Suppression of pathogens in properly refrigerated raw milk.

Conflicting claims exist regarding pathogen growth in raw milk. A small pilot study was designed to provide definitive data on trends for pathogen growth and decline in raw bovine milk hygienically produced for direct human consumption. An independent laboratory conducted the study, monitoring growth and decline of pathogens inoculated into raw milk. Raw milk samples were inoculated with foodborne pathogens (Campylobacter, E. coli O157:H7, Listeria monocytogenes, or Salmonella) at lower (<162 colony forming units (CFU) per mL) and higher levels (<8,300 CFU/mL). Samples were stored at 4.4°C and quantified over time after inoculation (days 0, 3, 6, 9, 12, and 14) by standard culture-based methods. Statistical analysis of trends using the Mann-Kendall Trend Test and Analysis of Variance were conducted for 48 time series observations. Evidence of pathogen growth was documented for L. monocytogenes in 8 of 12 replicates (P = 0.001 to P = 0.028). Analysis of variance confirmed significant increases for L. monocytogenes at both initial levels in week 2. No evidence of growth was documented over 14 days for the three pathogens predominantly associated with raw milk outbreaks in the US (Campylobacter, E. coli O157:H7, and Salmonella). Further research is needed to characterize parameters for pathogen growth and decline to support re-assessment of risks that were based on incorrect assumptions about interactions of pathogens with the raw milk microbiota.

Process risk model for Salmonella and ground chicken.

AIMS: To develop a process risk model (PRM) for evaluating the safety of individual lots of ground chicken (GC) contaminated with Salmonella (Salm). METHODS AND RESULTS: Data for prevalence, number and serotype of Salm were collected with 25 g samples of GC using a combination of methods (whole sample enrichment, quantitative polymerase chain reaction, cultural isolation and serotyping). These data were used to develop a predictive model for Salm contamination of GC as a function of serving size from 25 to 300 g. This model was combined with a model for thermal inactivation of Salm in GC and a dose-response model for Salm to develop a PRM in Excel that was simulated with NeuralTools and @Risk. Of 100, 25 g samples of GC examined, 19 tested positive for Salm. Three serotypes were isolated: Infantis (n = 13), Enteritidis (n = 5) and Typhimurium (n = 1). The number of Salm ranged from 0 to 2·56 log with a median of 0·93 log per 25 g of GC. The PRM predicted that Salm prevalence would increase (P < 0·05) from 19 to 57% to 82 to 93% as serving size increased from 25 to 100 g to 200 to 300 g. However, the total number of Salm in a 100-kg lot of GC and total severity of illness (TSI) were not affected (P> 0·05) by serving size. The PRM was also used to evaluate effects of serving size distribution, cooking, food consumption behaviour, consumer demographics and Salmonella virulence on TSI. CONCLUSIONS: How a lot of GC is partitioned and consumed does not affect TSI. Scenario analysis demonstrated that the PRM can integrate prevalence, number and serotype data for Salm with consumer handling, consumption and demographics data to identify safe and unsafe lots of GC for improved food safety and public health. SIGNIFICANCE AND IMPACT OF THE STUDY: Process-risk models like the one developed in this study represent a new, holistic approach to food safety that holds great promise for improving public health and reducing food recalls.

Risk of Salmonellosis from Chicken Parts Prepared from Whole Chickens Sold in Flow Pack Wrappers and Subjected to Temperature Abuse.

