Key points are not available for this paper at this time.
The derivation of quality control data from patient laboratory values presupposes, first, that the distribution of those values is stable over long periods of time; second, that it is possible to estimate accurately the mean of the entire population from a small sample of that population. That the first presupposition is true for erythrocyte indices has been shown elsewhere. The second presupposition is examined in this paper. Six estimators were tried in turn on two data batches, each consisting of 480 patients from two university hospitals. An estimator which smoothed the data (decreased the effect of oudying values) and at the same time showed the characteristics of a moving average was found sufficiently effective to serve as the major basis for quality control of an automated whole blood analyzer.
Bull et al. (1974) studied this question.