SUN-LB67 Revisit Serum Calcium Correction

Abstract

Context: The serum calcium level is one of most routinely ordered tests in clinical practice. Many factors can affect calcium level and its interpretation. There are challenges and barriers in applying calcium correction formulas to every-day practice. Objective: Revisit correlation between total and ionized calcium levels, and dependence of serum calcium on albumin, pH and creatinine levels. Methods: This study included 1537 subjects enrolled in a parathyroid disease clinical protocol. We examined calcium and relevant biochemistry tests collected simultaneously and repetitively over consecutive years. Histograms, repeated measures correlation, correlation plots, and liner regression plots were used to analyze and visualize the data. Results: We found that: 1) directly measured total serum calcium and ionized calcium had excellent correlation and dependence with p-value=2.2e-16, repeated measures correlation coefficient (rmcorr)=0.919, and 95% interval (CI) = 0.916 to 0.922; 2) there was a low dependence between total serum calcium and albumin levels (rmcorr=0.454, 95% CI=0.433 to 0.474), a low dependence between ionized calcium and pH levels (rmcorr=-0.309, 95% CI= -0.326 to -0.292), and no dependence between total calcium and creatinine levels (rmcorr=0.026 95% CI=0.012 to 0.040); 3) using the commonly applied correction formulas, to either adjust total calcium based on albumin levels or else adjust ionized calcium based on pH levels, did not improve dependence among them. Conclusions: We therefore suggest using directly measured total serum calcium and/or ionized calcium level to assess clinical calcium status in general patients tested for parathyroid related disorders.