The flow pack wrapper is a popular packaging choice for retail sale of whole chickens. However, it may provide a favorable environment for growth and spread of Salmonella within the package, leading to an outbreak of salmonellosis. To investigate this possibility, a process risk model was developed that predicted the risk of salmonellosis from chicken parts prepared from whole chickens sold in flow pack wrappers and subjected to proper storage (6 h at 4°C) or improper storage (72 h at 15°C) before preparation. The model had four unit operations (pathogen events): (i) preparation (contamination), (ii) cooking (death), (iii) serving (cross-contamination), and (iv) consumption (dose-response). Data for prevalence, number, and serotype of Salmonella on chicken parts were obtained by whole sample enrichment, real-time PCR. Improper storage increased (P < 0.05) prevalence of Salmonella on raw chicken parts from 10.6% (17 of 160) to 41.2% (66 of 160) and incidence of cross-contamination of cooked chicken from 10% (4 of 40) to 52.2% (24 of 46). Improper storage also increased (P < 0.05) the number (mean ± standard deviation) of Salmonella from 0.017 ± 0.030 to 3.51 ± 1.34 log per raw chicken part and from 0.048 ± 0.089 to 3.08 ± 1.50 log per cooked chicken part. The predominant serotypes isolated (n = 111) were Typhimurium (34.2%), Typhimurium var 5- (20.7%), Kentucky (12.6%), Enteritidis (11.7%), and Heidelberg (8.1%). When chicken was properly stored before preparation, the model predicted that risk of salmonellosis was low and sporadic with only six cases per 100 simulations of 105 chicken parts. However, when 0.1 to 1% of chickens were improperly stored before preparation, the model predicted that salmonellosis would increase (P < 0.05) linearly from a median of 7 (range, 1 to 15) to a median of 72 (range, 52 to 93) cases per 105 chicken parts. These results indicated that the flow pack wrapper provided a favorable environment for growth and spread of Salmonella within the package and that even when only a small percentage of packages were subjected to improper storage before preparation, the risk and size of an outbreak of salmonellosis from chicken parts increased significantly.

Neural Network Model for Thermal Inactivation of Salmonella Typhimurium to Elimination in Ground Chicken: Acquisition of Data by Whole Sample Enrichment, Miniature Most-Probable-Number Method.

Predictive models are valuable tools for assessing food safety. Existing thermal inactivation models for Salmonella and ground chicken do not provide predictions above 71°C, which is below the recommended final cooked temperature of 73.9°C for chicken. They also do not predict when all Salmonella are eliminated without extrapolating beyond the data used to develop them. Thus, a study was undertaken to develop a model for thermal inactivation of Salmonella to elimination in ground chicken at temperatures above those of existing models. Ground chicken thigh portions (0.76 cm3) in microcentrifuge tubes were inoculated with 4.45 ± 0.25 log most probable number (MPN) of a single strain of Salmonella Typhimurium (chicken isolate). They were cooked at 50 to 100°C in 2 or 2.5°C increments in a heating block that simulated two-sided pan frying. A whole sample enrichment, miniature MPN (WSE-mMPN) method was used for enumeration. The lower limit of detection was one Salmonella cell per portion. MPN data were used to develop a multiple-layer feedforward neural network model. Model performance was evaluated using the acceptable prediction zone (APZ) method. The proportion of residuals in an APZ (pAPZ) from -1 log (fail-safe) to 0.5 log (fail-dangerous) was 0.911 (379 of 416) for dependent data and 0.910 (162 of 178) for independent data for interpolation. A pAPZ ≥0.7 indicated that model predictions had acceptable bias and accuracy. There were no local prediction problems because pAPZ for individual thermal inactivation curves ranged from 0.813 to 1.000. Independent data for interpolation satisfied the test data criteria of the APZ method. Thus, the model was successfully validated. Predicted times for a 1-log reduction ranged from 9.6 min at 56°C to 0.71 min at 100°C. Predicted times for elimination ranged from 8.6 min at 60°C to 1.4 min at 100°C. The model will be a valuable new tool for predicting and managing this important risk to public health.

Neural Network Model for Survival and Growth of Salmonella enterica Serotype 8,20:-:z6 in Ground Chicken Thigh Meat during Cold Storage: Extrapolation to Other Serotypes.

