Total, ultrafiltrable, dialysable and ionised calcium measurements in randomly selected samples compared with clinical diagnoses.

Total, ultrafiltrable (Worthington 'Ultrafree'), dialysable and ionised (Orion SS-20) calcium levels were measured in 110 serum samples selected randomly from patients for whom total calcium was ordered as a routine test. The available medical records were used to categorize most of the 86 patients as a guide to determining if the calcium results were appropriate to the diagnosis. Total calcium measurement was precise and reliable but was affected by the albumin level. Dialysable calcium measurement was good for diagnostic purposes and was not affected by albumin. Of 25 patients with apparently normal calcium metabolism, six had elevated ultrafiltrable calcium results. Ionised calcium measurement required the most time and sample and offered few, if any, clear advantages in diagnosis over the other free calcium tests. Further improvements in ionised calcium measurements appear to be needed.