Formulas for Calculating Calcium Values Underestimate Hypocalcemia in Hemodialysis Patients.
Hemodialysis patients are exposed to high amounts of phosphate and citrate, causing hypocalcaemia and hyperparathyroidism (1). These perturbations have a substantial role in extravascular calcifications causing cardiovascular disease and bone abnormalities. The Kidney Disease: Improving Global Outcomes (KDIGO)  work group developed a clinical practice guideline on mineral and bone disorders and have advised frequent measurement of calcium, phosphorus, and parathyroid hormone (PTH) (1). Ideally, ionized calcium is measured, but for practical reasons adjusted calcium formulas are broadly used in the clinic. Therapeutic decisions depend on calcium status, in addition to PTH concentrations and clinical aspects (1). Until now, the adequacy of only a few formulas for calculating calcium values in hemodialysis patients has been evaluated. Here, results of the 3 best-performing formulas and the classic modified Orrell from a comprehensive evaluation of 24 formulas are presented.
Medline was searched using search terms related to formulas for calculating calcium. Ninety-one blood samples for routine control laboratory measurement from 56 hemodialysis patients were collected. The inclusion criterion was the availability of control panel samples for hemodialysis. Exclusion criteria were interference due to hemolysis, icteria or lipemia. Ethical considerations were addressed by following the code of conduct Human Tissue and Medical Research: Code of Conduct for Responsible Use (https://www.federa.org/sites/ default/files/digital_version_first_ part_code_of_conduct_in_uk_ 2011_12092012.pdf). Blood was collected before hemodialysis using SST II tubes from Becton Dickinson (REF367953) and were directly centrifuged (10 min, 10 [degrees]C, 2000g) upon arrival in the laboratory. Basic clinical chemistry tests were performed with the Abbott Architect c Systems. Intact PTH (L2KKP2; reference interval 1.6--6.8 pmol/L, serum stored at --20 [degrees]C within 3 h after collection) was measured on a Siemens Immulite 2000 XPi. Ionized calcium adjusted to pH 7.4 [[iCa.sup.2+] + (pH 7.4) X 0.5], pH, and bicarbonate were measured on a Radiometer Abl825 FLEX. Calculated calcium values were categorized as hypo-, normo- or hypercalcemia (reference interval 2.20-2.65 mmol/L) and compared with categorization of hypo-, normo- or hypercalcemia based on ionized calcium measurement (reference interval 1.20-1.30 mmol/L at pH 7.4).
The diagnostic sensitivity of a formula for hypocalcemia was defined as the percentage of identified cases that truly had hypocalcemia, based on [iCa.sup.2+]. Diagnostic specificity of a formula was the percentage of identified cases that truly had normocalcemia, based on [iCa.sup.2+]. Two patients were hypercalcemic, and therefore no diagnostic sensitivity for hypercalcemia was calculated. The KDIGO guideline advises calculating calcium in cases of hypoalbumenia. These results were inferior to the results of calculated calcium values regardless of albumin values, and so only the latter are presented.
Baseline characteristics of the included blood samples of end-stage kidney disease patients on dialysis are described in Table 1. Total protein (TP) was measured in 48 of the 91 samples. In samples with hypocalcemia, TP was 66 g/L (n = 37, SD 5 g/L); in normocalcemic samples this was 67 g/L (n = 10, SD 6 g/L). Total calcium (tCa; in mmol/L) showed 36% diagnostic sensitivity for hypocalcemia and 100% diagnostic specificity; 44 samples were misclassified as normocalcemic. The classic modified Orrell, adjusted calcium (adjCa; in mmol/L) = tCa + 0.02 X (40 -albumin (alb); in g/L) (2), showed 14% diagnostic sensitivity for hypocalcemia and 90% diagnostic specificity. The best-performing formula, adjCa = tCa + 0.01 X (30 - alb) (3), had 41% diagnostic sensitivity and 100% diagnostic specificity. The second-best formula [adjCa = tCa -((0.00019 X alb -(0.42 X (alb/4730) X (7.42 - pH)) + 0.00004 X (TP - alb) - (0.042 X ((TP - alb)/2500) X (7.42 - pH)))) (4)] had 32% diagnostic sensitivity and 100% diagnostic specificity. In third place, the unmodified Orrell, adjCa = tCa -0.0176 X (alb - 34) (5), had 26% diagnostic sensitivity and 100% diagnostic specificity.