Mathematical models that predict the behavior of human bacterial pathogens in food are valuable tools for assessing and managing this risk to public health. A study was undertaken to develop a model for predicting the behavior of Salmonella enterica serotype 8,20:-:z6 in chicken meat during cold storage and to determine how well the model would predict the behavior of other serotypes of Salmonella stored under the same conditions. To develop the model, ground chicken thigh meat (0.75 cm(3)) was inoculated with 1.7 log Salmonella 8,20:-:z6 and then stored for 0 to 8 -8 to 16°C. An automated miniaturized most-probable-number (MPN) method was developed and used for the enumeration of Salmonella. Commercial software (Excel and the add-in program NeuralTools) was used to develop a multilayer feedforward neural network model with one hidden layer of two nodes. The performance of the model was evaluated using the acceptable prediction zone (APZ) method. The number of Salmonella in ground chicken thigh meat stayed the same (P > 0.05) during 8 days of storage at -8 to 8°C but increased (P < 0.05) during storage at 9°C (+0.6 log) to 16°C (+5.1 log). The proportion of residual values (observed minus predicted values) in an APZ (pAPZ) from -1 log (fail-safe) to 0.5 log (fail-dangerous) was 0.939 for the data (n = 426 log MPN values) used in the development of the model. The model had a pAPZ of 0.944 or 0.954 when it was extrapolated to test data (n = 108 log MPN per serotype) for other serotypes (S. enterica serotype Typhimurium var 5-, Kentucky, Typhimurium, and Thompson) of Salmonella in ground chicken thigh meat stored for 0 to 8 days at -4, 4, 12, or 16°C under the same experimental conditions. A pAPZ of ≥0.7 indicates that a model provides predictions with acceptable bias and accuracy. Thus, the results indicated that the model provided valid predictions of the survival and growth of Salmonella 8,20:-:z6 in ground chicken thigh meat stored for 0 to 8 days at -8 to 16°C and that the model was validated for extrapolation to four other serotypes of Salmonella.

Chlorine inactivation of Salmonella Kentucky isolated from chicken carcasses: evaluation of strain variation.

The current study was undertaken to evaluate chlorine resistance among strains of Salmonella Kentucky isolated from chicken carcasses. Selected strains (n = 8) were exposed to 30 ppm of chlorine in 10% buffered peptone water (pH 7.4) for 0 to 10 min at 4°C and 150 rpm. The initial level (mean ± SD) of Salmonella Kentucky was 6.18 ± 0.09 log CFU/ml and did not differ (P > 0.05) among strains. A two-way analysis of variance indicated that the level of Salmonella Kentucky in chlorinated water was affected (P < 0.05) by a time by strain interaction. Differences among strains increased as a function of chlorine exposure time. After 10 min of chlorine exposure, the most resistant strain (SK145) was 5.63 ± 0.54 log CFU/ml, whereas the least resistant strain (SK275) was 3.07 ± 0.29 log CFU/ml. Significant differences in chlorine resistance were observed for most strain comparisons. Death of Salmonella Kentucky was nonlinear over time and fitted well to a power law model with a shape parameter of 0.34 (concave upward). Time (minutes) for a 1-log reduction of Salmonella Kentucky differed (P < 0.05) among strains: >10 min for SK145, 6.0 min for SK254, 1.5 min for SK179, and 0.3 to 0.65 min for other strains. Results of this study indicate that strain is an important variable to include in models that predict changes in levels of Salmonella Kentucky in chlorinated water.

Use of enrichment real-time PCR to enumerate salmonella on chicken parts.

Salmonella bacteria that survive cooking or that cross-contaminate other food during meal preparation and serving represent primary routes of consumer exposure to this pathogen from chicken. In the present study, enrichment real-time PCR (qPCR) was used to enumerate Salmonella bacteria that contaminate raw chicken parts at retail or that cross-contaminate cooked chicken during simulated meal preparation and serving. Whole raw chickens obtained at retail were partitioned into wings, breasts, thighs, and drumsticks using a sterilized knife and cutting board, which were then used to partition a cooked chicken breast to assess cross-contamination. After enrichment in buffered peptone water (400 ml, 8 h, 40°C, 80 rpm), subsamples were used for qPCR and cultural isolation of Salmonella. In some experiments, chicken parts were spiked with 0 to 3.6 log of Salmonella Typhimurium var. 5- to generate a standard curve for enumeration by qPCR. Of 10 raw chickens examined, 7 (70%) had one or more parts contaminated with Salmonella. Of 80 raw parts examined, 15 (19%) were contaminated with Salmonella. Of 20 cooked chicken parts examined, 2 (10%) were cross-contaminated with Salmonella. Predominant serotypes identified were Typhimurium (71%) and its variants (var. 5-, monophasic, and nonmotile) and Kentucky (18%). The number of Salmonella bacteria on contaminated parts ranged from one to two per part. Results of this study indicated that retail chicken parts examined were contaminated with low levels of Salmonella, which resulted in low levels of cross-contamination during simulated meal preparation and serving. Thus, if consumers properly handle and prepare the chicken, it should pose no or very low risk of consumer exposure to Salmonella.

Chlorine inactivation of nonresistant and antibiotic-resistant strains of salmonella typhimurium isolated from chicken carcasses.

