Superiority of dynamic over static reference intervals for intact, midmolecule, and C-terminal parathyrin in...

We compared the clinical performance of assays of intact, C-terminal, and midmolecule parathyrin (PTH), when used with either a dynamic reference interval (based on the range of serum PTH concentrations observed in 35 healthy individuals during acute modifications of their blood calcium concentrations) or a gaussian (mean +/- 2 SD) interval derived from normocalcemic individuals. Dynamic intervals were substantially different from gaussian intervals, with half of the area delimited by the gaussian limits for calcium and intact PTH concentrations, and one-third for both C-terminal and mid-PTH assays, being outside the range of values observed during the dynamic tests. Use of the dynamic intervals increased the average clinical sensitivity of the three assays for detecting primary hyper- and hypoparathyroidism from 68% to 97% (and up to 100% for the intact and C-PTH assays). Even though only the intact PTH assay allowed complete separation between primary hyperparathyroid and nonparathyroidal hypercalcemic patients, the average proportion of patients correctly classified in this latter category was increased from 40% to 70% by the use of dynamic intervals. We conclude that gaussian reference intervals are largely responsible for the poor clinical sensitivity of many types of PTH immunoassays and that they should be replaced by dynamic reference intervals when evaluating calcemic disorders.