Raw, frozen chicken nuggets and strips have been identified as a significant risk factor in contracting foodborne salmonellosis. Cases of salmonellosis as a result of consuming partly cooked chicken nuggets may be due in part to Salmonella strains originating in broiler feed. This study was undertaken to determine the occurrence and characterize the strains of Salmonella contaminating chicken nuggets, strips, and pelleted feeds, in an attempt to demonstrate whether the same Salmonella strains present in broiler feed could be isolated from raw, frozen chicken nuggets and strips available for human consumption. Salmonellae were recovered using the Health Canada MFHPB-20 method for the isolation and identification of Salmonella from foods. Strains were characterized by serotyping, phage typing, antimicrobial resistance typing (R-typing), and by pulsed-field gel electrophoresis (PFGE). Salmonellae were isolated from 25-g samples in 27% ( n = 92) of nugget and strip samples, 95% ( n = 20) of chicken nugget meat samples, and from 9% ( n = 111) of pelleted feed samples. Salmonella Heidelberg, Salmonella Enteritidis, and Salmonella Orion were the most commonly isolated serovars from chicken nuggets and strips, nugget and strip meat, and pelleted broiler feeds, respectively. Salmonella Enteritidis phage type (PT) 13a with PFGE pattern SENXAI.0006 and R-type sensitive as well as Salmonella Enteritidis PT13a with PFGE pattern SENXAI.0068 and R-type sensitive were isolated from pelleted feed, and chicken nugget and strip meat in two separate instances. Data showed that Salmonella strains isolated from broiler feed were indistinguishable from strains isolated from packaged raw, frozen chicken nuggets and strips. However, results did not rule out the possibility that breeding stock or contamination during processing may have contributed to chicken meat contamination by Salmonella .

The present study compared the sensitivity of the BAX automated fluorometric and the recently discontinued BAX gel electrophoresis systems with a standard culture method to detect Salmonella in 333 high-moisture and 171 low-moisture foods. A total of 95 naturally contaminated foods, including 63 high-moisture and 32 low-moisture foods, were detected by the standard culture method. No contaminated samples were identified exclusively by the BAX systems. By means of the analytical protocol stipulated by the manufacturer, the BAX fluorometric system detected 36 (57.1%) and 29 (90.6%) of the contaminated high- and low-moisture foods, respectively. Similar results were obtained with the BAX gel electrophoresis system, which identified 40 (63.5%) and 26 (81.3%) of the contaminated high- and low-moisture foods. The rate of false-positive reactions with the BAX systems was low. Our results indicate that the low sensitivity of the BAX systems with high-moisture foods, notably raw meats and poultry products, was serovar-independent. The high levels of background microflora that commonly occur in raw meat and on fresh fruit and vegetable products, and the high successive dilutions of test materials for PCR analysis, suggestively undermined the sensitivity of the gel and the fluorometric BAX assays. The potential benefits of immunomagnetic separation of Salmonella in preenrichment cultures, of selective broth enrichment following preenrichment to markedly reduce levels of background microflora in PCR test materials, and the use of larger portions of test materials in PCR analyses should be investigated.

The prevalence of microwave ovens in North American homes has increased dramatically within the last decade. Although microwave ovens are primarily used for reheating of foods, they are now more commonly being applied to the cooking of raw foods. Although cooking of raw foods, according to manufacturers' instructions targets an organoleptically acceptable end product, the process does not address the microbiological safety of the cooked food. Seventeen microwave ovens from various commercial suppliers were used to cook naturally contaminated whole raw broilers (≤1.8 kg) and roasters (>1.8 kg) according to manufacturers' instructions. Temperature probes (six per chicken) were used to measure the temperature of chickens immediately after cooking and during the holding period. Of 81 Listeria -positive raw broilers and 93 raw roasters, 1 (1.2%) and 9 (9.7%), respectively, yielded viable Listeria spp. after microwave cooking. Of these, two were undercooked (visual inspection), one was over the maximum weight stipulated by the oven manufacturer and another one was over the maximum weight and undercooked. A significantly greater proportion of contaminated cooked birds was observed with roasters than with broilers, where for one of these contaminated roasters, the temperature at all six measured sites was ≥87°C. Most of the postcook Listeria -positive birds were associated with 2 of the 17 microwave ovens. Factors such as wattage, cavity size, and the presence or absence of a turntable seemingly did not play a significant role in the survival of Listeria spp. in microwave-cooked chicken. However, the general inability of microwave ovens to uniformly heat chicken carcasses was noted. In order to promote greater safety of microwave-cooked foods, general recommendations for consumers are provided.

