Study 1 (Pilot)

Report prepared by:- J de Louvois and G Nichols

INTRODUCTION

In 1993 the Public Health Laboratory Service and the Local Authorities Coordinating Body on Food and Trading Standards (LACOTS) agreed to undertake joint surveillance projects on foods which had been proposed or identified by local Food Liaison Groups as giving cause for concern. This joint exercise was to be known as the LACOTS/PHLS Coordinated Food Liaison Group Microbiological Sampling Programme. Two studies which would act as pilots for future projects were planned for 1994/5. The first was undertaken in April/May on ready-to-eat take away chicken pieces from fast food establishments.

Food Liaison Groups were invited to submit up to 100 samples of cooked ready-to-eat take away chicken pieces which had been collected from fast food establishments to the laboratory they normally used for the microbiological examination foods.

Samples were collected from any fast food take-away catering establishments which provided cooked ready-to-eat chicken, including fish and chip shops and burger bars. Mobile and stall based caterers such as those in markets, at football matches, fairs and in lay-bys were also sampled. The study did not include pubs, supermarkets, schools, works canteens/restaurants, vending machines and garages offering a microwave service. Sampling officers were asked to take samples from a representative cross section of appropriate establishments in their area. No attempt was made to structure the study so as to represent market share or retail size, since accurate information on these matters was not available.

NUMBER AND TYPE OF SAMPLES

Each participating Food Liaison Group was invited to submit up to 100 samples during the study period. A sample consisted of at least 150g of cooked, ready-to-eat chicken. Chicken coated with batter, breadcrumbs, barbecue sauce etc was acceptable. The study did not include compound chicken products such as chicken nuggets, pies, pasties, chow mein, curry etc nor products incorporating fillings such as chicken kiev. Samples were collected at the point of sale as if being purchased by a customer.

SAMPLING TIMES

EHO’s were requested to collect samples at times when the business was characteristically busy, for example, peak lunchtime trade.

METHODS

SAMPLING METHOD

The sampling method used for this study was that detailed in the Food Safety Act 1990. Code of Practice No. 7:- Sampling for Analysis and Examination. Particular reference was paid to Part III, "Samples for Examination". At the time of collection a standard proforma was completed (Appendix A).

Samples, which were served by the vendor according to his usual customer practice, were placed in a suitable sterile container. Prior to closure the surface temperature of the sample was taken using a sterile food temperature probe. Once secured and labelled the sample container was placed in a cool box or mobile refrigerator at 0-5^oC. Hot samples were transported in a separate cool box from chilled samples. The protocol required samples collected during normal laboratory working hours to be delivered within 2 hours and only in exceptional circumstances after more than 4 hours. Samples taken outside normal laboratory hours were delivered to a holding refrigerator at 0-5^oC for examination later.

LABORATORY PROCEDURES

The detailed laboratory procedure is attached as Appendix B.

In brief the sample was prepared by separating chicken flesh, including skin and any other coating, from bones, cutting the meat into small pieces and stomaching 25g aliquots in either buffered peptone water or Campylobacter broth. The aerobic plate count (APC) was undertaken at 30^oC according to British Standard 5763 using either a spiral plater or surface spread method. The method used to detect Salmonella species was based on BS 5763 Part 4 with brilliant green agar and another selective agar medium used for selection. The Campylobacter method was based on that of Humphrey and Muscat (1989) which has recently received approval from the US Food & Drug Administration (FDA).

REPORTING

Results on each sample examined were returned to the Environmental Health Department concerned. The completed questionnaires were collected by Environmental Services (PHLS Headquarters) for analysis. Local authorities using Public Analysts, in-house or independent laboratories were asked to send their results to LACOTS. These were then passed on to the PHLS, for inclusion in the analysis of results.

The results presented in this report are those for the study as a whole. Each Food Liaison Group will receive tables containing all the results and also information on samples which their particular group submitted to enable them to compare local and national data. Any conclusions relating to the results from a single Food Liaison Group will be appended to the individual report.

RESULTS

Fifty-two Food Liaison Groups submitted a total of 4055 chicken samples to their local laboratory (50 PHLS laboratories, 2 other laboratories) during April/May 1994. The majority of samples came from chip shops (61% (Table 1) and consisted either of chicken quarters (40%) or chicken breast (23%) (Table 2). The means of cooking is recorded in Table 3, from which it can be seen that 65% of samples were deep fried. This represents the predominance of samples collected from fish and chip shops. In many instances these samples had already been cooked by other means and were immersed in hot fat for a short period in order to heat them prior to serving. The aerobic plate count for all samples as assessed on a 25g sample of chopped chicken meat is recorded in Table 4. In 62% of samples the aerobic plate count was less than 100 organisms per gram. In 5% of samples the count exceeded 10⁵ bacteria per gram. A comparison of the aerobic plate counts of samples from different retail outlets is shown in Table 5. Camplylobacter species were isolated from 1 of 4055 samples. Salmonella was not isolated from any samples.

DISCUSSION

The isolation of the pathogens sought from only one of more than 4,000 ready-to-eat retail chicken samples from high street retail outlets indicates that this food is not a major public health risk. It is noteworthy that the only sample from which a pathogen was isolated was reported to be significantly undercooked. It is of concern that 5% of samples had bacterial counts of 100,000 bacteria per gram or more and would therefore be considered as unsatisfactory according to the PHLS provisional microbiological guidelines on ready to eat foods¹. It is probable that these high counts arose as a result of poor hygiene and handling practices of chickens once they had been cooked. Since these counts refer to a mean figure of all edible portions of the samples examined, it is probable that had the surface coverings and skin of samples been examined separately there would have been a higher proportion of samples with high bacterial counts.

Samples submitted to one laboratory had a disproportionate number of aerobic plate counts in excess of 10⁵ and in the light of the overall results the local Food Liaison Group and Public Health Laboratory Director are carrying out further investigations.

The majority of samples with high bacterial counts came from outlets under the heading "Other". This group of retail outlets includes Chinese, Indian and other ethnic take aways and butchers selling cooked chicken. Based on the results of this study it would seem reasonable that future local investigations of ready-to-eat take away chicken etc might with benefit concentrate on these outlets, and that the sampling protocol might provide for the separate examination of the outer surface (skin/breadcrumb covering etc) and chicken meat. While relatively few samples were received from butcher’s shops, this might also be an area worthy of further investigation because of the inherent problems associated with the sequential handling of uncooked and ready-to-eat meats.

CONCLUSION

This first pilot study has demonstrated that "bottom up" studies involving foods proposed by local Food Liaison Groups can be operated efficiently and that there is a high level of support for such studies. With the use of a common protocol and uniform laboratory methods, based on established standards, the data from such studies can rapidly provide information on the microbiological status of individual foods from a variety of retail outlets in England and Wales which will be of value to local authorities in targeting their resources and undertaking risk based inspections.

ACKNOWLEDGEMENTS

We are grateful to our laboratory and environmental health colleagues for their enthusiastic support for this new approach to structured food sampling. We wish to thank Janet Norcup for analysing the data and Ros Lewis for preparing the report.