Evira creates model to calculate microbes in food chain

A model for statistically calculating the amounts of microbes in the production chain has been developed by the Finnish Food Safety Authority (Evira).

The model makes it possible to account for the combined uncertainty of factors relating to the risk assessment, conditionally on the number of samples and analysis results. 

It also assesses the impact of limit values on risk that consumers are exposed to. Limit values are determined using microbiological criteria specifying the acceptable level of Campylobacter in samples from broiler meat at various processing stages.

Model for risk-based management

The Bayesian method evaluates and compares effects of microbiological criteria (MC) in broiler production on consumer risk.

Parameter uncertainty was represented by a joint posterior distribution, which was then used to predict the risk and evaluate the criteria for acceptance of production batches.

A fully risk-based evaluation of MC involves several uncertainties related to the underlying Quantitative Microbiological Risk Assessment (QMRA) model and production-specific sample data on prevalence and concentrations of microbes in production batches, said the agency.

Based on the sample data, together with the QMRA model, one could achieve a relative risk of 0.4 by insisting that the default criterion be fulfilled for acceptance of each batch,” it said.

Evira said it is possible to compare optional sets of criteria and identify the optimum solution where the risk and cost are at a minimum by using the model.

For example, this can be done by adjusting the number of samples or concentration levels applied to the samples.

The model can be used with other information for determining how changes in the number of samples affect the accuracy of the assessment.

Uncertainty factors in approach

Jukka Ranta, senior researcher of Evira, said the effect of microbiological criteria on consumer risk can be determined theoretically by samples taken in the production chain.

“Consequently, approval of the production lot for sale to consumers may be based on the measured concentrations of microorganisms and the assessment made on this basis," he said.

"The uncertainty factors include the occurrence of production lots containing microbes and the prevalence of microbes within such production lots as well as the average microbe counts per gram throughout the entire production chain and within each individual lot in such occurrences.

“Variation in concentrations within and between lots is another uncertainty factor.”

The work was part of an inter-Nordic project using Nordic data on Campylobacter levels in broiler meat production.

Evira's task was to develop a Bayesian approach to the determination of microbiological levels in foods as a probability distribution based on samples.

Shifting from traditional hazard-based management toward risk-based requires statistical methods for evaluating intermediate targets in food production, such as microbiological criteria, in terms of effect on risk of illness.

The analysis was illustrated with two data sets that contained partial but complementary evidence.

Pork meat testing

Meanwhile, antibiotic-resistant MRSA bacteria may be found in Finnish pork meat, according to Evira.

Evira is analyzing 300 samples of fresh pork meat for the methicillin-resistant Staphylococcus aureus (MRSA) bacterium. 

Of the 275 samples mostly of domestic origin analysed by November, six were found to contain MRSA, of which five were Finnish meat.

The remaining results will be ready by the end of January 2016.

MRSA bacteria are destroyed when pork is heated to at least 75°C. Transmission of MRSA from meat to people is not considered to be a major risk as long as it is handled hygienically and cooked thoroughly, said the agency.

Source: The Annals of Applied Statistics

Online doi: 10.1214/15-AOAS845

A BAYESIAN APPROACH TO THE EVALUATION OF RISK-BASED MICROBIOLOGICAL CRITERIA FOR CAMPYLOBACTER IN BROILER MEAT”

Authors: Jukka Ranta, Roland Lindqvist, Ingrid Hansson, Pirkko Tuominen and Maarten Nauta