[Blood ionized calcium fraction in various serum albumin levels surveyed in non-dialysis patients].

Blood ionized calcium (iCa) fraction is affected by the serum albumin(Alb) level, even though this effect might not be appropriately estimated by the formulae proposed previously. To clarify a reasonable regimen for predicting iCa from serum total Ca (tCa), we investigated the relationship of blood iCa, tCa, and serum Alb levels through 124 samples from 116 non-dialysis patients requiring iCa measurement at the Nephrology Section of Toride Kyodo General Hospital. The patients comprised 61 males and 55 females with the mean age of 66.9 +/- 1.4 years, including 9 cases of hypercalcemia, 110 of normocalcemia, and 5 of hypocalcemia based on their iCa levels. Their background diseases were 25 cases of chronic glomerulonephritis, 17 of nephrotic syndrome, 40 of diabetes mellitus, 4 of collagen diseases, and 30 of others. Their mean serum Cr was 2.44 +/- 0.21 mg/dl, and 77 patients showed elevated Cr levels. Four adjustment formulae: one derived from Payne's, two from the proposal of K/DOQI Clinical Practice Guidelines, and a theoretical one based on the previous in vitro experiments, were compared with the non-adjusted value (tCa itself) with respect to their suitability for estimating iCa. The correlation coefficient of tCa with iCa was higher than the values adjusted by the above four formulae. The difference of iCa from tCa divided by eight, which concisely predicted iCa based on the assumption that half the serum Ca is bound to protein, was less than 1/8th of the other adjusted Ca levels. Hence none of the adjusted Ca by the above formulae was superior to nonadjusted tCa from the point of estimating the iCa level. Moreover, the sensitivity for predicting hypocalcemia was the highest in tCa, even though its specificity was lower than the other adjusted values. In conclusion, no adjustment formula is required to predict ionized Ca from tCa, and to screen hypo- or hypercalcemia.