The sensitivity of bacteriological culturing of milk samples as a diagnostic tool for detecting S. aureus infection in a Hungarian dairy herd : Literature review and research work

Previous studies had shown the difficulty which is faced when we are trying to culture S. aureus from milk samples of infected cows. Numerous variations of culturing technique did not provide any satisfactory level of sensitivity in any previous research. We wanted to find out what level of sensitivity we could achieve and to compare the results achieved when using mixed samples , quarter samples, before milking samples and after milking samples. We, in this study, took a group of 13 S. aureus-infected cows and, each day that we visited them, took one mixed sample (from all quarters) before milking, individual quarter samples after milking and mixed samples after milking. We visited the farm 6 times in total. We attempted to culture S. aureus from these samples using a standard technique. This work has again (in agreement with many previous studies) found that the sensitivity of standard culturing of milk samples to detect S. aureus (regardless of what type of sample it is) is unsatisfactory. The best type of sampling was individual quarter samples as this identified more infected cows (51.5% positive) than the mixed samples. 16.7% of infections identified by the mixed samples (either before or after or both) would have been unidentified if we had just relied on quarter samples alone. As we have discussed, however, the majority (75% of them) of the mixed samples that were positive when all of the quarter samples were negative, occurred in the before milking mixed sample. I, therefore, would suggest that in some cases it seems as though particular cows may only shed bacteria in the early phase of milking and that the infection may not be detected after this (not detected in the after milking samples). I think that, to investigate this idea in a subsequent study, I would take two sets of quarter samples from each cow on each day of sampling; one before milking and one after milking. It would be interesting to see if the same quarters which are positive before milking are usually positive after milking and vice versa. Another idea in which this study agrees with earlier studies is that we clearly see the need for repeated sampling on different days. If repeated samples are not taken then the sensitivity (which is unsatisfactory even with repeated samples) will be even worse due to the periodic shedding of the pathogen in the milk of infected cows. Our results make this very obvious (48.5% of cows on a given day produced no positive quarter samples). Finally, if we do continue to use culturing from milk to detect S. aureus then I believe that we need to continue to try to develop new improved protocols to make it more sensitive. If we cannot find any way to improve the sensitivity of the culturing process then I think we should look for other techniques. New PCR-based procedures may be one example of a potential improvement in our ability to identify S. aureus intramammary infections