Elaboration of microbiological guidelines as an element of codes of hygienic practices for small and/or less developed businesses to verify compliance with hazard analysis critical control point.

Effective assurance of microbiological food safety practices in small and/or less developed businesses is not yet resolved. Although a start has been made by drafting hygiene codes, feasible methods for verifying manufacturing processes that rely on strict and meaningful criteria to be applied to process points are still lacking. This investigation is a model study with various types of ready-to-eat foods aimed at verifying adequate processing for safety and subsequent meticulous hygienic handling and safe storage of foods in small and/or less developed businesses by the use of quantitative methods for selected indicator organisms, as standardized by the International Organization for Standardization. The manufacture of the foods that were used in this study invariably included (i) a heat processing step that ensured a level of lethality of nonsporing organisms at least equivalent to the reduction of such organisms attained in the pasteurization of milk and (ii) effective means of prevention of postprocess recontamination and recolorization. The results of this study indicate that simple microbiological criteria used for this purpose, including aerobic mesophilic colony (standard plate) counts, Enterobacteriaceae counts, and, in some instances, enumeration of yeast propagules, allow adequate verification of good practices throughout. This verification through monitoring of samples taken during processing in small and/or less developed businesses was found to be an attractive alternative to the conventional examination of end products.

Mould spoilage of cereals during transportation by sea from Latin America to Europe mechanisms, impact and management.

Safe water contents of consignments of cereals to be shipped overseas can be calculated from the relation between mould-free storage time and storage conditions (temperature of the environment, aw of the cereal), corrected for heterogeneity of water distribution, content of damaged kernels and degree of infestation by insects. The validity of this model was substantiated by the inspection of shipments and theoretical data from the literature. This predictive model can usefully be substituted for previously used, ill-defined criteria like average or any portion's water content and should prompt the trade to sell consignments of cereals on the basis of dry substance.

Adequate protection of the public against food-transmitted diseases of microbial aetiology. Achievements and challenges, half a century after the introduction of the Prescott-Meyer-Wilson strategy of active intervention.

No dramatic change in the approach to supplying microbiologically safe foods and meals to the consumer occurred immediately upon the introduction, in the 1930's, by S.C. Prescott and K.F. Meyer in the U.S. and Sir Graham Wilson in the U.K. of the principle of taking preventive remedial measures rather than merely examining final product samples. Thus morbidity and the economic impact of food-borne disease of microbial aetiology continued to rise and new food-transmitted pathogens continued to be identified. The 1930 strategy was not implemented until 1970 when food processing specialists in the U.S. launched a similar approach under the new acronym 'HACCP' (hazard analysis: critical control points). This relies on (i) identification, by meticulous ecological studies, of hazardous practices and locations, termed 'critical points' by the new generation; (ii) elimination of such hazards by developing processing techniques designed to control contamination and colonization of foods--also given an innovative term, viz. GMP; (iii) validation of the elaborated procedures by a new stem to the Science of Public Health: risk assessment, perhaps better termed Safety Science; (iv) upon introduction of the validated GMP's monitoring limited numbers of samples--which have become fully representative because of process management--for ecologically valid marker organisms, indicating incidental going-out-of-control of manufacture or distribution. This implementation of the Prescott-Meyer-Wilson maxim has been given the name longitudinally integrated safety assurance, charmingly acronymized to LISA. Introduction of LISA eliminates all previously existing or perceived difficulties about microbiological target values ('standards') for foods. These can henceforth be empirically assessed from surveys on samples obtained from factories or catering establishments previously inspected for adherence to LISA, deficiencies having, where necessary, been rectified prior to the examination of samples. Unfortunately, a more general adoption of the LISA maxim in food technology occurred at almost the same time as people became concerned about the perceived negative health impact of processing food for safety. This culminated in widespread rejection of foods decontaminated by low dose gamma irradiation. While behaviourial scientists seriously attempt to identify the mechanisms of this anxiety and devise means of reassurance, food scientists should continue to elaborate innovative LISA procedures. In this context three promising areas of research and development are briefly discussed.

Listeria monocytogenes in foods. Isolation, characterization and control.