A performance assessment of the GENE-TRAK ® colorimetric probe assay using pure cultures and naturally contaminated foods and animal feeds underlined the high sensitivity and specificity of this DNA-rRNA diagnostic system. The probe effectively identified 110 (100%) strains of Salmonella spp. and yielded no false-positive reactions in the examination of 61 pure cultures of nonsalmonellae. Of the 53 contaminated raw meats and other high-moisture samples examined in this study, 47 (88.7%) were detected using the GENE-TRAK ® analytical protocol. The system also identified 21 (95.5%) of 22 contaminated low-moisture foods and animal feeds. The overall sensitivity and specificity of the GENE-TRAK assay procedure with the naturally contaminated foods examined in this study was 90.7 and 100%, respectively. Ancillary work showed that the choice of selective enrichment conditions played a determinant role in the performance of the probe system. Although the GENE-TRAK protocol relies on varying temporal regimens of selective enrichment in tetrathionate brillant green (TBG 35 ) and selenite cystine (SC 35 ), and postenrichment in gram-negative (GN) broth for high and low-moisture food products, our results suggest that, irrespective of product type, probing of GN (6-h) postenrichment cultures inoculated from homologous TBG 43 and SC 35 (24-h) cultures and subsequent combination of these postenrichment cultures into a single probe assay would enhance the performance of the GENE-TRAK system to a level of unfailing sensitivity and specificity. Attempts at method brevity through direct probing of TBG 43 , TBG 35 , SC 35 , and Rappaport-Vassiliadis (RV 43 ) enrichment cultures proved unsuccessful where probing of short (6-h) enrichment cultures identified ≤15.1 % and ≤72.7% of contaminated high- and low-moisture foods, respectively. Direct probing of 24 h enrichment cultures yielded homologous values of ≤17.0% and ≤77.3%. These findings suggest that components in the selective enrichment broths inhibit the probe assay. This negative effect was most prominent with RV 43 enrichment cultures where most samples known to contain Salmonella spp. yielded false-negative reactions. Treatment of RV 43 cultures, with 0.1 M sodium citrate prior to probe assay was partially effective in neutralizing the inhibitory agent(s).

During the past two decades, there have been many studies on the efficacy of competitive exclusion for the control of Salmonella in poultry. Undefined preparations of cultured fecal or cecal microflora generally reduce the prevalence of infected chicks upon challenge with a standard dose of Salmonella under laboratory conditions; in contrast, results under field conditions are more variable. The protective capacity of undefined cultures can be affected by several factors including the source of microflora, method for protective culture administration, presence of poultry feed additives, in-laboratory or natural environmental challenge, and hygienic practices on the farm. The formulation of effective defined cultures is most difficult because of insufficient knowledge on the underlying protective mechanism(s) and interactions between gut microflora. Defined cultures are less effective than undefined cultures under laboratory conditions and afford little protection against natural Salmonella challenge; their potency decreases upon storage and manipulation of single or mixtures of defined culture isolates.

The propensity of short (6 h) selective enrichment combined with a preenrichment to enrichment transfer volume ratio of 1:100 to provide greater method brevity could not be demonstrated. Inoculation of tetrathionate brilliant green (35 and 43°C), Rappaport-Vassiliadis (43°C), and selenite cystine (35°C) enrichment broths (9.0 ml) with 0.1 ml of preenrichment culture and incubation for 6 h identified, respectively, 107 (84.9%), 104 (82.5%), 112 (88.9%), and 113 (89.7%) of the 126 contaminated samples detected in the present study; homologous results with the 1.0-ml transfer volume showed a marginal increase in sensitivity. Recoveries of foodborne salmonellae with the standard (24-h) period of selective enrichment were generally transfer volume-independent and consistently exceeded that obtained with 6-h enrichment cultures. Results further underlined the importance of enrichment at an elevated (43°C) temperature for the effective repression of competitive microflora, and the facilitated isolation of Salmonella on plating media.

