Jeanne-Marie Membre - Microbiological Risk Assessment Associated with the Food Processing and Distribution Chain

The objective of this study was therefore to explore the possibility of identifying heat treatments milder than 10 minutes at 90°C, integrating this latency time with the classic thermal inactivation effect. An exposure assessment model, based on the concept of the DoP, was developed.

The principle of the approach used to calculate the DoP, combining the thermal inactivation effect and the thermal damage effect, is based on similar studies previously presented in the literature (Hauschild 1982; Lund 1993; Schillinger et al. 1996). Details are provided in equations [1.1]– [1.4]:

[1.1]

with ΔR being the decimal reduction due to thermal inactivation and ThI the DoP due to thermal stress. Both are expressed as the decimal logarithm of the reciprocal of a probability:

[1.2]

[1.3]

Pr indicates the probability that a spore will survive thermal treatment and Pi the probability that the latency period will be shorter than the SL:

[1.4]

Once the model has been built, the results can be set out in two ways. The first is to present the DoP, including thermal inactivation and stress, for different processes, formulations (pH and aw) and refrigerated storage conditions at different shelf lives.

The second is to choose a targeted level of protection and present the model results as a set of combinations of processes, formulations, storage conditions and SLs that achieve that target protection. This second presentation of the results can be compared to an isoprobability method, because the DoPs are derived from probability calculations. The second approach was the one chosen in this study. An overall DoP of 6 was chosen because it is a reference value currently applied in the management of food safety, in the form of inactivation (Gould 1999). Work by Membré et al. (2009), which led to this approach, along with bibliographical references and a number of illustrations are presented in the Appendices.

It was shown that the hazardous C. botulinum was controlled in a product with a pH of 6.0 kept in a cold place for 25 days, despite the thermal processing being reduced to 85°C for 10 minutes. These results were obtained by taking into account the potential variability of refrigeration temperatures at the premises of both the distributor and the consumer. Probabilistic approaches were implemented. Such approaches are discussed again in the chapter on exposure assessment and in the chapter on risk characterization.

A Web of Science search (January 2021) using the query “TITLE: (hazard AND identification) AND TOPIC: (microbial OR microbiological OR microorganism*)” first of all revealed the very small number of scientific articles in the field of hazard identification. Indeed, since 1950, there have been only 61 articles which directly address (in the sense that the keywords are in the title) the topic of microbiological hazard identification.

Once the off-topic articles had been discarded, and focusing on the most frequently cited articles (10 times or more), only 16 articles were left ( Table 1.2). Their detailed analysis sheds valuable light on both the most frequent object of study and the new trends/methods used in hazard identification.

Hazard identification was most often done as part of HACCP systems or risk (or exposure) assessment, which is consistent with what has already been said in section 1.2. Finally, the articles reporting on molecular approaches were published from 2015. The two cited (Franz et al. 2015; Pielaat et al. 2015) were identified in the literature search because the term “hazard identification” was in the title of the article, but there are other articles that deal with the use of molecular methods and in particular with WGS in hazard identification; see, for example, the review by Rantsiou et al. (2018).

As noted in the Introduction, hazard identification is the first step in microbiological risk assessment. This chapter does not develop examples; they are merely provided to illustrate the process. The reader is invited to consult the first book in this series (Haddad 2022) to see numerous examples of hazard identification.

Table 1.2. List of 16 scientific papers dealing with the identification of the most frequently cited hazards (based on a search of the Web of Science database in January 2021)