • Preservation Types

    Other Preservative
  • Diagnoses

    Normal
  • Platforms

    Cell count/volumeHematology/ auto analyzer
    Cell count/volumeFlow cytometry
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formaldehyde
    Glutaraldehyde
    Hematology/ auto analyzer SpecificTechnology platform
    Abbott
    Beckman Coulter
    Horiba ABX
    Siemens
    Sysmex