The intent for examining foods for Listeria monocytogenes, i.e. for surveying, for epidemiological purposes, or to inspect consignments of foods for microbiological safety, determines which analytical method is to be used. For instance resuscitation of debilitated cells may be required, and the degree of accuracy and precision necessary should be considered. Moreover, in the case of acceptance-or-rejection monitoring target values for 'absence' of the pathogen have been amply but not always effectively discussed. Recommendations are given for assessing adequate repair of sublethally damaged populations of L. monocytogenes and for performance testing of selective enrichment and isolation media to be used for the isolation and enumeration of L. monocytogenes. An approach to the empirical assessment of reference values for L. monocytogenes in foods processed for safety is also presented. This relies on a data base of results obtained when examining foods manufactured and distributed according to practices previously validated by longitudinally integrated ('holistic') quantitative risk analysis.

Public health implication of refrigerated pasteurized ('sous-vide') foods.

Food that upon pasteurization is stored in hermetically sealed containers at food temperatures not exceeding 3 degrees C could be designated by the generic term Refrigerated Pasteurized Foods of Extended Durability, REPFEDs. If not properly processed or protected against recontamination, or if temperature-abused, REPFEDs may present serious health risks. However, control is readily available. Sound microbial ecology, supported by expert risk assessment, allows the design and introduction of longitudinally integrated manufacture, distribution, handling by outlets and consumers and culinary preparation, which result in the assurance of the wholesomeness of the commodity as eaten. Recent progress, including intrinsic preservation by the incorporation of starter cultures, bacteriocins or particular enzymes, opens vistas for attractive future developments. Once microbiological safety has been built into the REPFED-line, monitoring can be limited to (i) real-time tests particularly applied to the factory environment; and (ii) rapid, simple examination for marker organisms of freshly manufactured products versus those approaching expiration dates. Such audits will allow rapid retrieval of incidental process failure and its rectification. It also serves to substantiate measurements of food temperature and spot checks on intrinsic inhibitory attributes. The application of scientific knowledge and technological expertise should primarily be entrusted to the industry itself, heeding Lord Plumb's strategy of "partnership along the food production chain from farm to fork." It should be supported and validated by Public Health Authorities. At all stages safety communication with the public should be ensured.

Control of the transmission of Vibrio cholerae and other enteropathogens by foods originating from endemic areas in South America and elsewhere as a model situation.

The cholera-pandemic raging in South and Middle America and endemic cholera in other countries call for measures of health protection of the local population, but particularly with respect to the young, old, pregnant and immunocompromised citizens of countries importing food from the areas where the disease has struck. Instead of harshly barring importation, a more humanitarian policy is recommended, relying on assistance of areas presenting risks, with the introduction of and adherence to rigorous measures of longitudinally integrated microbiological safety assurance. This model is equally applicable to other enteric diseases transmitted by food. Examples are given of how canned foods of neutral pH, fishery products, vegetables and certain fruits should be processed for safety. Importation monitoring strategies, linked to this proactive approach to consumer protection, are briefly summarized.

Lactic acid decontamination of fresh pork carcasses: a pilot plant study.

Lactic acid decontamination (LAD) was carried out in an abattoir on pork carcasses, artificially contaminated with Salmonella typhimurium in faeces suspensions. The surface contamination with S. typhimurium ranged from 1-2 log10 cfu/cm2. Before cold and hot LAD was undertaken, the inoculum was allowed to adhere to the meat surface for 20 min. Cold LAD consisted of treatment for 60 s with 2% (pH 2.3) or 5% (pH 1.9) lactic acid (LA); for hot LAD the exposure times were 30, 60, 90 and 120 s. The spray nozzle temperatures were 11 degrees C and 55 degrees C, and that of the treated meat surface 16-18 degrees C and 36-38 degrees C, respectively. Treatment with cold 2% and 5% LA for 60 s eliminated S. typhimurium from pork carcasses inoculated with ca. 1 log10 cfu/cm2, but not from those inoculated at ca. 2 log10 cfu/cm2. However, this could be achieved by hot 2% and 5% LA sprayed for 60-120 s. Also exposures of at least 30 s using these hot LA solutions eliminated S. typhimurium consistently from carcasses inoculated with ca. 1 log10 cfu/cm2. Rinsing-off contributed only marginally to contamination reduction. Application of 2% or 5% LA for 120 s led to an unacceptable deterioration of the organoleptic qualities of the meat. Addition of nicotinic and ascorbic acid as colour stabilizers to the spraying solutions reduced these changes to just acceptable levels when 2% LA was used.

Bacteriological quality of broiler carcasses as affected by in-plant lactic acid decontamination.

