RESUMEN
Background: Most of the laboratories use previously published regression equations for estimation of calcium which may not fit for their population. So deriving locally a regression equation for albumin-adjusted calcium (CaAd) is a mainstay to avoid population-based differences. Aims & Objective: To derive regression equation for albumin-adjusted calcium in our laboratory and validate it for the local population. Material and Methods: Total 575 normal healthy individuals of 35-65 years were included in the present study and were estimated for serum total calcium (CaT), ionized calcium (Ca2+), and albumin. The linear regression equation for the binding of calcium and albumin was derived in a cohort of 450 normal healthy individuals of 35-65 years, and the albumin-adjusted calcium equation was internally validated in a separate cohort of 125 subjects. The performance of this equation was compared with a previously published equation: CaAd (mmol/L) = CaT (mmol/L) + 0.02 (40 - [albumin] (g/L). Results: The local adjustment equation obtained from the derivation subset was expressed by the relationship; CaAd (mmol/L) = CaT (mmol/L) + 0.03 (37.33 – [albumin] (g/L)). The equation was internally validated with an adjusted r2 shrinkage value of 0.0009 in a validation subset. Bland-Altman plot showed statistically significant difference (Mean = 0.13 mmol/L) when both formulae were compared for the population. Conclusion: A locally derived and internally validated albumin-adjusted calcium equation differed significantly from previously published equations. Individual laboratories should determine their own linear albumin-adjusted regression equation for calcium rather than relying on published formulas.