The ability of the Bactigen® Salmonella Shigella (BSST), the Microscreen® (MS), and the Spectate® (SPECT) latex agglutination kits to detect Salmonella in pure cultures and in naturally contaminated foods was examined. Of 190 Salmonella strains tested, the MS, BSST, and SPECT systems correctly identified 89.5, 81.6, and 66.3% of the test cultures, respectively. The sensitivity of SPECT increased to 92.7% when only strains belonging to the targeted serogroups (somatic A to E plus G) and strains harboring the Vi antigen were considered. The lack of specificity of the MS (3.4%), SPECT (17.0%), and BSST (33.9%) systems with 59 cultures of nonsalmonellae varied widely, with Citrobacter freundii and Escherichia coli accounting for many of the false-positive reactions. Examination of foods according to the prescribed MS and SPECT analytical test protocols identified respectively, 18 (75%) and 19 (79.2%) of the 24 food samples found to contain Salmonella spp. by a standard cultural method. Although instructions with the BSST kit indicate that the product is intended for the analysis of clinical samples, the system nevertheless identified 21 (87.5%) contaminated food samples under homologous MS and SPECT test conditions. The concurrent use of TBG 43 with enrichment media recommended by kit manufactures enhanced the sensitivities of MS (83.3%), SPECT (91.7%), and BSST (91.7%). Attempts to effect greater method brevity through the application of latex kits at various stages of the standard cultural procedure were counterproductive.

The developmental BacTrace ™ ELISA system which recognizes a common structural antigen (CSA-1) in the cell wall of target microorganisms was tested with pure cultures and naturally contaminated foods. The system readily detected all of the 104 Salmonella test strains but produced 38 (52.1%) false-positive reactions upon examination of 73 nonsalmonellae cultures. Citrobacter freundii , Escherichia coli , and Proteus mirabilis were primarily responsible for erroneous results. Of 119 foods tested, 37 (31.1%) were found to contain Salmonella by a standard cultural procedure. Parallel BacTrace™ testing of the nutrient broth (NB), tetrathionate brilliant green (TBG 43 ), and selenite cystine (SC 35 ) broth cultures arising from standard cultural analyses identified 24 (64.9%), 25 (67.6%), and 31 (83.8%) Salmonella contaminated foods, respectively. Maximum sensitivity of the test system (89.2%) could be attained through combination of ELISA results from both TBG 43 and SC 35 . False-positive reactions were particularly prominent with high moisture foods.

Short (6 h) enrichment under five different selective conditions adversely affected the isolation of Salmonella from pre-enriched samples of naturally contaminated foods. Of the 109 high moisture and 18 low moisture foods found to contain salmonellae following conventional (24 h) enrichment, combined results of the abbreviated enrichment procedures identified only 99 (90.8%) and 13 (72.2%) contaminated samples, respectively. The productivities of tetrathionate brilliant green (TBG 43 ) and Muller-Kauffman tetrathionate (MKTBG 43 ) broths consistently exceeded that obtained with the modified Rappaport (RV 43 ), tetrathionate brilliant green (TBG 35 ), and selenite cystine (SC 35 ) media after 6 and 24 h of incubation. Semi-quantitative analyses of growth under all enrichment conditions indicated that short (6 h) enrichment negatively affected method sensitivity through reduced numbers of Salmonella colonies and heavy growth of nonsalmonellae on bismuth sulfite (BSA) and brilliant green sulfa (BGS) plating media. These findings raise concerns on the dependability of commercial diagnostic schemes that incorporate abbreviated (6 h) enrichment in TBG 35 and/or SC 35 in their analytical protocol.

The prevalence of Salmonella , Campylobacter , and Yersinia spp. in the food chain, and the more recent emergence of Listeria monocytogenes and hemorrhagic Escherichia coli as foodborne pathogens, are of public health concern. The ability of some of these bacterial agents to grow in milk and dairy products, to survive prolonged periods of refrigerated storage, and to withstand thermal treatments of raw milk at subpasteurizing temperatures, place new emphasis on the need for stringent control of milk processing operations and plant environment. Mandatory use of pasteurized milk may provide the only viable option for production of pathogen-free milk products.

Twenty-nine confirmed cases of Salmonella nima enterocolitis in Canada and four cases in the United States were traced to gold-foil wrapped chocolate coins from Belgium. Illness in Canadian cases occurred between September 1985 and October 1986 where the primary case in each of 24 affected families was a child ≤ 14 years of age. A product recall was issued on October 3, 1986 in Canada. Quantitative analysis of four composite samples of suspect chocolate by the most probable number (MPN) technique showed levels of 4.3 to 24.0 S. nima per 100 g product. These levels of contamination and consumption of approximately 25 g of chocolate by primary cases suggest that small numbers of S. nima precipitated clinical symptoms.

