Multicenter evaluation of an enzymatic method for glycated albumin

R. Paleari, G. Bonetti, C. Callà, M. Carta, F. Ceriotti, N. Di Gaetano, M. Ferri, E. Guerra, G. Lavalle, C.L. Cascio, F.G. Martino, M. Montagnana, M. Moretti, G. Santini, D. Scribano, R. Testa, A. Vero, A. Mosca

Research output: Contribution to journalArticlepeer-review


Background The use of glycated albumin (GA) has been proposed as an additional glycemic control marker particularly useful in intermediate-term monitoring and in situation when HbA1c test is not reliable. Methods We have performed the first multicenter evaluation of the analytical performance of the enzymatic method quantILab Glycated Albumin assay implemented on the most widely used clinical chemistry analyzers (i.e. Abbott Architect C8000, Beckman Coulter AU 480 and 680, Roche Cobas C6000, Siemens ADVIA 2400 and 2400 XPT). Results The repeatability of the GA measurement (expressed as CV, %) implemented in the participating centers ranged between 0.9% and 1.2%. The within-laboratory CVs ranged between 1.2% and 1.6%. A good alignment between laboratories was found, with correlation coefficients from 0.996 to 0.998. Linearity was confirmed in the range from 7.6 to 84.7%. Conclusion The new enzymatic method for glycated albumin evaluated by our investigation is suitable for clinical use. © 2017 Elsevier B.V.
Original languageUndefined/Unknown
Pages (from-to)81-86
Number of pages6
JournalClinica Chimica Acta
Publication statusPublished - 2017


  • glycosylated albumin
  • enzyme
  • glycosylated serum albumin
  • serum albumin, Article
  • chemical analyzer
  • colorimetry
  • correlation coefficient
  • enzymatic assay
  • human
  • intermethod comparison
  • measurement repeatability
  • protein determination
  • blood analysis
  • clinical trial
  • evaluation study
  • metabolism
  • multicenter study
  • procedures
  • reproducibility
  • statistical model, Blood Chemical Analysis
  • Enzymes
  • Humans
  • Linear Models
  • Reproducibility of Results
  • Serum Albumin

Cite this