ABSTRACT
A commercially available holotranscobalamin (holo-TC) radioimmunoassay (RIA) (Axis-Shield, Dundee, Scotland) was evaluated in four laboratories and compared with a holoTC ELISA run in one laboratory. The performance of the holoTC RIA assay was comparable in three of the four participating laboratories. The results from these three laboratories, involving at least 20 initial runs of "low", "medium" and "high" serum-based controls (mean holoTC concentrations 34, 60 and 110 pmol/L, respectively) yielded an intra-laboratory imprecision of 6-10%. No systematic inter-laboratory deviations were observed on runs involving 72 patient samples (holoTC concentration range 10-160 pmol/L). A fourth laboratory demonstrated higher assay imprecision for control samples and systematic deviation of results for the patient samples. Measurement of holoTC by ELISA showed an imprecision of 4-5%, and slightly higher mean values for the controls (mean holoTC concentrations 40, 70 and 114 pmol/L, respectively). Comparable results were obtained for the patient samples. The long-term intra-laboratory imprecision was 12% for the holoTC RIA and 6% for the ELISA. In conclusion, it would be prudent to check the calibration and precision prior to starting to use these holoTC assays in research or clinical practice. The results obtained using the holoTC RIA were similar to those obtained using the holoTC ELISA assay.
