The study highlights the important role that good kitchen hygiene practice has in preventing infection through cross contamination, and the need to clean surfaces and utensils thoroughly.

But the problem of biofilms extends beyond the kitchen, and is in fact an issue at many points in the production chain. It is one of the reasons why campylobacter is proving such a tricky customer.

One such area of focus for the poultry industry is crates and modules. These are pieces of equipment used in the catching process. Modules create a drawer-like structure into which multiple crates are placed. This kit shuttles between farms and the slaughterhouse and is a potential route of how campylobacter gets onto sites.

Getting equipment such as this visually clean is a minimum standard applied across the industry, but it has been found that campylobacter can survive in a biofilm similar to that described in the IFR paper. So what does that mean for crates and modules?

Traditional design of this equipment has meant that new ideas are needed in both their use and their cleaning. In their use we’re seeing new practices that minimise the time that crates and modules spend in and around chicken houses, but this is always going to be compromised by the amount of faecal matter chickens produce during catching and transport.

So cleaning of crates and modules becomes even more important. Work is underway on novel washing systems that disrupt the biofilm as well as removing the visible dirt. Trials are also ongoing on using higher temperatures in the water used for cleaning. Previously the plastic of the crates warped or became brittle if the temperature exceeded a certain value.

Ultimately this part of the effort against campylobacter is going to require a combination of novel design, new cleaning techniques, and good practice in use.

What we’re interested in is whether work from other areas can contribute to our understanding in this specific one. If you think you can help please get in touch.

Project outline

The project consisted of 3 stages

• Off line application of steam to carcases assess effect on campylobacter.
• Building a tunnel with steam jets on the processing line.
• Assessment of carcases after steam treatment for customer acceptability.

A steam chamber was constructed around the main process line just after the inside / outside washer and prior to entering the main in-line chill.  With the steam switched on the impact on product quality (particularly skin structure) and microbiological loading was assessed.

The machine was in operation for approximately 5 minutes and achieved an operating temperature of 90^oC. The birds were in contact with the steam for 5.2 seconds. 5 micro samples were collected pre-steam treatment and 5 samples were collected post steam treatment samples were sent to the lab for campylobacter, TVC, coliforms and pseudomonas analysis.

10 samples were visually assessed pre- & post- steam treatment and post-chilling and trussing; this was to visually assess the birds for any physical defects.

Results

Results showed that at best the treatment had no impact on the microbial loading of the neck skins.

Campylobacter levels were slightly higher on the post-treatment samples, but this was not statistically significant.

It was noted that product which underwent steam treatment had very ‘delicate’ skin and when collecting the product it was noted that the skin was very ‘fragile’. During trussing the skin was torn in the thigh area and the vent.

The microbial benefit of this treatment on campylobacter and other common microflora was not realised during this trial.  This point, coupled the amount of skin damage and disruption, would render the treated whole birds as unsuitable for retail sale in the whole bird, or skin-on portion formats.

Conclusion

Steam is an effective microbiological decontaminant. However, applied directly to the carcase during processing, steam did not provide any significant microbiological reduction.

The organolepic effects on the carcase resulted in changes to the colour and consistency of the skin. The skin was significantly damaged/torn during further processing and handling. The birds were assessed as unacceptable for retail sale.

Project outline

• 5 birds were left with standard neck flap length and 5 with neck skin totally removed. Levels of campylobacter were ascertained using a whole carcase rinse method and testing via ISO 10272
• Work was undertaken to assess the acceptability to the consumer of carcases with short neck skin. Neck skin was removed at various lengths and assessments made of the raw product in pack and the cooked product
• Trial 2 was conducted on 9 carcases with a neck skin length 3.5 cm. Levels of campylobacter were ascertained using a whole carcase rinse method and testing via ISO 10272
• To assess results across the shelf life of the product, 3 carcases were tested on each of the following: Pack day +1, Pack day +5, Pack day +9.
• Trial 3 was conducted on 40 carcases to reduce any variation due to bird size and effectiveness of the total carcase wash in removal of bacteria. A more sophisticated and controlled method of carcase rinse was used and the results expressed as cfu/cm² carcase using a carcase weight to carcase surface area conversion using the formula reported by Thomas.(Thomas N L 1978 J Food Technol 13 81-86)

Results

First Trial

Initial work carried out with complete removal of neck skin indicated a potential reduction in total campylobacter using a whole bird rinse method. The results indicated a potential decrease in Campylobacter of 0.9 log by comparing mean log values between the two groups.

