Abstract
During a 12 month period over 5000 calcium estimations have been performed in our laboratory, the vast majority as part either of an admission profile or routine outpatient screening. In the absence of a reference range appropriate to the local Saudi population, these results had been interpreted by comparison with figures provided by the manufacturers of our autoanalyzer system (ACA - DuPont Reference Range: 8.7–10.2 mg/100 ml; 2.18–2.55 mmol/1). No formal account had been taken of abnormal protein levels or of acid-base status. The author has set out to establish a more relevant reference range for total calcium and to assess, by a preliminary study, the extent to which nomogram estimations of free calcium might better identify those patients with a genuine disturbance of calcium metabolism.
The study population included those patients (n = 478) in whom both albumin and total protein had been concomitantly determined. Assuming a blood pH of 7.40, since only rarely had this been measured, free calcium levels were estimated according to both albumin and total protein levels. On the basis of the assumption that requests for blood calcium estimations are many while genuine disorders of calcium metabolism are few, approximate 95% reference ranges were computed for each of the three parameters: total calcium (7.4–10.2 mg/100 ml; 1.85–2.55 mmol/1), free Ca2+prot.(3.96–5.44 mg/100 ml; 0.99–1.36 mmol/1) and free Ca2+alb. (4.06–5.38 mg/100 ml; 1.02–1.35 mmol/1) after exclusion of outliers greater than three standard deviations from the mean of the whole population. Histograms for free calcium, whether estimated from total protein or albumin, more closely approximated a normal Gaussian distribution than did total calcium. For each estimation patients were deemed ‘hypocalcaemic’ or ‘hypercalcaemic’ if more than three, and ‘borderline’ if between two and three, standard deviations from a population mean. Cross comparison between the calcium status determined by each of the three methods and a retrospective review of any relevant clinical findings was made for all patients lying outside any one of the three 95% confidence ranges.