[The congruence and repeatability of electronic determinations of somatic cell counts in milk by diagnostic...

Six joint trials were conducted in six diagnostic laboratories equipped with the Coulter Counter Model D electronic computers. The congruence and repeatability of the determination of somatic cell counts in milk were determined in these trials. The results were confronted with similar trials performed in the Netherlands. The average differences in the results of determinations performed in the different laboratories exhibited a tendency of improvement as the joint tests proceeded, so that a satisfactory congruence of the results in all laboratories was obtained in the last trial. The standard deviation of the average difference and the average difference between parallel samples were characterized by high variedness . As found, the calibration of the computers by the plateau-and-trough method is not satisfactory and should be replaced by the calibration method of half the number and by the setting of an equal threshold size of the counted cells. The repeatability of determination will have to be improved by controlling rationing devices for milk dilution and by controlling the standard volumes counted by the apparatuses.