Analytical variance and definition of a reference change as a function of calcium concentration.

Using data from the Centers for Disease Control (CDC) Proficiency Testing (PT) Surveys, we obtained estimates of repeatability (intralaboratory variability between results on the same material) and reproducibility (interlaboratory variability between results on the same material) for the Technicon SMA 6 (or 12/60) and SMAC 1 (or II) systems used with cresolphthalein complexone methodology to measure serum calcium. The two systems were comparable in terms of short-term (within-day) repeatability, long-term (three to six months) repeatability, short-term (one to two weeks) reproducibility, and long-term (three to six months) reproducibility. The long-term repeatability was essentially the same as the long-term reproducibility. Short-term repeatability, long-term repeatability, and long-term reproducibility increased linearly with increased calcium concentration over the range 1.75 to 2.95 mmol/L; short-term reproducibility showed no significant change over this range. The effect of analytical variance on the definition of a reference change in semiannual calcium measurements was demonstrated.