In an attempt to improve the bacteriological quality of broiler carcasses the bactericidal effect of treatments with 1% and 2% lactic acid was investigated. Bacterial colonisation was determined immediately after treatment, after the carcasses had been chilled and during storage at 0 degrees C. Examination included numbers of mesophilic aerobic and psychrotrophic aerobic colony-forming units (CFU), CFU of Enterobacteriaceae at 37 degrees C and CFU of Staphylococcus aureus. Immediately after treatment colonisation per gram skin was generally reduced by about 1 log. Initially 2% lactic acid was not found significantly more effective in reducing colony counts than 1%. However, treatment with 2% lactic acid suppressed post-decontamination colonisation with Enterobacteriaceae more effectively than 1% lactic acid, as determined after 15-18 days storage at about 0 degrees C. Lactic acid treatment was most effective when applied shortly before chilling. Successive treatment at three different stages during slaughtering did not increase reduction of colony counts. It is concluded that decontamination with 1-2% lactic acid at pH 2, when applied shortly before chilling, will markedly improve the bacterial safety and increase the refrigerated shelf life of broiler carcasses.

A selective and diagnostic medium for use in the enumeration of Listeria spp. in foods.

A new medium, called RAPAMY agar, has been elaborated for the isolation from and the enumeration of Listeria spp. in foods. It is based on Ralovich's nalidixic acid-trypaflavin-agar with the following modifications: (i) the slight inhibitory properties of that medium were overcome by the use of Columbia Blood agar base instead of tryptose agar and the addition of 0.05% ferric ammonium citrate and 2.5% egg yolk emulsion; (ii) selectivity was improved by the addition of 0.25% 2-phenyl ethanol and incubation under microaerobic conditions; (iii) the medium was provided with two diagnostic traits by the addition of (a) aesculin + ferric ammonium citrate; and (b) D-mannitol and phenol red. The growth of Enterococcus spp., the only organisms other than Listeria spp. which grow on RAPAMY agar, was not inhibited by the addition of 20 microgram.ml-1 Cefoxitin (Moxolactam). Higher levels inhibited some Listeria spp., but not the enterococci. The medium recovered Listeria spp. quantitatively and allowed recovery from foods colonized by Enterococcus spp. at levels upto 10(2) per g.

Assessment of cleaning and disinfection in the food industry with the rapid ATP-bioluminescence technique combined with the tissue fluid contamination test and a conventional microbiological method.

A quantitative ATP bioluminescence procedure has been used to determine the cleanliness of food processing factories and the results have been compared with those from conventional microbiological culture methods. ATP measurements were combined with the tissue or tissue fluid contamination (TTFC) assessment method to obtain an impression of the amount of inanimate contamination on the sampled surfaces. It was found that, in the sampled food factories, there was poor relation between the two assessment techniques: ATP bio-luminescence combined with TTFC and contact plating. However, either method in its own right is useful to check cleanliness of food industries. ATP measurements do have in addition the great advantage that it is a fast method and is easy to perform.

Liquid and solid selective differential media for the detection and enumeration of L. monocytogenes and other Listeria spp.

A selective and differential medium (PALCAM agar) was elaborated for the isolation and enumeration of Listeria monocytogenes. PALCAM is based on Columbia agar with 0.05% glucose made selective by the addition of 0.001% polymyxin B, 0.0005% acriflavin, 1.5% lithium chloride and 0.002% ceftazidime. The diagnostic traits were attained by the incorporation of (i) 0.08% aesculin and 0.05% ferric salt; and (ii) 1% mannitol plus 0.008% phenol red. PALCAM recovered test strains of L. monocytogenes and other Listeria spp quantitatively and suppressed most other bacteria of common occurrence in fresh food. L. monocytogenes colonies were approximately 2 mm grey-green with a black sunken centre and a black halo on a cherry-red background. The occasional Enterococcus or Staphylococcus strains developing on the medium gave rise to grey colonies with a brown-green halo or yellow colonies with a yellow halo. PALCAM was the preferred medium out of 13 tested Listeria selective agars in current use. A similar differential enrichment broth, L-PALCAMY was developed based on peptone yeast extract broth with 2.5% egg yolk emulsion. The diagnostic traits and inhibitors used in this medium were the same as in PALCAM agar, through in different concentrations. Growth rate and cellcrop of L. monocytogenes in L-PALCAMY were of the same order as in Columbia broth. The growth of the majority of other bacteria of common occurrence in fresh foods was inhibited. The medium recovered L. monocytogenes more effectively from severely contaminated food than other current enrichment media.

