Assessment of the reference method for determination of total calcium

for determination of calcium in serum has been critically assessed, both to evaluate its performance in one labo-ratory and to identify ambiguities and errors in the pub-lished texts. Within-run and between-run precision and bias have been estimated, and a revised protocol is pre-sented. In reviewing quality control in clinical chemistry, Whitby et a!. (1) stressed the need for assessing the accuracy of analytical techniques by means of recov-ery experiments and by comparison with what might have been called “reference methods. ” Caraway (2) emphasized the importance of specificity for high ac-curacy and discussed the difficulties associated with recovery experiments. He did not discuss reference methods, perhaps because of a general awareness of the logical and practical difficulties of establishing that any one method of analysis for a particular con-stituent is more accurate than another. Cali (3) pointed out some implications of the Refer-ence Method (4) for the determination of calcium in serum, the accuracy of which had been defined by comparison with an absolute (“definitive”) method, isotope dilution-mass spectrometry (5). In principle, the Reference Method permits the introduction of true accuracy control into clinical chemistry, and it was therefore important to assess both its practicabi-lity, and its accuracy and precision, as it was evalu-ated in one laboratory. The field trial which preceded publication of the protocol yielded a limited amount of data on each of these points; the present study was designed to complement these findings.