Accuracy of methods to estimate ionized and "corrected" serum calcium concentrations in critically ill multiple trauma patients receiving specialized nutrition support

BACKGROUND: The purpose of this study was to determine the accuracy of 22 published methods to estimate serum ionized calcium (iCa) and "corrected" total serum calcium (totCa) concentrations in critically ill, multiple trauma patients. Seven of these formulas estimated iCa and 15 were directed toward predicting a "corrected" totCa. METHODS: Adult patients admitted to the trauma intensive care unit who received specialized nutrition support were consecutively recruited for study. Patients who received blood products, i.v. calcium, or therapeutic doses of heparin within 24 hours before the laboratory measurements or had a history of cancer, bone disease, parathyroid disease, hyperphosphatemia (> or = 6 mg/dL), hyperbilirubinemia (> 3.5 mg/dL), or renal failure requiring dialysis were excluded. The 22 published methods were analyzed for sensitivity, specificity, percentage false negatives, and percentage false positives for predicting hypocalcemia or hypercalcemia. RESULTS: One hundred patients were studied 4.9 +/- 3.3 days postinjury and were receiving enteral nutrition (n = 81), parenteral nutrition (n = 18), or both (n = 1) at the time of study. Twenty-one patients were hypocalcemic (iCa < or = 1.12 mmol/L) and 6 were hypercalcemic (iCa > or = 1.32 mmol/L). The mean sensitivity of the 22 methods for assessing hypocalcemia was 25% +/- 32% and the specificity was 90% +/- 18%. Although the average percentage of false positives for assessing hypocalcemia was 10% +/- 18%, the mean percentage of false negatives was inordinately high at 75% +/- 32%. The most common method for determination of "corrected" totCa concentration ["corrected" calcium = totCa + (0.8 x (4-serum albumin concentration))] had a sensitivity of only 5%. The McLean-Hastings nomogram method, the most common method for estimating serum iCa concentration, had a sensitivity of 67% but unfortunately also had a significant false-positive rate of 27%. Serum totCa correlated modestly with iCa (r2 = .334, p < .001). Those patients with a serum albumin < or = 2 g/dL (n = 43) had a significantly higher prevalence of hypocalcemia than those with a higher serum albumin concentration (37% incidence of hypocalcemia vs 10%, respectively, p < .002). CONCLUSIONS: Aberrations in calcium homeostasis are frequent (27%) in postresuscitative critically ill multiple trauma patients. Methods for predicting hypocalcemia lack sensitivity and are often associated with an unacceptable rate of false negatives. Predictive methods for estimating ionized or corrected serum concentrations should not be used. Direct measurement of serum iCa concentration is indicated for assessing calcium status for this population.