Identification and quantification of risk factors regarding Salmonella spp. on pork carcasses.

The main elements of a descriptive epidemiological model for Salmonella spp. in Dutch pig slaughterlines, and the subsequent quantification of risk factors regarding the contamination of carcasses, are described. There is a strong correlation between the number of live animals that carry Salmonella spp. in their faeces and the number of contaminated carcasses at the end of the slaughterline. Live animals that carry Salmonella spp. are 3-4 times more likely to end up as a positive carcass than Salmonella-free animals. Currently, about 70% of all carcass contamination results from the animals themselves being carriers, and 30% because other animals were carriers (i.e. cross contamination). Furthermore, it is estimated that in general between 5-30% of the carcasses produced may contain Salmonella spp. With respect to carcass contamination with Enterobacteriaceae and Salmonella spp., inadequately cleaned polishing machines (odds ratio, OR, 6) and 'inapt procedures during evisceration' (OR 11), i.e. faulty evisceration and hygiene practices, are the most important risk factors. An estimated 5-15% of all carcass contamination with Salmonella spp. occurs during polishing after singeing. The remainder is the result of current evisceration practices (55-90%) and, to a lesser extent, further processing (5-35%), i.e dressing, splitting and meat inspection. Less likely Salmonella spp. already present on the skin of the live animals survive scalding and singeing. However, because pigs are the only important source for the Salmonella contamination of the line and the carcasses produced, it can also be concluded that if Salmonella-free pigs were produced, consumers could be provided with virtually Salmonella-free pork. As long as Salmonella-positive animals enter abattoirs, there will always be transmission of Salmonella spp. to consumers, even if the process is carried out according to stringent codes of good manufacturing practices (GMP). EU regulations should, therefore, allow for the decontamination of caracasses with a safe substance, e.g. lactic acid, on the condition that the slaughterhouse strictly adhers to GMP principles.

Modified methods for the enumeration of spores of mesophilic Clostridium species in dried foods.

Clostridia constitute markers of limited though definite importance for the microbiological integrity of particular foods processed for safety, provided their application and the results obtained are meticulously considered and guided by proper ecological awareness. Their selective diagnostic enumeration in food specimens relies on their ability to reduce sulphite in agar media, visualised by the presence of ferrous cations leading to the production of black colonies. The composition of the medium used substantially affects the productivity of the procedure. We established that (1) the sulphite activity and the ferrous ion should be rigorously standardised; (2) tryptose is one of the appropriate nitrogen sources for a limited number of clostridia; (3) the basal medium should be free of added acetate and lactate. Black colonies obtained in the newly elaborated medium, termed Differential clostridial agar (DCA) should be further examined for morphology and metronidazole sensitivity, since some bacilli might mimic clostridia under the conditions of the procedure. An elegant variant of the technique relies on using a bottom-layer of mannitol/egg yolk/polymyxin/bromocresol purple agar, inoculated with macerates of food in buffered cysteine hydrochloride peptone saline, immediately liberally overlayered with freshly prepared DCA. Plates are incubated and read in tightly closed bags of plastic with a low oxygen permeability coefficient, which eliminates the need for using anaerobic jars. Colony identification is relying on assessment of sulphite reduction, egg yolk dissimilation, the mode of attack on mannitol and when required to be supported by classical other physiological traits. The mandatory precautions to be observed in this procedure call for extreme caution when introducing reference ranges ("standards") for clostridial spores in foods, particularly in the international commerce.

Identification, assessment and management of food-related microbiological hazards: historical, fundamental and psycho-social essentials.