• No neck skin:- mean count 632 cfu/ml rinse
• Standard neck skin:- mean count 4740 cfu/ml rinse

Second trial

• 1 (packing +1 day) / 750 (standard neck skin, cfu/ml) / 300 (short neck skin, cfu/ml)
• 1 / 5000 / 750
• 1 / 190 / 1300
• 5 / 280 / 170
• 5 / 130 / 60
• 5 / 160 / 160
• 9 / 60 / 100
• 9 / 840 / 500
• 9 / 1800 / 200

Although the bird appearance was acceptable prior to cooking, it was assessed as less so after cook because the neck cavity was exposed and the skin did not cover all the breast meat resulting a reduction in meat quality due to possible drying effects. 5cm was considered the shortest acceptable neck skin length.

Third trial

Standard neck skin (cfu/cm² carcass): 4, 13, 49, 55, 74, 87, 151, 166, 218, 306, 653, 656, 656, 665, 680, 697, 705, 912, 925, 1118

Short neck skin (cfu/cm² carcass): 22, 52, 75, 107, 130, 141, 143, 152, 484, 638, 643, 651, 661, 715, 744, 817, 1069, 1874, 4712, 21927

The results were statistically analysed by a two-way analysis of variance. This indicated that any improvement, i.e. lower counts, associated with short neck skins was tiny. There was weak evidence that carcases with short neck skins had a slightly greater risk of carrying campylobacter.

Conclusion

The results indicated that reducing the neck skin length on a chicken carcase to no more than 5cm did not impact on the risk associated with the level of campylobacter on whole chicken carcases.

Project outline

Electrolysed water is produced by applying a low voltage charge to water containing a salt.

A hydroxide compound is produced which acts as a disinfectant to kill bacteria. It is rendered inactive during the disinfection process or over time. Electrolysed water is safe to eat and is used in the food and other industries. A series of trials and their controls were conducted on the same flocks of birds.

Neck skins were collected for testing before and after the in-line carcase washer from birds going through the normal washing process using potable water. The machine that produces the electrolysed water was activated and neck skin samples were collected for testing before and after the in-line carcase washer supplied with the electrolysed water.

Results

Initially machine performance was unreliable in producing a steady supply of electrolysed water but further work was effective in addressing this problem.

The results showed that there was no evidence in chicken that campylobacter levels were reduced either immediately after washing or when the birds had been stored for the period of their designated shelf life.

Conclusion

Using electrolysed water in the carcase wash is no more effective in reducing the level of contamination on the carcase than using potable water.

Project outline

To implement agreed biosecurity controls on selected poultry farms and assess their effectiveness at excluding campylobacter from the flocks or reducing the number of bacteria on carcases.

Measurement of the level of contamination was by enumeration of campylobacter in the caeca and on the neck skins.

The project analysed the results of on farm interventions from 16 farms (9 model farms) looking to provide a robust basis on which to base decisions to roll out the effective interventions to all sites. The farms were selected to represent the range of shed types and ages found in UK farms. Standard and enhanced welfare farms were selected.  Each retailer in the project each had at least 1 model farm supplying their product. All major production areas in the UK are represented. The information was generated from both small and large units and contract and company farms.

The results from the model farms were compared to farms populated at the same time and with the same welfare status. Outcomes were measured by comparing the campylobacter enumeration results in the caeca with the results from the random weekly routine test results. The project ran for 2 years from 1^st September 2011 collecting data from 16 crops.