The novel 1–2 Test™ is a self-contained tetrathionate enrichment and immunodiffusion-agglutination reaction vial for the detection of Salmonella in preenrichment and direct enrichment cultures of processed and non-processed foods, respectively. Of 186 foods tested, 46 (24.7%) were found to contain salmonellae by all analytical conditions combined. The standard cultural procedure identified 43 (93.5%) contaminated samples whereas the 1–2 Test™ system detected 20 (43.5%) and 25 (54.3%) positive samples after 8 h and 24 h of incubation, respectively. Although six false-positive reactions were obtained after short (8 h) incubation of test vials, homologous reactions were not detected after 24 h incubation. System deficiency likely stems from the low selectivity of applied cultural conditions and inability of the 1–2 Test™ to detect salmonellae in the presence of large numbers of competing microflora.

The performance of the double-antibody (MOPC 467 and 6H4) BioEnzabead ™ enzyme linked immunosorbent assay was evaluated with pure cultures and naturally contaminated foods. The immunoassay detected all but four (1.6%) of the 250 Salmonella test strains and showed high levels of cross-reactivity (30.7%) among the 75 strains of non salmonellae examined. Citrobacter freundii figured prominently as a source of erroneous results. Although the BioEnzabead ™ system identified the 41 foods found to be contaminated by a standard cultural procedure, a high rate of false positive reactions (46%) was nevertheless encountered with high moisture foods. Attempt at method brevity through application of the immunoassay at the preenrichment level was unsuccessful resulting in the detection of only 24 (58.5%) of the 41 contaminated foods. Greater recoveries were obtained with tetrathionate brilliant green enrichment broth incubated at 43°C than at 35°C.

The thermal resistance of Salmonella senftenberg 775W, Salmonella muenster previously isolated from raw fluid milk, and two mixtures each consisting of ten Salmonella strains commonly isolated from human or non-human sources was tested. Cells were suspended in whole milk at a final concentration of 10 5 cells/ml. The inoculated milk was thermally processed at temperatures ranging from 60°C to 74°C using a pilot-scale plate pasteurizer unit. The mean and minimum residence time of milk in the holding tube of the pasteurizer was 17.6 and 16.2 s, respectively. The maximum temperature at which viable salmonellae were detected in the human (61.5°C) and non-human (64.5°C) mixtures was considerably lower than that obtained with S. senftenberg 775 W (67.5°C). S. muenster failed to show any milk-adapted response and could not be recovered at temperatures greater than 63.0°C. Treatment at 63°C produced a 4 log 10 or greater reduction in the number of viable Salmonella including the heat resistant S. senftenberg 775 W, and a minimum 2 log 10 decrease at 60°C. These findings warrant caution in the use of subpasteurizing temperatures for thermal processing of fluid milk.

Levels of Salmonella typhimurium phage-type 10 in Cheddar cheese implicated in a major Canadian foodborne outbreak ranged from 0.36 to 9.3 salmonellae/100 g. Such a low level contamination likely accounted for the uneven distribution of the organism among subsamples of individual lots. Coliform and Escherichia coli counts were within acceptable limits, whereas three of the 11 lots tested contained ≥10 5 Staphylococcus aureus per gram but no staphylococcal enterotoxins. Campylobacter and Yersinia spp . were not detected in any of the 12 lots examined. Ability of S. typhimurium to survive up to 8 months in Cheddar cheese stored at refrigerator temperature (5°C) underlines the inadequacy of current regulations requiring a 60-d storage of cheese manufactured from heat-treated (unpasteurized) milk before sale. Results underlined the greater sensitivity of selective enrichment in tetrathionate brilliant green (43°C) than in selenite cystine (35°C) for detection of Salmonella in cheese.

The methods of the International Office of Cocoa and Chocolate and International Sugar Confectionery Manufacturers' Association (IOCC/ISCMA), and of the Health Protection Branch (HPB) were compared for their ability to detect Salmonella in chocolate and cocoa products. Of 152 samples tested, 13 contaminated samples were identified, 10 by the HPB and 8 by the IOCC/ISCMA method. Prolonged (48 h) incubation of enrichment media produced two false-negative results each with the Muller-Kauffman tetrathionate and the selenite cystine broths, exerted no effect on tetrathionate brilliant green, and identified one additional positive sample with the selenite brilliant green broth. More samples were found positive on bismuth sulfite than on brilliant green and brilliant green sulfa agar media. The present study underlines the limited sensitivity of both standard methods and questions the determinant role of casein in the neutralization of toxic agents in cocoa products.

