SOMATIC-CELL COUNTING IN MILK - COMPARATI VE-STUDIES ON THE MEASUREMENT USING THE FLUORESCENCE OPTICAL-PRINCIPLE (FOSSOMATIC-360) AND THE SOMASCOPE

Somatic cell counting was performed in comparative studies using the F ossomatic 360 and the ,,Somascope'', the aim being to determine the pr ecision of measurement (repeatability) and to objectify the specific i nfluences of the instruments as well as the effects which are a functi on of sample pretreatment. Sample pretreatment was done using varying preservatives (potassium dichromate, sodium azide, boric acid, Bronopo l, no preservative) without and with 24-h cold storage at 7-degrees-C. With both sodium azide preservation and without chemical preservation the influence on cell counts was markedly more pronounced compared wi th potassium dichromate-, Bronopol- or boric acid preservation. Compar ed with other procedures no chemical preservation and sodium azide did not yield absolutely reliable cell counts. When testing repeatability both systems gave low coefficients of variation, which means a high p recision. The cell counts were, however, different; here, both the ins trument and milk sample pretreatment must be taken into account. The r esults obtained with the quarter foremilk samples show that, although the instruments measure varying cell counts, these differences are sys tematical (high correlations). With the ,,Somascope'' one has to take into account that, compared with the Fossomatic 360, markedly higher c ell counts are measured in the low cell count range and markedly lower cell counts in the higher range. With potassium dichromate and Bronop ol (without storage and 24-h storage) the values are well comparable b etween both systems. International comparative test using the Fossomat ic and the ,,Somascope'' gave repeatabilities (r) and reproducibilitie s (R) which led, according to the IDF-standard 148:1991, to comparable results. For repeatability and reproducibiliy values of r=78.4 and R= 200.1 were obtained with the ,,Somascope'' for all laboratories (n=16) and all milk samples (n=10). Despite the restricted validity of these values indications can be derived from them. Between the milk samples 1-10 the values ranged, for all laboratories (n=16) between 27 and 11 8 for r and between 47 and 428 for R. Here, there exists naturally a d ependency from the level of the measuring value (VK(sr) 3.2-6.9 %, VR( sR) 7.5-13.7 %).