Project title:    Standardisation of sampling and analysis methods in poultry abattoirs in support of HACCP-based hygiene strategies.

Duration:         October 1st 2001 to May 31st 2005

Background and Objectives

Food production and processing in the European Union is about to undergo radical change as a result of European Union (EU) commission decision SANCO 4198/2001.  The legislation implemented as a consequence of this decision will oblige all food producers and processors within the EU to introduce safety procedures aimed at reducing hazards in their food manufacturing plants; thereby ensuring the safety of EU-manufactured food products for consumers.  Microbiological testing of food products will need to be undertaken to ensure that any hazard-control schemes that are put in place are effectively monitoring production processes.  This study was commissioned to identify appropriate microbiological process control indicators for poultry slaughterhouses, identify appropriate ways for carcasses sampling and set process performance limits for these indicators.

 Approaches

Industry interviews.  The project commenced with face-to-face interviews with 20 poultry slaughterhouse operators.  The interview were used to gather information describing the microbiological testing that is currently undertaken by UK slaughterhouses.  Information was recorded using a standardised questionnaire which simplified comparisons of results collected from different plants.  Plants were canvassed for their current practices to determine if there was a consensus in sampling and testing methods across the industry that could be adapted to suit the new regulations.  Although such an approach would cause minimum upheaval for the industry, there was no widely-used sampling and testing method at the plants interviewed.  Some individual plants owned by a common parent company also lacked standardised sampling and testing methods for carcasses.

Plant sampling. Since there were no standard protocols in use across the industry, studies to determine the appropriateness of bacterial indicator numbers on poultry carcasses for process verification purposes were undertaken in commercial slaughterhouses.  Sampling was by neck skin excision or whole carcase rinse.  Samples were analysed microbiologically for a range of potentially useful process hygiene indicator bacteria including total aerobic viable counts (TVC), Enterobacteriaceae, E. coli, Pseudomonas, and coliforms.  The variation was calculated for each bacterial indicator and found to be significantly lower in excised samples indicating a more consistent recovery of bacteria by this sampling method.  TVC, Enterobacteriaceae and Pseudomonas in samples collected by excision had the lowest variation when compared with other indicators and were therefore taken forward and used for further study. 

National processing plant survey.  In order to define a representative range for the numbers of indicators on carcasses in UK poultry slaughterhouses, a national snapshot of bacterial numbers was undertaken.  Samples were collected from 18 matched plants spread throughout the UK in the summer and winter seasons.  Although considerable variation in numbers was observed for the indicators in different plants, no significant differences in indicators was determined between the different seasons.

Trends over time.  To more fully investigate the possibility of seasonal bias, the bacterial levels on carcases were followed at a typical medium-to-high throughput plant by weekly sampling for almost a year.  No seasonal trends, or significant changes in the bacterial indicator numbers on carcasses were observed for this plant during this time period despite a wide variety of changes in plant hygiene during this time.  To determine whether this finding was typical, three other plants were visited and bacterial numbers sampled at periodic intervals after the start of a days processing.  Plant hygiene deteriorates visibly as the days’ processing continues.  However we did not observe much change to the bacterial numbers as process hygiene diminished.

Laboratory analyses. The errors associated with the measurement of bacterial numbers for each indicator were calculated from dual analyses of single samples in a laboratory.   Errors were lowest overall for TVC.  In general, measurement errors were higher when lower bacterial numbers were being measured.  Using five neck skin samples at once for microbiological analyses was found not to be significantly different to the mean of individually-analysed samples.  Thus bulking samples together to save money on analyses costs is appropriate for neck flap samples.  Bacterial numbers increased by a single log when stored under chilled conditions similar to those used for sending samples to external analyses laboratories.  The increases were not statistically significant for Pseudomonas and TVC when storage time was less than 17 hours. 

Additional studies.  The Food Standards Agency requested we undertake additional work after this study commenced.  The additional studies concerned Campylobacter numbers on poultry during processing.  Campylobacter is the bacterial species which causes the most food borne illness in the UK.  The additional studies showed that a new method for the determining Campylobacter counts is not reliable when low numbers of Campylobacter are present.  The additional studies also clearly showed that Campylobacter contamination of poultry carcasses in processing plant across the UK is significantly lower in the winter when compared with summer.

What it means and why it’s important

Only weak relationships were identified between bacterial indicator numbers and the length of time of processing, although visible cleanliness of the processing environment diminished during this time.  Therefore we conclude that, at the plants visited for this study, there was a poor relationship between presumptive bacterial indicator numbers and actual process hygiene and consequently bacterial analyses for process verification purposes in poultry slaughterhouses may not be appropriate.

Project publications (click here to request a paper from the authors by email)

Hutchison, M. L., Walters, L. D., Allen, V. M., Mead G. C. and Howell, M. 2006. Measurement of uncertainty of numbers of Campylobacter in British poultry slaughterhouses. J. Food Prot. 69: 421-424.

Hutchison, M. L., Walters, L. D., Mead, G. C., Howell, M. and Allen, V. M. 2006. An assessment of sampling methods and microbiological hygiene indicators for process verification purposes in poultry slaughterhouses. J. Food Prot. 69:145-153.