Food Safety and Hygiene, May 2000

A bulletin for the Australian Food Industry May 2000

Contents: Packaged minimally-processed fresh-cut vegetables | Decontamination of fresh fruit and vegetables | Improving the safety of fresh produce | Use of sanitisers at low temperature

Available issues | Search for articles | Index to articles

Packaged minimally-processed fresh-cut vegetables

For almost 100 years, vegetables contaminated in the field have been recognised as a source of human infection and illness. Many of the viruses, bacteria and protozoa on vegetables which have caused food-poisoning are derived from human faeces. These include hepatitis A and Norwalk viruses, Shigella and the protozoan parasite Cyclospora cayetanensis. Contamination from animal faeces can cause non-typhoid Salmonella, enterohaemorrhagic Escherichia coli, and the protozoan Cryptosporidium parvum. Other microorganisms which occur naturally in soil and decaying plant matter are Bacillus cereus, Listeria monocytogenes, and Clostridium botulinum. Much of the micro-flora on fresh vegetables reflects the types present in the growing and harvest environment. However, microorganisms can be added after harvesting, and conditions of storage, transport and display can allow any pathogens present to grow.

Managing the risks

For fresh-cut vegetables that are eaten raw, there is no treatment that can be relied on to substantially reduce the numbers of contaminating micro-organisms. Washing with antimicrobial compounds, while important, often brings about only a relatively small reduction. Eliminating the risks is difficult. Management of them is based on identifying and controlling those factors that are important in preventing contamination or limiting growth of pathogenic microorganisms between farm and plate.

A number of guidance documents has been developed. The advice in such documents as Food Safety Guidelines for the Australian Fresh-cut Produce Industry and the International Fresh-cut Produce Association's guidelines for fresh fruit and vegetables, is based on principles of HACCP, Good Agricultural Practices, Good Handling Practices and Good Hygienic Practices. Sources of contamination

Direct or indirect contamination from animal and human faeces can occur at many points in the fresh-produce chain. Pre-harvest contamination can come from irrigation water, improperly composted manure used as fertiliser, faecal contamination from feral and domestic animals and from workers. On harvesting, contamination may be caused by improperly cleaned harvesting equipment and transport. After harvesting, contamination may occur from wash water and cross-contamination from other vegetables.

During processing, proper storage, control of temperature and hygiene are important in limiting contamination and growth of pathogens. The quality of the water used in production and processing is of prime importance since water can contaminate, or spread contamination through, a large amount of product. Pathogens can be transmitted to product handled by infected workers. Cutting and shredding can be a major source of in-plant contamination since the equipment used can be difficult to clean and has been identified as an important site for the accumulation of pathogens, particularly of L. monocytogenes.

Effects of chilling and short shelf-life

The short shelf-life and refrigerated temperatures at which packaged fresh-cut vegetables are held limits the range of pathogens that can grow. Bacteria such as Staphylococcus aureus will not grow at temperatures below 7-8°C. Salmonellae, shigellae and enterohaemorrhagic E. coli will not grow at this temperature either but are able to survive past the shelf-life of the product.

Though L. monocytogenes is able to grow at chill temperatures, the short shelf-life normally ensures that the extent of growth is limited. On many vegetables, increases in numbers of L. monocytogenes are less than 10-fold after storage for 6-7 days at 5-7°C and on some (e.g. grated carrots and trimmed brussels sprouts) numbers of inoculated viable L. monocytogenes cells decline. Modified atmospheres can increase the shelf-life of fresh vegetable products without affecting the growth of L. monocytogenes. This increases the risk of listeriosis because of the extra time available for growth. Temperature abuse of modified atmosphere packs of vegetables can result in anaerobic conditions within the pack, allowing growth and toxin production by Cl. botulinum. The danger is that botulinum toxin will be formed before overt spoilage, and this has been reported in modified-atmosphere packed shredded cabbage stored at room temperature.

Testing and analysis

It is clearly not possible to test all lots of fresh-cut vegetables for all of the viral, bacterial and parasitic pathogens that might be present. Furthermore, for some (e.g. Cyclospora cayetanensis) methods are limited, not routine and of unknown sensitivity. Even when sensitive methods are used, a large number of samples need to be analysed from each lot to have a reasonable assurance of detecting contamination. If a pathogen were present in every 250 grams of a lot, about 30 samples of 25g would need to be tested to have a 95% chance of detecting this level of contamination. For microorganisms that appear to be able to cause infection at very low doses, such as enterohaemorrhagic E. coli, testing clearly cannot provide sufficient safety.

