Precision of autoantibody assays in clinical diagnostic laboratories: What is the reality?

Marie Senant, Lucile Musset, Georges Chyderiotis, Laurence Guis-Cabanne, Jan Damoiseaux, Nicole Fabien, Marie-Agnes Dragon-Durey*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Web of Science)
63 Downloads (Pure)

Abstract

Background: ISO 15189 accreditation remains a challenge for specialized laboratories. In the field of autoimmunity, beside the crucial problem of absence of standardization, laboratories have to manage the analytical performances of the large panel of assays in terms of sensitivity and specificity, but also on their measurement precision for which no reference values are available on biorepositories.

Methods: As an initiative of the French EASI (European Autoimmunity Standardization Initiative) group, French clinical diagnostic laboratories were requested to participate in a survey aiming to analyze the coefficients of variation (CVs) of intra-run and inter-run variability obtained with assays quantifying 14 different autoantibodies. Two performance goals corresponding to the 90th percentile and the 50th percentile (lowest CV values reached by 90% and 50% of laboratories respectively) defined for three levels of concentration were calculated. The impact on the assay performances of the number of measurements, of the nature of the internal quality control (IQC) and the type of immunoassay, was also analyzed.

Results: 414 and 616 values of intra-run and inter-run CVs were collected, respectively. The 50th percentile performance goals were comprised between 1.0% and 8.9% for the intra-run CVs, and between 1.8% and 14.6% for the inter-run CVs. At 90th percentile, the performance goals were comprised between 3.2% and 13.5% for the intra-run CVs, and between 7.3% and 30.8% for the inter-run CVs. CVs calculated from 10 values were similar to those obtained from more values. Higher imprecision was observed when the antibody levels of the IQC was lower than 2 fold the positive threshold. Commercial IQCs gave lower CVs than IQCs derived from patient samples.

Conclusion: Our results allow proposing some acceptability limits for the precision performances of the autoantibody assays, compatible with the reality of life in diagnostic laboratories and clinical care.

Original languageEnglish
Pages (from-to)57-64
Number of pages8
JournalClinical Biochemistry
Volume83
DOIs
Publication statusPublished - Sep 2020

Keywords

  • Autoantibody
  • Laboratory accreditation
  • Coefficient of variation
  • Analytic performances
  • Internal quality control
  • BIOLOGICAL VARIATION
  • CLASSIFICATION CRITERIA
  • QUALITY-CONTROL
  • LIMITS

Cite this