Our results show that calculated calcium values have poor diagnostic sensitivity for hypocalcemia. Remarkably, total calcium performs better than the widely used modified Orrell formula (2) in classifying calcium status in hemodialysis patients, in agreement with previous reports. The Jain et al. formula (3), which was specifically created for this purpose, performed best in classifying calcium status in hemodialysis patients. Still, about 60% of the hypocalcemic patients were missed.
Serum albumin concentration is only one of numerous factors influencing ionized calcium concentrations. In addition to these limitations, the formulas from Orrell were created using the analytically less specific bromcresol green method for albumin measurements (2, 5).A caveat of blood gas analyzers is their inaccuracy at very low ionized calcium values. A possible limitation of this study is the measurement of ionized calcium at pH 7.4.
The tendency of calculated calcium to underestimate hypocalcemia has important clinical consequences since the KDIGO guideline advises altering therapy guided by calcium values to prevent bone and mineral disorders (1). We strongly advocate not using any formulas to calculate calcium values to assess calcium status in hemodialysis patients, but instead to use ionized calcium measurements.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article. Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
(1.) Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD Work Group. KDIGO clinical practice guideline for the diagnosis, evaluation, prevention, and treatment of chronic kidney disease-mineral and bone disorder (CKD-MBD). Kidney Int Suppl. 2009; 76(Suppl113):S1-130.
(2.) Correcting the calcium. Br Med J 1977; 1:598.
(3.) Jain A, Bhayana S, Vlasschaert M, House A. A formula to predict corrected calcium in haemodialysis patients. Nephrol Dial Transplant 2008; 23:2884-8.
(4.) Moore EW. Ionized calcium in normal serum, ultrafiltrates, and whole blood determined by ion-exchange electrodes. J Clin Invest 1970; 49:318-34.
(5.) Orrell DH. Albumin as an aid to the interpretation of serum calcium. Clin Chim Acta 1971; 35:483-9.
Thijs C. van Holten [2*], Frans A.L. van der Horst , Sebastiaan J. Huisman , Ference J. Loupatty 
 Nonstandard abbreviations: KDIGO, Kidney Disease: Improving Global Outcomes; PTH, parathormone; TP, total protein; tCa, total calcium; adjCa, adjusted calcium; alb, albumin.
 Clinical Chemistry Department Reinier Haga Medisch Diagnostic Centre Delft, The Netherlands
 Nephrology Department Reinier de Graaf Gasthuis Delft, The Netherlands
* Address correspondence to this author at:
Reinier de Graafweg 7 Delft
Previously published online at DOI: 10.1373/clinchem.2016.262261
Table 1. Baseline characteristics of the included patients, categorized on measured ionized calcium. (a) Hypocalcemic Normocalcemic samples samples (n = 70) (n = 19) Male, % 57% 74% Age, years; median 72(54-81) 68 (57-80) (IQR) eGFR, (b) mL x 6(2) 7 (3) [min.sup.-1] x [(1.73[m.sup.2]).sup.-1]; mean (SD) Creatinine, mol/L; 757 (249) 682 (261) mean (SD) Urea, mmol/L; mean 23.1 (5.2) 21.7(6.1) (SD) Albumin, g/L; mean 33 (4) 34 (3) (SD) Parathormone, pmol/ 40.3 (23.6) 33.3 (30.5) L; mean (SD) Total calcium, mmol/ 2.21 (0.18) 2.41 (0.13) L; mean (SD) Ionized calcium, 1.10(0.10) 1.24 (0.03) pmol/L; mean (SD) Phosphate, mmol/L; 1.74 (0.46) 1.60 (0.53) mean (SD) Sodium, mmol/L; mean 138 (4) 138 (2) (SD) Potassium, mmol/L; 5.0 (0.7) 5.0 (0.8) mean (SD) pH; mean (SD) 7.6 (0.08) 7.6(0.14) Bicarbonate, mmol/ 24(3) 25 (3) L; mean (SD) (a) Normally distributed values reported as mean (SD); nonparametric values reported as median and interquartile range (IQR). (b) eGFR, estimated glomerular filtration rate.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Letters to the Editor|
|Author:||van Holten, Thijs C.; van der Horst, Frans A.L.; Huisman, Sebastiaan J.; Loupatty, Ference J.|
|Article Type:||Letter to the editor|
|Date:||Nov 1, 2016|
|Previous Article:||Glycated Albumin Identifies Prediabetes Not Detected by Hemoglobin [A.sub.1c]: The Africans in America Study.|
|Next Article:||Measurement of Hematocrit in Dried Blood Spots Using Near Infrared Spectroscopy: Robust, Fast, and Nondestructive.|