Microbiological risk assessment aimed at devising measures of hazard management, should take into account all perceived hazards, including those not empirically identified. It should also recognise that safety cannot be "inspected into" a food. Rather hazard management should be the product of intervention strategies in accordance with the approach made mandatory in the EU Directive 93/43 and the USDA FSIS Pathogen Reduction HACCP system; Final Rule. It is essential too that the inherent variability of the biological attributes affecting food safety is recognised in any risk assessment. The above strategic principles may be conceptualised as a four-step sequence, involving (i) identification and quantification of hazards; (ii) design and codification of longitudinally integrated ("holistic") technological processes and procedures to eliminate, or control growth and metabolism of, pathogenic and toxinogenic organisms; (iii) elaboration of microbiological analytical standard operating procedures, permitting validation of "due diligence" or responsible care, i.e. adherence to adopted intervention strategies. This should be supported by empirically assessed reference ranges, particularly for marker organisms, while the term "zero tolerance" is refined throughout to tolerable safety limit; (iv) when called for, the need to address concerns arising from lay perceptions of risk which may lack scientific foundation. In relation to infectious and toxic hazards in the practical context the following general models for quantitative holistic risk assessment are presented: (i) the first order, basic lethality model; (ii) a second approximation taking into account the amount of food ingested in a given period of time; (iii) a further adjustment accounting for changes in colonization levels during storage and distribution of food commodities and the effects of these on proliferation of pathogens and toxin production by bacteria and moulds. Guidelines are provided to address: (i) unsubstantiated consumer concern over the wholesomeness of foods processed by an innovative procedure; and (ii) reluctance of small food businesses to adopt novel strategies in food safety. Progress here calls for close cooperation with behavioural scientists to ensure that investment in developing measures to contain risk deliver real benefit.

Identification, assessment and management of food-related microbiological hazards: historical, fundamental and psycho-social essentials.

Microbiological risk assessment aimed at devising measures of hazard management, should take into account all perceived hazards, including those not empirically identified. It should also recognise that safety cannot be "inspected into" a food. Rather hazard management should be the product of intervention strategies in accordance with the approach made mandatory in the EU Directive 93/43 and the USDA FSIS Pathogen Reduction HACCP system; Final Rule. It is essential too that the inherent variability of the biological attributes affecting food safety is recognised in any risk assessment. The above strategic principles may be conceptualised as a four-step sequence, involving (i) identification and quantification of hazards; (ii) design and codification of longitudinally integrated ("holistic") technological processes and procedures to eliminate, or control growth and metabolism of, pathogenic and toxinogenic organisms; (iii) elaboration of microbiological analytical standard operating procedures, permitting validation of "due diligence" or responsible care, i.e. adherence to adopted intervention strategies. This should be supported by empirically assessed reference ranges, particularly for marker organisms, while the term "zero tolerance" is refined throughout to tolerable safety limit; (iv) when called for, the need to address concerns arising from lay perceptions of risk which may lack scientific foundation. In relation to infectious and toxic hazards in the practical context the following general models for quantitative holistic risk assessment are presented: (i) the first order, basic lethality model; (ii) a second approximation taking into account the amount of food ingested in a given period of time; (iii) a further adjustment accounting for changes in colonization levels during storage and distribution of food commodities and the effects of these on proliferation of pathogens and toxin production by bacteria and moulds. Guidelines are provided to address: (i) unsubstantiated consumer concern over the wholesomeness of foods processed by an innovative procedure; and (ii) reluctance of small food businesses to adopt novel strategies in food safety. Progress here calls for close cooperation with behavioural scientists to ensure that investment in developing measures to contain risk deliver real benefit.

Salmonella spp. on pork at cutting plants and at the retail level and the influence of particular risk factors.

