Albumin‐corrected calcium and ionized calcium in stable haemodialysis patients

variety of formulae have been proposed to permit Background. It is ionized calcium that is physiologic- calculation of the ionized calcium or of the ‘corrected’ ally active and under homeostatic control; however, total calcium (i.e. an estimation of the expected total total calcium is more conveniently measured. Formulae calcium were the serum albumin normal ) from the for correction of calcium to account for albumin total calcium and protein concentration. The methodbinding have not been validated in a dialysis setting. ologies used in the development of these formulae Methods. We measured ionized calcium simultan- varied: few were derived from clinically relevant mateously with total calcium (t[Ca]), albumin, total pro- erial and none were developed and validated in separtein and pH before dialysis in 50 stable outpatients ate, independent data sets. The literature on correction and convalescent inpatients. of calcium is, therefore, deficient in evidence supporting Results. Although 92% of patients were taking calcium the use of formulae for the correction of calcium in supplements and 70% taking alphacalcidol, 11 patients any population. A previous large, well-designed study (22%) had ionized hypocalcaemia. To facilitate com- comparing correction algorithms with measured ionparison of calculated ionized calcium, measured total ized calcium in unselected patients found that none calcium (t[Ca]), and ‘corrected’ calcium (c[Ca]), with performed significantly better than measured total the criterion measure of ionized calcium, all measure- calcium [1]. ments were converted to z scores, standardized on the Calcium in serum exists in three fractions: bound to normal range for each variable. Results are expressed plasma proteins (approximately 40%), chelated to as intraclass correlation coefficients (ICC: 0, all differ- serum anions (13%) and free ionized calcium (47%) ences are due to error; 1, all differences are due to [2,3]. It is this last fraction which has biological activity between patient variation). and is under homeostatic control. The equilibrium Conclusions. None of the published formulae greatly between the fractions is dependent on a number of improved the test characteristics beyond simply using variables, most importantly, the concentrations of the total calcium. A correction formula in widespread serum proteins and the pH. While the measurement of use (Payne), quoted in reference texts, agreed less well total calcium, albumin and total protein is available in with ionized calcium than did the unadjusted measured standard laboratories, measurement of ionized calcium calcium. Correction formulae should be abandoned in remains more difficult and is generally performed only favour of the use of uncorrected calcium. In cases of in reference laboratories. In addition, great care must doubt, ionized calcium should be directly measured. be taken with the method of venepuncture and subsequent sample handling [4]. Patient posture and the