Use of MALDI-TOF to identify cryopreserved mastitis pathogens collected from 2003 to 2011 that were originally identified using conventional microbiological methods

Abstract

For decades, conventional microbiological methods have been used to identify bacterial causes of bovine mastitis. Although these methods are relatively accurate for identification of important mastitis pathogens, all diagnostic tests are imperfect, and as testing technologies advance, widespread use of newer technologies may result in differences in the distribution of etiologies that are identified. As a result, historical research using conventional microbiological methods may not be comparable to results of current studies. The objective of this study was to compare agreement between the original identification of mastitis pathogens from cows enrolled in mastitis studies between 2003 and 2011 with identification of the same isolates using MALDI-TOF. Cryopreserved bacterial isolates (n = 308) that had been recovered from quarter milk samples and originally identified using conventional microbiological techniques were used. Bacterial identification was performed using MALDI-TOF. Among all isolates, 277 were able to be identified using MALDI-TOF, and the overall observed levels of agreement were 86% and 64% for identification at the genus level and genus-species level, respectively. The kappa statistic for agreement between methods at the genus level was substantial at 0.80 (95% CI: 0.78, 0.82) but dropped to 0.64 (95% CI: 0.62, 0.66) for agreement at both the genus and species level. For gram-positive isolates, agreement at both the genus level and the genus and species level was substantial. In contrast, for gram-negative isolates, the genus-level agreement was substantial, but agreement at both the genus and species levels was moderate. Our findings suggest substantial agreement between MALDI-TOF and conventional methods for determining genus-level identification, but some discrepancies occur at the species level. These results indicate that historical mastitis research using conventional microbiological methods are comparable at the genus level to current results using MALDI-TOF, but some caution should be applied when making species-level comparisons, especially for gram-negative pathogens.