On May 1, the FSIS announced Salmonella as an adulterant specifically stating that “not ready-to-eat (NRTE) breaded stuffed chicken products that contain Salmonella at levels of 1 Colony Forming Unit per gram or higher are adulterated within the meaning of the Poultry Products Inspection Act (PPIA).” I’ve already written other blogs and have spoken publicly many times about how this will have zero impact on public health metrics and why so I won’t go into great detail in this blog, but briefly, we are not addressing the primary sources of Salmonella and we are not considering the highly pathogenic strains of Salmonella which matter. There is no science or legitimate risk-based data to drive this decision. As a scientist, I know we can do better to address Salmonella. As the industry, you know we can too.

Something I have not discussed yet is the fact that the agency claims that this is a quantitative standard meaning that there has to be a certain number of cells of Salmonella in order for the product to be “adulterated”, but it is indeed not a quantitative standard. I am very disappointed in the fact that a public document made it through all of the vetting of FSIS, USDA and OMB with a public standard of “1 cfu/g”. The agency has an amazing team of scientists. I will say that they are the best and I have a high amount of respect for them. While I do not know for sure, knowing the strength of the scientific team, I know they would have caught this very obvious flaw in this announcement had adequate time been given for thorough review and vetting.

Why is this a problem? I am a scientist, but also a teacher and cling to basic microbiological rules from sources such as the FDA-BAM, the FSIS-MLG and Bergey’s Manuel of Determinative Microbiology thus it is important to address accuracy and precision so let’s review a few things. First of all, let’s define a “cfu” as a “colony forming unit”. The reason it is called a colony forming unit is the fact that on a plate of microbiological media, a colony is formed and the theory is that the colony arises from one cell. A cfu could also come from multiple cells, not one if they are clumped together. A cfu must be measured on a plate of microbiological media after it has been incubated and grows on the plate, not by any other method. In other words, a colony must be formed. It is inaccurate to use the terminology “cfu” in any other context from a scientific perspective. An alternative standard method exists for quantification, the MPN, or most probably number method. An MPN is the nightmare of any graduate student because the labor is rough. MPNs are designed to quantify bacteria when the numbers are below that of a countable plate. They work and are very accurate, but are not reported as “cfu”. They are reported as MPN/g and they are an estimate.

I am very well aware of the methods used by the agency and the industry that result in rapid quantification of Salmonella. As a matter of fact, I strongly support, use and embrace these methods. They are awesome! For these methods, an estimate of a cfu or MPN is developed by making a curve that correlates to the cfu or mpn resulting in an estimated cfu range, not an exact number (ie 1 cfu/g). The result comes with a standard error and range of values, not just an absolute number. I want to emphasize that these methods measure the amplification of DNA through PCR, they are not measuring a colony and thus they are not a “cfu”.

Another basic rule of microbiology, with a few exceptions, is that a microbiological plate must have 25-250 colonies to be considered “countable”. If the plate has less than 25 colonies, again, it is an estimate. One could take a larger portion of the sample, perhaps 100 g and dilute it, but if it really had 1 cfu/g, it would never achieve a countable number on a plate. You can do the math, but only 1 ml can be plated at a time and thus, it is theoretically impossible to get 1 cfu/g without estimating the number if it is a solid sample. Liquid samples can be filtered and I will argue that if the sample is liquid then bacteria can be collected on a filter and plated accurately to determine a value of 1 cfu, but raw, breaded, stuffed chicken products are not liquid.

The bottom line is that Salmonella is a public health problem and it has to be addressed with real data and real science to prevent illnesses and deaths. Not only will this make no difference in public health metrics, it is not even accurate or possible to measure. We can do better.

Why CFUs Matter

Join for free* to continue reading.