This article describes the contamination of pork with Salmonella spp. in cutting plants and butchers' shops in The Netherlands and quantifies the influence of several risk factors. When contaminated carcasses are being processed, the main risk factors regarding cross contamination are inapt cleaning and disinfection (OR 12.8), manipulation of contaminated materials as such (OR 4.7) and (re)contaminated surfaces (OR 4.4). However, in the current situation, where contaminated carcasses are constantly being brought into cutting lines, interim cleaning and disinfection of surfaces and utensils during breaks and at the end of the working day will most likely prevent not more than about 10% of all cross contamination that takes place during a working day. Thus, as long as contaminated carcasses are being processed, about 90% of the cross contamination that occurs in cutting plants is practically unavoidable. It can therefore also be concluded that under these circumstances the implementation of codes of good manufacturing practices (GMP) and Hazard Analysis Critical Control Point (HACCP)-inspired production methods will only be marginally effective in the control of Salmonella spp. cross contamination in cutting lines. The same is more or less true for the processing of contaminated cuts or carcasses by butchers in shops and supermarkets. Furthermore, in contrast to the situation in cutting plants, it may be that up to 10% of butcher's shops or kitchens of restaurants become colonized for several weeks or months with their own endemic 'house flora' of Salmonella spp., which are originally introduced via the purchased contaminated products of animal origin. Though there are no hard data to substantiate this, it can be suspected that these shops and restaurants represent the more badly managed, i.e. poorly cleaned and disinfected, enterprises. However, several analytical limitations hinder an exact determination of the prevalence of Salmonella spp. contaminated pork and an exact quantification the influence of risk factors. The diagnostic value (i.e. the sensitivity, specificity, precision and predictive value) of the combination of swabbing of carcasses and cuts and the usually employed culturing methods, in particular, is largely unknown, and there are indications that it may be seriously questioned. Without a more thorough knowledge about the diagnostic value of current and future methods of sampling and identification, it is impossible to provide for more accurate estimations of the prevalence of Salmonella positive carcasses and cuts. Based on the research data, the incidence of contaminated cuts and retail-ready pork can not be estimated more precise than as somewhere between 5-40%. When compensating for the discussed methodological flaws, it must be assumed that currently the true prevalence of contaminated primal cuts and retail-ready pork in butchers' shops is about 25-30%, and that of minced pork and pork sausages about 50-55%. Lastly it is concluded that if carcasses were Salmonella-free, consumers could in principle be provided with virtually Salmonella-free pork. It is therefore recommended that the EU allows for a decontamination step in slaughterhouses with a substance that is generally recognized as safe, provided that the producers strictly adhere to GMP-principles.

Impact on human health of Salmonella spp. on pork in The Netherlands and the anticipated effects of some currently proposed control strategies.

The impact on human health of Salmonella spp. on pork in The Netherlands is described. Subsequently, the effects of some currently proposed control strategies in the Dutch pork production chain are evaluated and quantified with the aid of a simple mathematical model. The estimated average incidence of cases of salmonellosis in the Netherlands is about 450 cases per 100,000 person years at risk (pyar). Some special risk groups for which the risks could be quantified are (1) persons with underlying diseases, such as neoplasms or diabetes mellitus (1200 cases/100,000 pyar); (2) persons with achlorhydria or who excessively use antacids (1100 cases/100,000 pyar); (3) persons who have recently been treated with antibiotics that disturb the normal gut flora (1700 cases/100,000 pyar); (4) nurses (900 cases/100,000 pyar); (5) caterers (900 cases/100,000 pyar); (6) slaughterline personnel (1800 cases/100,000 pyar). Furthermore, it is estimated that 15% (5-25%) of all cases of salmonellosis in The Netherlands are associated with the consumption of pork. Currently, proposed control measures regarding Salmonella in pigs and on pork in The Netherlands are codes of good manufacturing practices (GMP) that, in fact, formalize recommendations that can be found in many handbooks about pig breeding and pig slaughtering. When evaluated by a mathematical model constructed for this purpose, the proposed GMP codes from farm to cutting/retail could, at best, reduce the current levels of Salmonella-positive pigs and pork by 50-60%. If pigs were bred according to the rather costly specific pathogen-free concept (SPF), the prevalence of contaminated carcasses and pork could in total be reduced by 95% or more. However, implementing GMP codes from the transport phase up to the cutting/retail phase coupled with a decontamination step at the end of the slaughterline would be just as effective as GMP in combination with breeding using the SPF-concept. It is therefore concluded that the most efficient and cost-effective way of reducing the 'Salmonella problem' entailed by the consumption of pork would be to decontaminate carcasses under the precondition that the entire production chain strictly adheres to GMP principles. Therefore, the EU should also allow for more possibilities regarding the decontamination of carcasses than is currently the case. It is also concluded that current EU regulations relying on hazard analysis of critical control points (HACCP)-inspired production in cutting plants will not be effective in reducing the prevalence of Salmonella spp. on pork. This is mainly because (1) there is currently an almost steady stream of Salmonella-positive carcasses that enter the cutting process; (2) when contaminated carcasses are being processed, further cross-contamination during working hours is unavoidable; (3) no steps in the cutting process are intentionally designed to effectively reduce the risks or consequences of cross contamination of cuts and retail-ready products.

Duplex (thermotroph-psychrotroph) quadrant plates: convenient, error-avoiding tools for monitoring of HACCP-contained food lines and for epidemiological investigations under conditions of military or other constraints.

