Determination of reference values for serum folate and vitamin B12 using three different immunoassays: is it worth making an effort to produce them in our laboratory?

ABSTRACT

BACKGROUND: Despite being a widely studied concept, the reference interval is the most widely used medical decision-making tool. As such, it is vital that these limits are correctly established and regularly reviewed in the clinical laboratory. METHODS: The reference population comprised 315 healthy individuals selected a priori from Bizkaia province. Blood and serum samples were sent for subsequent assay of vitamin B12 and folate using three immunochemical methods. Reference values were calculated using non-parametric methods. RESULTS: The reference values for serum vitamin B12 and folate were almost identical to those obtained previously using the same methods. Use of new reference values led to an increase in the kappa value despite the low agreement in the case of vitamin B12 (0.4 - 0.62). However, precision obtained for vitamin B12 (94.48 - 96.55%) and folate (95.77 - 97.18%) was very high. The intraclass correlation coefficient ranged from 0.723 to 0.894. Furthermore, a Passing-Bablok regression analysis gave acceptable correlation coefficients of 0.75 - 0.94 for vitamin B12 and 0.92 - 0.95 for folate. CONCLUSIONS: Vitamin B12 and folate deficiencies are currently being over-diagnosed leading to an increase in the number of unnecessary consultations. The main conclusion that can be drawn from our study has resulted in a change in reference values in our laboratory, with a subsequent increase in our ability to accurately detect possible deficiencies. Furthermore, as this study involved all methods currently in use in the Basque healthcare network, its conclusions can be extrapolated to the whole population covered by Osakidetza, thereby improving the rational use of healthcare funding.