Membrane filter-deoxyribonucleic acid method of somatic cell counting: collaborative study

In the membrane filter-deoxyribonucleic acid test, somatic cells are trapped on a special filter. A chemical determination of the DNA of the trapped material gives an estimate of the number of cells. The MF-DNA was evaluated and compared with the direct microscopic somatic cell counts (DMSCCs) in 10 laboratories. Cell levels in the 10 bulk milk samples analyzed ranged from 380,000 to 1,700,000/mL. Regression analysis gave a relation of 0.181 absorbance at 490 nm, when 2.5 mL of milk was analyzed and 5.0 mL reagent was used for color development. The 95% confidence limits on the slope were 0.158-0.204. Analysis of variance showed that both procedures had significant (P less than 0.001) systematic errors, but that only the DMSCC method showed a significant laboratory X sample interaction (P less than 0.001). The repeatability standard deviation of the MD-DNA procedure was 117,000 cells/mL (14%), and the reproducibility standard deviation was 198,000/mL (23%). This compared favorably with the DMSCC method which has a repeatability of 162,000/mL (19%) and a reproducibility of 187,000/mL (34%). The error of the MF-DNA was constant at all cell counts, whereas there was a tendency for the error to increase in the DMSCC method at high cell counts. This method has been adopted as official first action.