Farm types

Of the 16 farms chosen:

• 10 were enhanced welfare and 6 were standard welfare
• 3 in Northern Ireland, 4 in East Anglia, 2 in the Midlands, 3 in the South West, 3 in Scotland, 1 in Wales
• There was an average of 5 sheds per arm (min. 1, max. 10)

Interventions used

• Standard biosecurity for control farms: Red Tractor assurance scheme for poultry farms
• Additional biosecurity for model farms: Each shed was a biosecure unit with all protective clothing and equipment being shed specific
• Thinning: Additional clothing changes and biosecurity practices during thinning, including module, crate, and transport initiatives
• Depopulation: Clearing sheds within five days of thin to reduce campylobacter load if introduced during thinning
• Shed rest period: Extending the time between flocks while not compromising commercial operations

Each farm was visited and the biosecurity assessed. Regular visits were mad by area managers and the results were provided to the farmers.

Sampling regime

The following tests were carried out to measure the level of campylobacter and assess the success of biosecurity interventions. All testing was done to the ISO standard for enumeration (10272:2) at certified laboratories. Testing of caeca involved 5 samples pooled into 1, and neck skin testing 3 samples pooled into 1.

• Caeca from birds of every model shed at thin (approx. 357 tests per year)
• Caeca from birds of every model shed at depopulation unless campylobacter status is known to be positive (approx. 315 tests per year)
• Neck skins from every model farm (approx. 91 tests per year)

Results

• The data was collected by Industry on 883 crops from 16 farms over 2 years.
• Three quarters of crops have >1000 cfu/g colonisation, while a quarter of flocks have <100 cfu/g colonisation (caeca samples at thin and depopulation)
• There was marked seasonality
• 50% of those found negative at thinning were positive at depopulation (further analysis is being carried out on this)
• There was a very slight significant association between a period of more than 5 days between thinning and depopulation and the crop being positive. However, the data set was quite large which could have the effect of picking up associations which may not have much effect biologically
• The proportion of high risk crops (days between thinning and depopulation >5 days) if reduced to under 5 days will decrease the numbers of positives, but only by 12%. From these results stopping thinning would decrease the Campylobacter colonisation rate-but only by 12%.
• An increase in the number of days that sheds are left empty are strongly linked to a decrease in the caecal campylobacter count at thinning (more than that seen at depopulation)
• There was a trend in lower values of caecal campylobacter count at thin with increasing days of sheds left empty (needs to be confirmed that this is not an outlier (e.g. shed rest days > 20 days) were excluded)

Conclusion

The bio-security standards used on the model farms was unable to totally protect the chicken from colonisation with campylobacter. The project demonstrated that not all sheds on a farm became positive with campylobacter when good bio-security was applied between sheds. In farms using the higher bio-security standards fewer sheds became positive.

Thinning is the term used when some birds are removed from a shed, usually 25-30% and the remained left to grow on. High level bio-security is broken during thinning. Stopping thinning and completely depopulating sheds may result in a reduction of around 12% reduction in campylobacter positive sheds. The results were obtained from a small number of farms and further work is being undertaken to evaluate the effects of stopping thinning.

Further detailed analysis is being  undertaken by an independent organisation.

Project Outline

Chlorine Dioxide ClO₂ up to a maximum level of 0.5ppm was added to the water used for the washing of carcases during processing, and comparing the effect against a previous baseline study using potable water.

A minimum of 1 flock per week was sampled.

The neck skins of the carcases were tested for both Campylobacter and for the total viable count (TCVs).

Results

The addition of ClO₂ in wash after evisceration and to the inside /outside wash and ISOS wash made no significant difference to microbiological status of the neck flaps of birds results compared with the baseline study.

ClO₂ was also introduced in the pre-chill water for approx 6 weeks. The results of these flocks appear to show a small improvement compared to baseline but this was based on a small number of flocks and is not statistically significant.

An independent assessment of impact of chlorine dioxide treatment also found no evidence of an effect of the addition of chlorine dioxide to the water in the supply tank on Campylobacter counts.

Conclusion

The addition of Chlorine Dioxide at 0.5ppm to mains water for carcase washing does not reduce the enumeration of Campylobacter on the neck skins.