Build in margin for error when interpreting individual cow cell counts

Milk producers need to build in a 40,000-cells/ml margin for error to individual cow somatic cell count data before planning dry cow antibiotic or teat sealant protocols, according to findings from a study at The Institute for Animal Health (IAH).

Single individual cow cell count values have long been recognised as too inaccurate to use for mastitis management purposes, so usually the last three measurements are used. The IAH study set out to examine whether the current approach is giving vets and dairy farmers a reliable enough picture of individual cow infection status.

"Cell count readings are usually only taken once a month, but the real question is how much variance do you get from day-to-day, and between different cows," says Eric Hillerton from The Institute. "According to our trial, the answer is lots!"

In the study, Institute researchers took milk samples every day for 33 days from a sample of 21 cows and measured the individual cow somatic cell count levels. The variation in values day-to-day was significant, as was the variation between cows.

"We want to be confident that individual cow cell counts are giving us information about infection status and can highlight differences between cows. However, this study flags up high within-cow variability," Eric Hillerton says.

"It means that for individual cow sampling to be worthwhile, it is very important to take several serial measurements over a short time period from individual cows if the cell count is to be truly representative of the individual cow mean. Unfortunately, repeated sampling is costly and not particularly practical."

On the basis of the study, Eric Hillerton suggests farmers and their advisers should build in a margin for error when interpreting individual cow cell count figures when the results are based on samples taken a month apart.

"What we can say is that below a cell count of 100,000 cells/ml the cow is most likely to be uninfected in all quarters. But at 200,000 cells/ml the probability changes to one or more quarters being infected, so always assume a plus or minus 40,000 cells/ml error margin when planning your dry cow management treatment protocol," he advises.