A study was conducted to test the hypothesis that strains of Salmonella Typhimurium that are resistant to antibiotics are more resistant to chlorine in chilled water than strains of Salmonella Typhimurium that are not resistant to antibiotics. To test this hypothesis, strains (n = 16) of Salmonella Typhimurium with four antibiotic resistance profiles were tested for their inactivation kinetics in chlorinated (30 ppm, pH 6) water at 4°C. The four antibiotic resistance profiles were (i) none; (ii) tetracycline-sulfisoxazole (T-Su); (iii) tetracycline-ampicillin-amoxicillin-cefoxitin-ceftiofur-sulfisoxazole (T-A-Am-C-Ce-Su); and (iv) tetracycline-ampicillin-amoxicillin-cefoxitin-ceftiofur-sulfisoxazole-kanamycin (T-A-Am-C-Ce-Su-K). Inactivation of Salmonella Typhimurium in chlorinated water displayed nonlinear kinetics with a concave downward curve that fit well (R(2) = 0.964) to the power law model, with a shape parameter of 1.37. The time for a single log reduction (D-value) of Salmonella Typhimurium from an initial concentration of 5.36 log/ml did not differ (P > 0.05) among the four antibiotic resistance groups and ranged from 3.8 to 4.3 min for n = 4 strains per group. Thus, the hypothesis was rejected, and it was concluded that expression of an antibiotic resistance phenotype does not confer cross-protection in Salmonella Typhimurium to chlorine inactivation in chilled water.

Validation of a predictive model for survival and growth of Salmonella typhimurium DT104 on chicken skin for extrapolation to a previous history of frozen storage.

The U.S. Department of Agriculture's tertiary Pathogen Modeling Program (PMP) model for survival and growth of Salmonella enterica ser. Typhimurium definitive type 104 (DT104) on chicken skin stored for 0 to 8 h at 5 to 50°C was evaluated for its ability to predict survival and growth of the same organism on chicken skin after frozen storage for 6 days at -20°C. Experimental design and methods used to collect data for model development (dependent data) were the same as those used to collect data for survival and growth after frozen storage (independent data for extrapolation). This was done to provide a valid comparison of observed and predicted values. The model was classified as providing acceptable predictions of the test data when the proportion of residuals in an acceptable prediction zone (pAPZ) from -1 log (fail-safe) to 0.5 log (fail-dangerous) was ≥0.7. The pAPZ for dependent data, independent data for interpolation, and independent data for extrapolation to a new independent variable of previous frozen storage were all acceptable (pAPZ ≥0.7), with the exception of the pAPZ for dependent data at 50°C, where an unacceptable pAPZ of 0.625 was obtained. Although a majority of observed log counts were less than predicted log counts, indicating that frozen storage of chicken skin for 6 days at -20°C had injured some Salmonella Typhimurium DT104, the injury was not large enough to cause the tertiary PMP model to provide unacceptable predictions. Thus, it was concluded that the tertiary PMP model provided valid predictions of survival and growth of Salmonella Typhimurium DT104 on chicken skin that had a previous history of frozen storage for 6 days at -20°C. Additional research is needed to determine how broadly the model can be applied to other conditions of previous frozen storage.

Initial contamination of chicken parts with Salmonella at retail and cross-contamination of cooked chicken with Salmonella from raw chicken during meal preparation.

The current study was undertaken to acquire data on contamination of chicken parts with Salmonella at retail and to acquire data on cross-contamination of cooked chicken with Salmonella from raw chicken during meal preparation. Whole raw chickens (n = 31) were obtained from local retail stores and cut into two wings, two breasts without skin or bones, two thighs, and two drumsticks. Data for cross-contamination were obtained by cutting up a sterile, cooked chicken breast with the same board and knife used to cut up the raw chicken. The board, knife, and latex gloves used by the food handler were not rinsed or washed before cutting up the sterile, cooked chicken breast, thus providing a worst-case scenario for cross-contamination. Standard curves for the concentration of Salmonella bacteria in 400 ml of buffered peptone water after 6 h of incubation of chicken parts as a function of the initial log number of Salmonella bacteria inoculated onto chicken parts were developed and used to enumerate Salmonella bacteria. Standard curves were not affected by the type of chicken part but did differ (P < 0.05) among the five isolates of Salmonella examined. Consequently, Salmonella bacteria were enumerated on naturally contaminated chicken parts using a standard curve developed with the serotype of Salmonella that was isolated from the original sample. The prevalence of contamination was 3 % (4 of 132), whereas the incidence of cross-contamination was 1.8 % (1 of 57). The positive chicken parts were a thigh from chicken 4, which contained 3 CFU of Salmonella enterica serotype Kentucky, and both wings, one thigh, and one cooked breast portion from chicken 15, which all contained 1 CFU of serotype 8,20:-:z(6). These results indicated that the poultry industry is providing consumers in the studied area with chicken that has a low prevalence and low number of Salmonella bacteria at retail and that has a low incidence and low level of cross-contamination of cooked chicken with Salmonella from raw chicken during meal preparation under a worst-case scenario.

