Use of a slaughter hygiene indicator (Escherichia coli) to quantify the risk of human salmonellosis related to pork in Denmark – an approach for risk based meat control?

Abstract

Food chain information does not per se allow an effective distinction of herds according to shedding of Salmonella. Thus, no effective sorting of pigs at slaughter according to Salmonella risk is possible and hygiene improvement is the only effective mitigating tool so far at the slaughter. From a large study of 1906 slaughter pigs we quantified Salmonella and established quantitative hygiene data (E. coli) on pig carcasses (paired data) at slaughter. Based on the results, we found a positive correlation between level of E. coli and the prevalence of Salmonella positive carcasses. The odds ratio for Salmonella being present on the carcass was found to increase by 1.87 for every one log10 unit increase of E. coli found on the carcass. A simple Salmonella consumer risk model was constructed using the observed levels of E. coli contamination as input and the model, established a positive correlation between slaughter hygiene (E. coli) and consumer risk. Further, we analysed two years’ own control data on Salmonella (prevalence data) and E. coli (quantitative data) from four large slaughterhouses in Denmark and found a similar positive correlation between the E. coli level and the carcass prevalence of Salmonella. The aim of this study was to propose a principle for setting risk based hygiene targets on E. coli on carcasses at pig slaughter. As such we provide input to the discussion on how to develop a risk based meat control procedure, based on statistical process control. Introduction The Salmonella is widespread in the slaughter pig production in Europe, and Salmonella from pork constitutes a significant risk for consumers. In recent years the ability of the classical meat control to provide consumer protection against food borne pathogens, has been discussed. A consensus of basing the meat control on food chain information of the pathogens is emerging in Europe. Food chain information do however not per se allow for an effective distinction between pigs from Salmonella positive and negative herds, and improvement of the general slaughter hygiene is the only mitigating tool to use. A modernisation of the meat control could include a risk based statistical process control but so far there have been no reports describing how the hygiene level at slaughter associates to Salmonella risk. We have established quantitative hygiene data (E. coli) and quantified Salmonella on pig carcasses at from 2880 pigs slaughter. Moreover we have analysed two years own control data on E. coli and Salmonella from five large pig slaughter houses in in Denmark. The objective is to establish the correlation between the hygiene level and presence of Salmonella and to provide the first suggestion for a method to set risk based process hygiene criteria at pig slaughter. Material and methods Sample collection 1906 carcasses from pigs slaughtered at five large Danish pig slaughterhouses were sampled in the period May 2005 to June 2007. Carcass swabs (2800 cm2) from was taken just before cooling and analysed both quantitatively for E. coli and semi-quantitatively for Salmonella. A total of 75 ml peptone water was added to stomacher bags with carcass swabs containing approximately 12.5 ml of peptone water and tissue fluid. The sample was stomached and one millilitre of 10-fold dilutions were spread on Petrifilm and subsequently incubated at 41.5 °C for 23-25 h. The number of E. coli was determined using Select E. coli Count Plate Petrifilm (3M Microbiology, St. Paul, MN, USA) in accordance with the supplier’s instructions. Cell counts were determined by automated reading using a Petrifilm plate reader MI649 9 (3M Microbiology, St. Paul, MN, USA). From a the ten-fold dilution of the homogenate, a semi-quantitative analysis for Salmonella was performed. All stomached samples were analysed for Salmonella using MSRV agar (ISO 6579, Annex D, Anonymous, 2007). Own control data from the same five slaughter houses were obtained based on swabbing of 300 cm2 mandatory for slaughter-houses exporting to USA. Each day one pooled sample of five swab samples were analysed for Salmonella and one sample per 1000 carcasses were analysed for E. coli with a range of 5-12 samples per day depending of slaughterhouse.