Have you caught that outlier yet? Evaluate the utility of repeat testing in nutritional and toxic element assessment

Eric A. Smith, Rebecca Parker, Bryce Genesi, Yifei K. Yang

Research output: Contribution to journalArticlepeer-review

Abstract

Repeat testing is routinely required by regulatory bodies as a measure to rule out contamination in trace elements and heavy metal analysis, especially when the initial analysis result is outside the reference interval. However, its clinical utilities in detecting analytical measurement outliers have not been systematically evaluated in different clinical testing scenarios. In this study, we present an extensive evaluation of repeat testing and its comparison with the initial analysis in four serum and plasma trace element assays performed by inductively coupled plasma mass spectrometry. We demonstrate that the patient population distributions for these elements differ significantly from the reference interval established by healthy individuals. Accordingly, a significant proportion of the patient specimens would require repeat testing when using reference intervals as the threshold to perform repeat analysis. Crucially, comparison of the first analysis and repeat analysis reveals the limited utility of performing repeat measurements. The relative differences between the first and second measurements are consistent with the observed analytical imprecision of the assay and the likelihood of detecting actual analytical outliers is very low.

Original languageEnglish (US)
Pages (from-to)84-89
Number of pages6
JournalClinica Chimica Acta
Volume528
DOIs
StatePublished - Mar 1 2022

Keywords

  • Analytical reproducibility
  • Element analysis
  • ICP-MS
  • Laboratory operation efficiency

ASJC Scopus subject areas

  • Biochemistry
  • Clinical Biochemistry
  • Biochemistry, medical

Fingerprint

Dive into the research topics of 'Have you caught that outlier yet? Evaluate the utility of repeat testing in nutritional and toxic element assessment'. Together they form a unique fingerprint.

Cite this