Comparison of urea content in milk, measured in different laboratories

In sechs Labors (drei in Slowenien und drei in Tschechien) wurde der Harnstoffgehalt von zehn Milchproben mit zwei bis drei Wiederholungen bestimmt. Es wurden unterschiedliche Methoden benutzt, alle basierten auf enzymatischer oderIn sechs Labors (drei in Slowenien und drei in Tschechien) wurde der Harnstoffgehalt von zehn Milchproben mit zwei bis drei Wiederholungen bestimmt. Es wurden unterschiedliche Methoden benutzt, alle basierten auf enzymatischer oderphotometrischer Basis. Der durchschnittliche Harnstoffgehalt variierte von 2,38 mmol/l im Labor E bis 3,43 mmol/l im Labor D. Die höchste Standardabweichung wurde im Labor 6, die niedrigste im Labor Ffestgestellt. Der höchste Variationskoeffizient wurde in Labor E (51%) und Labor 6 (44,7%) festgestellt. Die höchste Differenz zwischen zwei Labors bestand zwischen Labors D und E, die niedrigste zwischen 6 und C. Die Standardabweichungen der Differenzen zwischen den Labors variierte von 0,12 bisKeywords: cows; milk; urea content; determination methods 0,46. Die t-Teste zeigten hoch signifikante Unterschiede zwischen den meisten Labors. Die berechneten Korrelationskoeffizienten in den Labors variierten von 0,96 bis 0,99 und waren statistisch hoch signifikant. Es wurden auch lineare Regressionskoeffizienten berechnet, um den Harnstoffgehalt der Milch aus den Daten in einem Labor auch in anderen Labors schätzen zu können. (Schlüsselwörter: Kuh, Milch, Harnstoffgehalt, Methode)Urea content was determined by the analysis of ten milk samples with 2 to 3 successions in each of the 6 laboratories (3 in Slovenia, 3 in Czech Republic). Determination method was different, but the principle of urea determination in these laboratories was based on enzymatic and/or photometric method. The average urea content varied from 2.38 mmolfl in the laboratory E to 3.43 mmol in the laboratory D. The highest standard deviation was found in the laboratory A, and the lowest in the laboratory F. The highest VC percent was noticed in the laboratory E (51.0%) and laboratory A (44.7%). The analysis results for the difference in urea content in milk measured in different laboratories indicated the lowest difference between the laboratories A and C, and the highest between the laboratories D and E. Standard deviation of differences ranged from 0.12 and 0.46. With t-test we examined the differences between laboratories, which were in almost all instances highly statistically significant. We also calculated the correlation coefficients among laboratories, ranging from 0.96 to 0.99, and were highly statistically significant. We determined regression coefficients among these laboratories and with linear regression equation predicted values for each laboratory if a certain trait was measured in any of the other laboratories.Urea content was determined by the analysis of ten milk samples with 2 to 3 successions in each of the 6 laboratories (3 in Slovenia, 3 in Czech Republic). Determination method was different, but the principle of urea determination in these laboratories was based on enzymatic and/or photometric method. The average urea content varied from 2.38 mmolfl in the laboratory E to 3.43 mmol in the laboratory D. The highest standard deviation was found in the laboratory A, and the lowest in the laboratory F. The highest VC percent was noticed in the laboratory E (51.0%) and laboratory A (44.7%). The analysis results for the difference in urea content in milk measured in different laboratories indicated the lowest difference between the laboratories A and C, and the highest between the laboratories D and E. Standard deviation of differences ranged from 0.12 and 0.46. With t-test we examined the differences between laboratories, which were in almost all instances highly statistically significant. We also calculated the correlation coefficients among laboratories, ranging from 0.96 to 0.99, and were highly statistically significant. We determined regression coefficients among these laboratories and with linear regression equation predicted values for each laboratory if a certain trait was measured in any of the other laboratories