Current "corrected" calcium concept challenged.

There is wide individual variation in the number of millimoles of calcium bound per gram of albumin in the serum. Individual regression coefficients for serum calcium concentration on serum albumin concentration have been determined in 62 people (25 of our own patients and 37 reported by others). The 95 percentile range was 0-007-0-053 mmol/g, with a median value of 0-025 mmol/g. Accordingly, it is not valid to "correct" a person's measured serum total calcium concentration for variations in serum albumin concentration using an average regression coefficient. Rather, the individual's own regression coefficient must be used. A tourniquet test seems to be the simplest technique for determining this value. Even then, precise interpretation of an individual's corrected serum calcium concentration is possible only when an appropriate reference range for corrected serum calcium concentration has been established. Such an appropriate reference range must be determined from an adequate number of normal people using the individual's own correction factor in each case.