Managing time-temperature risk in fresh produce using predictive models for Listeria

N/A

N/A

Many foods are perishable and require time/temperature control throughout their shelf life. In many cases this control is required to ensure quality, but in an uncertain number of situations, it may also be required for food safety (to control foodborne pathogen growth). This issue is often discussed by fresh produce producers and buyers and is now exacerbated by several federal regulations and policies including the Preventive Controls Rule and Sanitary Transportation Rule. There is an urgent need for short-term science-based parameters on this topic. This project focuses on the pathogen most likely to grow at the temperature range of interest (Listeria monocytogenes) and will use "off-the-shelf" computer models in the form of ComBase Predictor (https://browser.combase.cc/ComBase_Predictor.aspx?model=1. This progress report has a variety of predictions, comparing relative risk of Listeria monocytogenes growth for different conditions to guide science-based risk management decisions. This preliminary work will not consider the effect of spoilage organisms or competitive microflora. This work will also not explicitly consider whole versus cut produce, although the availability of moisture and nutrients are known to affect bacterial growth rates. This work assumes that the predictions from ComBase are correct. Since these predictions are based on growth of microorganisms in microbiological growth media without any competitive microflora, such predictions are generally conservative (i.e. organisms grow more slowly in the real world). The model prediction shown below assume that the organism experiences no lag phase, and a very high (0.997) water activity, and that there is no upper population limit, which are all conservative (failsafe) biases. These assumptions will make the models highly conservative, and thus quite robust and able to withstand scrutiny from regulatory agencies or over-zealous inspectors. It is quite likely that actual pathogen growth on specific produce commodities are much less than the model predictions, and in some cases no growth or slow decline would actually be observed in the real world.

Although this project is completed with the submission of this final report, additional research is ongoing in collaboration with a separately funded CPS project with Laura Strawn at Virginia Tech. We are planning to poster/technical presentations at the IAFP meeting this summer covering both data collected as part of a literature review, as well as data collected in the strong lab. Although not part of this report, those data validate that the models proposed here are failsafe (i.e. conservative or airing in the direction of food safety). Additional research on going in the Schaffner lab (not funded by CPS) will further validate these models.