Plenary lecture: innovative modeling approaches applicable to risk assessments.

Proper identification of safe and unsafe food at the processing plant is important for maximizing the public health benefit of food by ensuring both its consumption and safety. Risk assessment is a holistic approach to food safety that consists of four steps: 1) hazard identification; 2) exposure assessment; 3) hazard characterization; and 4) risk characterization. Risk assessments are modeled by mapping the risk pathway as a series of unit operations and associated pathogen events and then using probability distributions and a random sampling method to simulate the rare, random, variable and uncertain nature of pathogen events in the risk pathway. To model pathogen events, a rare event modeling approach is used that links a discrete distribution for incidence of the pathogen event with a continuous distribution for extent of the pathogen event. When applied to risk assessment, rare event modeling leads to the conclusion that the most highly contaminated food at the processing plant does not necessarily pose the highest risk to public health because of differences in post-processing risk factors among distribution channels and consumer populations. Predictive microbiology models for individual pathogen events can be integrated with risk assessment models using the rare event modeling method.

Qualitative map of Salmonella contamination on young chicken carcasses.

Salmonella contamination of poultry is a global public health problem. The objective of this study was to map the distribution of Salmonella on the young chicken carcass, to improve poultry inspection and food safety. Young chickens (n = 70) in the Cornish game hen class were obtained at retail over a 3-year period. Carcasses were aseptically sectioned into 12 parts, and then Salmonella was isolated from whole-part incubations by conventional culture methods. Isolates were characterized for serotype and antibiotic resistance, and by pulsed-field gel electrophoresis (PFGE). Salmonella incidence was 21.5% (181 of 840) for parts and 57.1% (40 of 70) for carcasses. The number of contaminated parts per carcass ranged from 0 to 12, with a mean of 4.5 among contaminated carcasses. Chi-square analysis indicated that Salmonella incidence differed (P < 0.05) among parts, with rib back (38.6%) and sacral back (34.3%) being the most contaminated. Among the 40 contaminated carcasses, there were 37 different patterns of contamination among parts. Of the 33 carcasses with more than one contaminated part, 12.1% contained two serotypes, 33.3% contained two or more antibiotic resistance profiles, and 100% contained two or more PFGE patterns. The most common serotype was Typhimurium (94.5%), and most (97.2%) isolates were resistant to multiple antibiotics. These results indicated a diverse pattern of Salmonella contamination among carcasses and that multiple subtypes of Salmonella were often present on contaminated carcasses. Thus, whole-carcass incubation succeeded by characterization of multiple isolates per carcass is needed to properly assess and manage this risk to public health.

Predictive model for survival and growth of Salmonella typhimurium DT104 on chicken skin during temperature abuse.

To better predict risk of Salmonella infection from chicken subjected to temperature abuse, a study was undertaken to develop a predictive model for survival and growth of Salmonella Typhimurium DT104 on chicken skin with native flora. For model development, chicken skin portions (2.14 cm2) were inoculated with 0.85 log of Salmonella Typhimurium DT104 (ATCC 700408) and then stored at 5 to 50 degrees C for 8 h. Kinetic data from the storage trials were fit to a primary model to determine lag time (lamda), specific growth rate (micrro), and the 95% prediction interval (PI). Secondary models for lamda, mu, and PI as a function of storage temperature were developed and then combined with the primary model to create a tertiary model. Performance of the tertiary model was evaluated against dependent data, independent data for interpolation, and independent data for extrapolation to kosher chicken skin by using an acceptable prediction zone from -1 (fail-safe) to 0.5 (fail-dangerous) log per skin portion. Survival of Salmonella Typhimurium DT104 on chicken skin was observed during 8 h of storage at 5 to 20 degrees C and at 50 degrees C, whereas growth was observed from 25 to 45 degrees C and was optimal at 40 degrees C with a lamda of 2.5 h and a mu of 1.1 log/h. Variation of pathogen growth, as assessed by PI, increased in a nonlinear manner as a function of temperature and was greater for growth conditions than no-growth conditions. The percentage of acceptable prediction errors was 82.6% for dependent data, 83.7% for independent data for interpolation, and 81.6% for independent data for extrapolation to kosher skin, which all exceeded the performance criterion of 70% acceptable predictions. Thus, it was concluded that the tertiary model provided valid predictions for survival and growth of Salmonella Typhimurium DT104 from a low initial dose on both nonkosher and kosher chicken skin with native flora.

