Misclassification of calcium status in end-stage kidney disease using albumin-adjusted calcium levels

Albumin-adjusted calcium remains widely used in clinical practice with guidelines for chronic kidney disease (CKD) mineral bone disorder recommending the use of serum calcium for monitoring. This is despite ionized calcium being the biologically active fraction. This study aimed to investigate the ability of total calcium and albumin-adjusted calcium to correctly assign calcium status in stage 5/5D CKD across non-dialysis, haemodialysis and peritoneal dialysis patients.

Over a 6-months, 352 paired serum and ionized calcium samples were collected from stage 5 (n = 58) and 5D (n = 294, 196 haemodialysis, 98 peritoneal dialysis) CKD patients in a tertiary-hospital setting. Albumin-adjusted calcium was calculated using the modified-Payne formula. Ionized calcium was the reference standard. The agreement between the two methods in assigning calcium status was assessed using Cohen's weighted kappa (κ) statistic.

Albumin-adjusted calcium was a poor predictor of calcium status compared to ionized calcium in stage 5/5D CKD (observed agreement 0.42, weighted κ 0.20, 95% CI 0.15-0.26). Dialysis dependence was associated with worse agreement (observed agreement 0.38, weighted κ 0.14, 95% CI 0.09-0.19). Total calcium was more reliable, however, remained inaccurate. Calcium status was not more accurately classified in those with higher albumin levels ≥30 g/L (observed agreement 0.47, weighted κ 0.23, 95% CI 0.10-0.36).

Total calcium provides better approximation of calcium status than albumin-adjusted calcium in stage 5/5D CKD. Albumin-adjusted calcium tends to 'overcorrect' serum calcium upward. Clinicians should use ionized calcium where accurate measure of calcium is indicated, with total calcium used as the next best option where resources are limited.