A set of two "two-tier" (thermotroph-psychrotroph) single quadrant plates (QPs) was developed previously to allow convenient enumeration of numbers of colony-forming units of most pertinent pathogenic bacteria or marker bacteria in foods. These include Listeria monocytogenes, Staphylococcus aureus, Bacillus cereus, thermotrophic and psychrotrophic Enterobacteriaceae, Clostridium spp., and Enterococcus spp. As the QPs had given excellent results in monitoring samples of marketed food products potentially involved in food-transmitted illnesses, the approach was tested for practicability under military deployment and other constraints. Three approaches were envisaged: (i) validating lapse-free adherence to meticulously codified good military catering practices; (ii) acceptance/rejection testing of locally procured foods or meals; and (iii) employing rapid culture in support of evidence obtained by microscopy in attempts to identify foods involved in infectious or toxic disease outbreaks occurring in the field. The method was found to be elegant, avoiding confusion when larger number of specimens were to be screened, as well as easy to teach to staff with little or no training in microbiology, and it provided entirely reliable results. For use outside the laboratory, preparation of food macerates by use of shake flasks containing glass or plastic beads and peptone saline as a substitute for stomaching was found acceptable, though the shake flask technique led to slightly diminished colony counts. Results obtained with incubation times shortened to ca. 12 h could be relied on only when the results were alarmingly positive, but not when the colony counts at the 12-h point did not yet indicate a reason for concern.

Hygienic aspects of porcine gullets.

In an attempt to elaborate good manufacturing practices, including the collection, processing and storage of porcine gullets, their bacterial condition immediately after collection (100 samples), as well as that of deep frozen gullets just before incorporation into meat products (40 samples), was assessed. Fresh gullets were found to be contaminated to a high degree: poured plate colony count at 30°C (PPCC) approximately 10(6) to 10(7) and Enterobacteriaceae approximately 10(3) to 10(4) cfu g(-1). Deep frozen gullets showed even higher counts: PPCC approximately 10(7) to 10(8) and Enterobacteriaceae approximately 10(4) to 10(5) cfu g(-1). Hygiene during collection was visually assessed in six abattoirs and found to be satisfactory in two, moderate in three and poor in one. The effects of processing, by cleaning or removal of the mucus membrane, on bacterial condition, pH, colour and odour were assessed before and during storage at 4°C and 20°C. Both cleaning and removal of the mucus membrane resulted in up to approximately a tenfold reduction of colony counts. After 7 days' storage at 4°C these were significantly lower than those of unprocessed gullets (P < 0·01). Processed gullets stored at 4°C were no longer fit for consumption after 4 days' storage. It being impossible to achieve a marked improvement in the bacteriological condition of gullets, the incorporation of these products into sausages should be discouraged and their use in petfoods only allowed under reasonable conditions of hygiene and chilling.

Bacteriological quality assurance (BQA) of mechanically deboned meat (MDM).

Adequacy of bacteriological quality assurance during the commercial production of mechanically deboned meat (MDM) was assessed. Lax standards of hygiene during production were observed, resulting in high numbers of Staphylococcus aureus, viz. 10(4) to 10(5) cfu g(-1), and severe contamination with Enterobacteriaceae: 10(5) to 10(6) cfu g(-1). These data indicate that measures of hygiene observed during boning of carcasses and during collection, storage and transport of bones or poultry parts should be markedly tightened, while conditions of refrigerated storage of raw materials and MDM should be improved. Use of bones of poor sensory quality (discoloration, abnormal smell) generally resulted in MDM of inferior bacteriological quality. Phage typing, biotyping and assessment of enterotoxin production was carried out with 136 St. aureus cultures, isolates from mechanically deboned pork produced at one plant. Fifty-five per cent of the isolates was not typable, 28% was typable with human phages, 8% with bovine phages. The majority of the strains could not be explicitly assigned to any Meyer and/or Hájek and Marsálek types. Applying the simplified system of Devriese to eighteen strains isolated in our investigation, ten were found to belong to the poultry ecovar, one to the bovine ecovar, while seven strains were non-host specific. None of the isolates produced enterotoxins A-E. Microbiological inspection of end products is recommended as part of an integrated quality assurance system. The following reference values for the final product (maximal colony counts to be expected under GMP conditions expressed as 95th percentile) were calculated: Pig MDM: log(10) mesophilic colony count 6·8 and log(10) cfu mesophilic Enterobacteriaceae g(-1) 4·8; Poultry MDM: log(10) mesophilic colony count 6·6 and log(10) cfu mesophilic Enterobacteriaceae g(-1) 4·7.