Analysis for specific pathogens, or even for indicators of faecal contamination, may detect only a high frequency of gross contamination. The sporadic nature of most contamination makes it unlikely that microbiological testing will identify a lot of contaminated vegetables with any degree of certainty. In addition, the time taken for many microbiological analyses is such that it is not possible to institute a test-and-hold procedure for products such as these as a result of their short shelf-life. Thus most of any tested lot of packaged fresh-cut vegetables will have been sold before the microbiological result is known. As the International Commission on Microbiological Specifications for Foods (ICMSF) stated in 1986, "Good agricultural practices in growing crops, combined with acceptable hygienic methods during harvesting, packing, and transporting of vegetables are more important than microbiological testing", and again, "routine microbiological examination of raw vegetables is unlikely to reduce hazards to any great extent and is not recommended".

Control programs

However, while microbiological testing cannot provide assurance for the safety of specific lots, appropriate microbial analysis is important in assessing the effectiveness of control programs. Microbial analysis may also be used as a general measure of the hygienic conditions of growing, harvesting, transporting and processing vegetables.

Choice of microorganism to monitor a control program will depend on the program. For instance, to control L. monocytogenes, product contact surfaces near the end of the processing line may be sampled for listeria to assess overall control of the pathogen in the process environment and the risk of contamination of fresh-cut vegetables. While cutting and shredding equipment might also be examined occasionally for listeria, it is simpler to rely on total aerobic counts to ensure that cleaning and sanitation of this machinery are adequate.

For monitoring general hygienic conditions, total aerobic bacterial counts of vegetables are unlikely to be useful since changes in the aerobic flora can occur which are not reflected in changes in the number or incidence of pathogens. For example, in vegetables stored at 4°C there can be significant bacterial growth without any increase in even the psychrotrophic pathogen L. monocytogenes. Since most of the pathogens of concern are derived from direct or indirect faecal contamination, indicators of faecal contamination are more relevant. Coliforms are unsuitable since they are found in both soil and decaying vegetation as part of the normal flora. They are mainly coliforms of non-faecal origin such as Enterobacter spp. and Klebsiella spp. The normal association of thermo-tolerant coliforms like Klebsiella spp. with vegetables also severely limits the value of 'faecal coliforms' as indicators of faecal contamination. Indole-positive strains of Klebsiella spp. can interfere with the detection of E. coli by some procedures.

Although a single species cannot be expected to mimic the survival and growth patterns of all faecally-derived pathogens, E. coli is the most appropriate indicator of faecal contamination. Significant numbers of E. coli in soil are thought to be a sign of relatively recent faecal contamination. However, there have been examples where salmonellae have been detected on vegetables but not E. coli.

If the presence of E. coli is to be used as a measure of good agricultural and manufacturing practices, then it is obvious that the numerical limits suggested must be attainable by the application of good agricultural and manufacturing practices. An individual processor can establish his own baseline and use this information to detect and correct any drift or sudden deviation from control. However, interpreting such a baseline is easier when it can be compared with that obtained by other manufacturers.

International guidelines

In the absence of published microbiological data for Australian packaged ready-to-eat minimally-processed vegetables produced under good manufacturing conditions, guidelines suggested in other countries can be considered. Much of the available data refers to shredded or salad vegetables in the market-place. German criteria are <100 E. coli/g. French specifications allow 2 of 5 samples to have between 10 and 1000 E. coli/g with none of the 5 over 1000/g. An early ICMSF recommendation for ready-to-eat vegetables was similar to the French specifications. However, this recommendation was removed in later ICMSF proposals, as routine microbiological examination was not recommended. For frozen vegetables, ICMSF recommendation was for none of 5 samples to have over 1000 E. coli/g, and, "in the absence of systematic data", suggested that 2 of the samples could have between 100 and 1000 E. coli/g.

A guideline of <100 E. coli/g in prepared ready-to-eat salads produced under good manufacturing conditions has been suggested by the IFST Professional Food Microbiology Group. "Occasional samples may exceed these limits and such results should not be used as rejection criteria without consideration of all the factors important in an entire sampling plan". A maximum value of 1000 E. coli/g was suggested with higher values indicating a failure of the process or hygienic procedures. UK Public Health Laboratory Service guidelines for fresh vegetables suggest that 20 E. coli/g is satisfactory, 20 to <100 is borderline, 100 to 10,000 unsatisfactory and >10,000 E. coli/g is unacceptable or potentially hazardous.

The predominant view is that, under good manufacturing practices, processors should be able to produce packaged, fresh-cut vegetables that regularly will have fewer than 100 E. coli/g. This recommendation applies to all packaged fresh cut vegetables with the exception of cultured seeds and sprouts and microbiological guidelines will be suggested for this class of products in the next issue of this bulletin.

A number of authorities have proposed that, at the point of consumption, foods should have fewer than 100 viable cells of L. monocytogenes/g. Where fresh-cut vegetables have a shelf-life of 6-7 days at 5-7° C and the expected growth of L. monocytogenes over this time is less than 10-fold, product should have fewer than 10 cells of this microorganism/g at the time of packaging. Where shelf-life is longer and growth will be greater, numbers permitted at production need to be reduced accordingly.