A-106 Albumin-corrected Calcium: Should we use it or not?

Abstract Background Accurate assessment of calcium (Ca) level is important for proper classification of Ca status. Biologically active ionized Ca (iCa) is the best marker for precisely estimating functional Ca levels. However, the cost and pre-analytical challenges limit wide use of this test. As an alternative, many facilities like ours report albumin-corrected Ca (ACa) and uncorrected total Ca (TCa) results concurrently with every metabolic panel order. This has led to frequent calls from our clinicians, specifically when a substantial difference is noted between the two results. Recent studies have challenged the widespread use of ACa in clinical practice, underscoring its potential inaccuracies in determining Ca status, especially in hospitalized patients needing frequent Ca level monitoring. Importantly, a subset of these investigations has affirmed the continued reliability of TCa in Ca assessment. Our study aims to evaluate the effectiveness of reporting ACa and TCa in our large urban academic medical center while exploring their selective utilization. Methods In this retrospective study, we analyzed 12,899 samples collected from 4,264 adult patients at our university hospital between January 2016 and December 2024. TCa, iCa, and ACa results reported from the same blood draw were included in the data analysis. The simplified Payne formula was employed for ACa calculation, with iCa serving as the reference method. Our facility*s reference ranges are 3.5-5.2 g/dl for albumin, 4.4-5.2 mg/dL for iCa and 8.4-10.2 mg/dL for TCa/ACa. Severe (critical) hypocalcemia and hypercalcemia were defined as <3.1 mg/dL and >6.3 mg/dL for iCa, and <6.0 mg/dL and >14.0 mg/dL for TCa/ACa, respectively. Statistical analysis was conducted using GraphPad Prism 9, with p<0.05 considered statistically significant. Results Among 4,264 patients, the majority (70%) had low albumin levels (<3.5 g/dl) and were inpatients (91%). ACa underestimated hypocalcemia and overestimated hypercalcemia significantly compared to iCa. In contrast, TCa overestimated hypocalcemia, particularly by several fold in severe hypocalcemia. In hypoalbuminemia, there was pronounced underestimation of hypocalcemia and overestimation of normocalcemia and hypercalcemia by ACa, while TCa showed opposite trends. ACa performed poorly compared to TCa in chronic kidney disease (CKD) patients with eGFR <60 and <30 ml/min/1.73m2. Importantly, both ACa and TCa overestimated severe hypocalcemia and hypercalcemia in patients with normal albumin levels. ACa and TCa showed comparable performance to iCa only when both albumin and Ca levels were within normal ranges. Conclusion ACa is unreliable for classifying clinical Ca status, especially in hypoalbuminemia. ACa performs worse than TCa in CKD patients. Contrary to previous literature, TCa is as unreliable as ACa unless both albumin and TCa levels are normal. We recommend iCa as the sole accurate test for monitoring Ca status in hospitalized and CKD patients.