An approach for mapping the number and distribution of Salmonella contamination on the poultry carcass.

Mapping the number and distribution of Salmonella on poultry carcasses will help guide better design of processing procedures to reduce or eliminate this human pathogen from poultry. A selective plating media with multiple antibiotics (xylose-lysine agar medium [XL] containing N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) and the antibiotics chloramphenicol, ampicillin, tetracycline, and streptomycin [XLH-CATS]) and a multiple-antibiotic-resistant strain (ATCC 700408) of Salmonella Typhimurium definitive phage type 104 (DT104) were used to develop an enumeration method for mapping the number and distribution of Salmonella Typhimurium DT104 on the carcasses of young chickens in the Cornish game hen class. The enumeration method was based on the concept that the time to detection by drop plating on XLH-CATS during incubation of whole chicken parts in buffered peptone water would be inversely related to the initial log number (N0) of Salmonella Typhimurium DT104 on the chicken part. The sampling plan for mapping involved dividing the chicken into 12 parts, which ranged in average size from 36 to 80 g. To develop the enumeration method, whole parts were spot inoculated with 0 to 6 log Salmonella Typhimurium DT104, incubated in 300 ml of buffered peptone water, and detected on XLH-CATS by drop plating. An inverse relationship between detection time on XLH-CATS and N0 was found (r = -0.984). The standard curve was similar for the individual chicken parts and therefore, a single standard curve for all 12 chicken parts was developed. The final standard curve, which contained a 95% prediction interval for providing stochastic results for N0, had high goodness of fit (r2 = 0.968) and was N0 (log) = 7.78 +/- 0.61 - (0.995 x detention time). Ninety-five percent of N0 were within +/- 0.61 log of the standard curve. The enumeration method and sampling plan will be used in future studies to map changes in the number and distribution of Salmonella on carcasses of young chickens fed the DT104 strain used in standard curve development and subjected to different processing procedures.

Development and validation of a stochastic model for predicting the growth of Salmonella typhimurium DT104 from a low initial density on chicken frankfurters with native microflora.

The presence of native microflora is associated with increased variation of Salmonella growth among batches and portions of chicken meat and as a function of temperature. However, variation of Salmonella growth can be modeled using a 95% prediction interval (PI). Because there are no reports of predictive models for growth of Salmonella on ready-to-eat poultry meat products with native microflora and because Salmonella is usually present at low levels on poultry meat, the current study was conducted to develop and validate a stochastic model for predicting the growth of Salmonella from a low initial density on chicken frankfurters with native microflora. One-gram portions of chicken frankfurters were inoculated with 0.5 log CFU of a single strain (ATCC 700408) of Salmonella Typhimurium DT104. Changes in pathogen numbers over time, N(t), were fit to a two-phase linear primary model to determine lag time (lambda), growth rate (mu), and the 95% PI, which characterized the variation of pathogen growth. Secondary quadratic polynomial models for natural log transformations of lambda, mu, and PI as a function of temperature (10 to 40 degrees C) were obtained by nonlinear regression. The primary and secondary models were combined in a computer spreadsheet to create a tertiary model that predicted the growth curve and PI. The pathogen did not grow on chicken frankfurters incubated at 10 to 12 degrees C, but mu ranged from 0.003 log CFU/g/h at 14 degrees C to 0.176 log CFU/ g/h at 30 degrees C to 0.1 log CFU/g/h at 40 degrees C. Variation of N(t) increased as a function of time (i.e., PI was lower during lag phase than during growth phase) and temperature (i.e., PI was higher at 18 to 40 degrees C than at 10 to 14 degrees C). For dependent data (n = 338), 90.5% of observed N(t) values were in the PI predicted by the tertiary model, whereas for independent data (n = 86), 89.5% of observed N(t) values were in the PI predicted by the tertiary model. Based on this performance evaluation, the tertiary model was considered acceptable and valid for stochastic predictions of Salmonella Typhimurium DT104 growth from a low initial density on chicken frankfurters with native microflora.

