Comparison of three commercial kits and a microbiological assay for the determination of vitamin B12 in serum

Different methods are available for cobalamin determination in serum. Microbiological and radio ligand binding assays are the most commonly used. Kits involving non-isotopic competitive-binding assay have been recently commercialized. In the present work, cobalamins were determined in 146 patient sera, using four methods: a microbiological method, two no boil radio ligand binding assay kits (Magic B12 FOL (NB) from Ciba-Corning and SimulTRAC SNB No Boil from Becton Dickinson) and a non-isotopic kit with acridinium ester labelled cobalamin (Magic Lite from Ciba-Corning). Median (range) cobalamin concentrations in pmol l-1 were 317 (15-1291) using the microbiological method, 355 (25-3469) using the Magic Lite kit, 355 (35-2312) using the Magic B12 FOL (NB) kit and 380 (37-2021) using the SimulTRAC SNB No Boil kit. The ANOVA test indicated that differences between methods were statistically significant (p < 0.01). Competitive-binding methods gave higher results than the microbiological method. Although correlation coefficients were not excellent (0.88 < r < 0.96), the results obtained with the different methods were generally similar and confirmed that competitive methods are useful for detecting low serum concentration of vitamin B12.