ABSTRACT
OBJECTIVE: The aim of proposed paper was to compare a three total 25-hydroxy-vitamin D immunoassays to that of HPLC with UV detection. MATERIAL AND METHODS: Serum 25-(OH) D levels were measured from blood samples of 109 patients with different immunoassays (ABBOTT, ROCHE, SIEMENS) and method of HPLC which was chosen as the reference. In the first step immunoassays were compared to HPLC. In the second step immunoassays were compared to each other. Further purpose of methods comparison the Passing-Bablok regression and Bland-Altman analysis were used. The limits of maximum acceptable differences were set at 21.5 %, according to Vitamin D Standardization-Certification Program (VDSCP). In the last step, the concordance in the interpretation of measured results was evaluated. RESULTS: None of the examined 25-(OH) D immunoassays was comparable to HPLC and to each other. Bland-Altman analysis revealed, in comparison to HPLC, that ROCHE showed positive bias +28.0 %, ABBOTT +0.2 % and SIEMENS -23.4 %. Although average bias of ABBOTT immunoassay is insignificant, particular results do deviate significantly (-89.4 % to 89.0 %). The concordance in the interpretation of measured results, in comparison to HPLC, was highest with ABBOTT (65.21 %), then with ROCHE (59.63 %) and lowest with SIEMENS (47.79 %). CONCLUSION: The results of the proposed papers suggest low levels of 25-(OH) D immunoassays standardization and an alternative to use assay-specific decision limits.