Survival and growth of Listeria monocytogenes in broth as a function of temperature, pH, and potassium lactate and sodium diacetate concentrations.

The objective of this study was to determine the antimicrobial effect of a combination of potassium lactate and sodium diacetate (0, 1.8, 3, and 4.5%; PURASAL P Opti. Form 4, 60% solution) on the survival and growth of Listeria monocytogenes Scott A in pH-adjusted broth (5.5, 6.0, 6.5, and 7.0) stored at 4, 10, 17, 24, 30, and 37 degrees C. Appropriate dilutions of broth were enumerated by spiral plating on tryptose agar and counted with an automated colony counter. Growth data were iteratively fit, using nonlinear regression analysis to a three-phase linear model, using GraphPad PRISM. At pH 5.5, the combination of lactate-diacetate fully inhibited (P < 0.001) the growth of L. monocytogenes at all four levels and six temperatures. At pH 6.0, addition of 1.8% lactate-diacetate reduced (P < 0.001) the specific growth rate of L. monocytogenes and increased lag time; however, 3 and 4.5% completely inhibited the growth at the six temperatures studied. Efficacy of the lactate-diacetate mixture was decreased as pH increased and incubation temperature increased. Thus, at pH 6.5, at least 3% was required to retard (P < 0.001) the growth of L. monocytogenes in broth. There was a limited effect of the lactate-diacetate level on the specific growth rate of the pathogen at pH 7.0. However, 1.8 and 3% significantly lengthened the lag time at 4 and 10 degrees C. These results suggest that 1.8% of lactate-diacetate mixture can be used as a substantial hurdle to the growth of L. monocytogenes when refrigerated temperatures are maintained for products with pH less than 6.5.

Predictive models for growth of Salmonella typhimurium DT104 from low and high initial density on ground chicken with a natural microflora.

A single strain (ATCC 700408) of Salmonella typhimurium DT104 was used to investigate and model growth from a low (1.12 log10 mpn g(-1)) and high (3.7 log10 cfu g(-1)) initial density on ground chicken with a natural microflora. Kinetic data for growth of the pathogen on ground chicken were fit to a primary model to determine lag time (lambda), maximum specific growth rate (mu) and maximum population density (Nmax). Secondary models for lambda, mu and Nmax, as a function of temperature (10-40 degrees C), were developed and compared among initial densities. Variation of pathogen growth among replicates (n=4 or 5) was higher at 10-18 degrees C than at 22-40 degrees C and was higher for Nmax than lambda and mu. Prediction problems were observed when secondary models developed with one initial density were used to predict lambda, mu and Nmax from the other initial density, especially at 10-18 degrees C and for Nmax. These results indicated that variation of growth among replicate challenge studies and initial density are important factors to consider when developing predictive models for growth of S. typhimurium DT104 on ground chicken with a natural microflora.

Validation of a tertiary model for predicting variation of Salmonella typhimurium DT104 (ATCC 700408) growth from a low initial density on ground chicken breast meat with a competitive microflora.