Efficacy of standard, 6-h standard and direct enrichment methods for detection of Salmonella in naturally contaminated feeds and feed ingredients was compared. Analysis by the standard method involved preenrichment of feed slurries in nutrient broth, selective enrichment in tetrathionate brilliant green (43°C) and selenite cystine (35°C), and isolation of presumptive isolates on bismuth sulfite and brilliant green sulfa agar media. Sample analysis by the 6-h standard method was identical to the above except that incubation of enrichment broths was reduced to 6 h; for direct enrichment. preenrichment in nutrient broth was omitted. Of 287 samples tested, 75 were found to contain salmonellae by the three methods combined. Ability of the standard and 6-h standard methods to identify the same 58 contaminated samples underlines the reliability of the 6-h standard method for the more rapid detection of Salmonella in animal feeds. Identification of 68 positive samples by direct enrichment presumably resulted from equilibration (3 to 4 h) of feed slurries at reduced water activity before analysis. Addition of novobiocin (40 μg/ml) to selective enrichment broths did not facilitate isolation of Salmonella through repression of competitive flora. Productivity of the six enrichment-plating combinations used in this study was comparable, and no single medium played a determinant role in recovery.

Twelve laboratories from 7 countries compared the productivity of refrigerated (72 h at 5 to 10°C) preenrichment and enrichment broth cultures with a standard cultural procedure for detection of Salmonella in 466 naturally contaminated low and high moisture foods. Refrigerated preenrichment and enrichment cultures identified 92.5 and 94.2% of contaminated samples, respectively. Variations in the ability of laboratories to successfully recover salmonellae under refrigeration test conditions were notable. Three laboratories found complete agreement between results by the standard and refrigeration test procedures and 5 additional laboratories reported >90% accuracy; lowest recovery rate for combined refrigeration results was 77%. Sensitivity of the refrigeration techniques was markedly greater with low than high moisture foods where the latter contributed all but two of the 62 false-negative results encountered in this study. Ability of individual laboratories to recover Salmonella from refrigerated preenrichment and enrichment broth cultures was not significantly different for given food categories. Productivity of paired enrichment-plating media differed widely with food type. Selective enrichment in tetrathionate brilliant green and plating on bismuth sulfite agar showed greatest sensitivity for isolation of Salmonella in high but not in low moisture foods where productivity of the 4 enrichment-plating conditions used in this study was comparable. Results on recoverability of Salmonella from refrigerated broth cultures concur with findings of an earlier comparative study and strongly support incorporation of this novel approach in standard cultural methods for detection of Salmonella in foods.

Inhibitory concentrations of 8 surfactants were determined for Salmonella typhimurium and Salmonella enteritidis . Pure culture work resulted in the exclusion of Tween 20, Teepol 610 and Brij 35 and retention of Tergitol-7 (T-7), Tween 80 (TW 80), Triton X-100 (TX), Myrj 52S (M), and Arlacel 80 + Tween 60 (AT) for a study on the quantitative recovery of Salmonella in 45 naturally contaminated fatty foods. Replicate food samples (100 g) were preenriched overnight at 35 C in nutrient broth supplemented with 3%(w/v) surfactant except AT (10%). Serial dilutions of preenrichment cultures were selectively enriched overnight in tetrathionate brilliant green (43 C) and selenite cystine (35 C) broths and streaked on bismuth sulfite and brilliant green sulfa agar media. Recovery with all test surfactants was comparable to that obtained with nutrient broth controls; of 270 preenrichment cultures tested, only 7 false-negative results attributable to TX (3), AT (2), M (1), and nutrient broth control (1) were obtained. None of the surfactants consistently yielded greater populations of Salmonella for given foods or food categories; median counts for preenrichment cultures were 10 4 –10 5 salmonellae/ml for low and high moisture foods and 10 6 –10 7 salmonellae/ml for animal feeds. These results suggest that use of surfactants to facilitate detection of Salmonella in fatty foods is not warranted.

Chicks ( Gallus domesticus ) were treated per os with 24-h-old anaerobic cultures of feces from mature chickens 1 day after hatching, challenged with Salmonella typhimurium in the drinking water 2 days later, and sacrificed on day 11 or 12; then the lower third of the intestinal tract was examined for salmonellae. Cultures of feces inoculated directly into the crop or added to the drinking water, even after holding at −70 C for 21 days, protected chicks against infection by S. typhimurium . Cultures serially subcultured daily up to four times were protective, and dilution to 1:80 in drinking water containing 4 % skim milk powder did not decrease their protective effect. Treated chicks were about 1000-fold more resistant to infection by Salmonella than untreated chicks.