Growth of a multiple antibiotic-resistant strain (ATCC 700408) of Salmonella Typhimurium definitive phage type 104 (DT104) from a low initial density (10(0.6) most probable number [MPN] or CFU/g) on ground chicken breast meat with a competitive microflora was investigated and modeled as a function of time and temperature (10 to 40 degrees C). MPN and viable counts (CFU) on a selective medium with four antibiotics enumerated the pathogen. Data from five replicate challenge studies per temperature were combined and fit to a primary model to determine maximum specific growth rate (micro), maximum population density (Nmax), and the 95% prediction interval (PI). Nonlinear regression was used to obtain secondary models as a function of temperature for micro, Nmax, and PI, which ranged from 0.04 to 0.4 h(-1), 1.6 to 9.4 log MPN or CFU/g, and 1.4 to 2.4 log MPN or CFU/g, respectively. Secondary models were combined with the primary model to create a tertiary model for predicting variation (95% PI) of pathogen growth among batches of ground chicken breast meat with a competitive microflora. The criterion for acceptable model performance was that 90% of observed MPN or CFU data had to be in the 95% PI predicted by the tertiary model. For data (n=344) used in model development, 93% of observed MPN and CFU data were in the 95% PI predicted by the tertiary model, whereas for data (n=236) not used in model development but collected using the same methods, 94% of observed MPN and CFU data were in the 95% PI predicted by the tertiary model. Thus, the tertiary model was successfully verified against dependent data and validated against independent data for predicting variation of Salmonella Typhimurium DT104 growth among batches of ground chicken breast meat with a competitive microflora and from a low initial density.

Transformation of Escherichia coli K-12 with a high-copy plasmid encoding the green fluorescent protein reduces growth: implications for predictive microbiology.

The green fluorescent protein (GFP) of the jellyfish Aequorea victoria has been widely used as a biomarker and has potential for use in developing predictive models for growth of pathogens on naturally contaminated food. However, constitutive production of GFP can reduce growth of transformed strains. Consequently, a high-copy plasmid with gfp under the control of a tetracycline-inducible promoter (pTGP) was constructed. The plasmid was first introduced into a tetracycline-resistant strain of Escherichia coli K-12 to propagate it for subsequent transformation of tetracycline-resistant strains of Salmonella. In contrast to transformed E. coli K-12, which only fluoresced in response to tetracycline, transformed Salmonella fluoresced maximally without tetracycline induction of gfp. Although pTGP did not function as intended in Salmonella, growth of parent and GFP E. coli K-12 was compared to test the hypothesis that induction of GFP production reduced growth. Although GFP production was not induced during growth on sterile chicken in the absence of tetracycline, maximum specific growth rate (mumax) of GFP E. coli K-12 was reduced 40 to 50% (P < 0.05) at 10, 25, and 40 degrees C compared with the parent strain. When growth of parent and GFP strains of E. coli K-12 was compared in sterile broth at 40 degrees C, mumax and maximum population density of the GFP strain were reduced (P < 0.05) to the same extent (50 to 60%) in the absence and presence of tetracycline. These results indicated that transformation reduced growth of E. coli K-12 independent of gfp induction. Thus, use of a low-copy plasmid or insertion of gfp into the chromosome may be required to construct valid strains for development of predictive models for growth of pathogens on naturally contaminated food.

Development and validation of primary, secondary, and tertiary models for growth of Salmonella Typhimurium on sterile chicken.

Models are used in the food industry to predict pathogen growth and to help assess food safety. However, criteria are needed to determine whether models provide acceptable predictions. In the current study, primary, secondary, and tertiary models for growth of Salmonella Typhimurium (10(4.8) CFU/g) on sterile chicken were developed and validated. Kinetic data obtained at 10 to 40 degrees C were fit to a primary model to determine initial density (N0), lag time (lambda), maximum specific growth rate (micromax), and maximum population density (Nmax). Secondary models for N0, lambda, micromax, and Nmax as a function of temperature were developed and combined with the primary model to create a tertiary model that predicted pathogen density (N) at times and temperatures used and not used in model development. Performance of models was evaluated using the acceptable prediction zone method in which experimental error associated with growth parameter determinations was used to set criteria for acceptable model performance. Models were evaluated against dependent and independent (validation) data. Models with 70% prediction or relative errors (RE) in an acceptable prediction zone from -0.3 to 0.15 for micromax, -0.6 to 0.3 for lambda, and -0.8 to 0.4 for N, N0, and Nmax were classified as acceptable. All secondary models had acceptable goodness of fit and were validated against independent (interpolation) data. Percent RE in the acceptable prediction zone for the tertiary model was 90.7 for dependent data and 97.5 for independent (interpolation) data. Although the tertiary model was validated for interpolation, an unacceptable %RE of 2.5 was obtained for independent (extrapolation) data obtained with a lower N0 (10(0.8) CFU/g). The tertiary model provided overly fail-dangerous predictions of N from a lower N0. Because Salmonella concentrations on chicken are closer to 10(0.8) than 10(4.8) CFU/g, the tertiary model should not be used to help assess chicken safety.