Application of Bayesian techniques to model the burden of human salmonellosis attributable to U.S. food commodities at the point of processing: adaptation of a Danish model.

Mathematical models that estimate the proportion of foodborne illnesses attributable to food commodities at specific points in the food chain may be useful to risk managers and policy makers to formulate public health goals, prioritize interventions, and document the effectiveness of mitigations aimed at reducing illness. Using human surveillance data on laboratory-confirmed Salmonella infections from the Centers for Disease Control and Prevention and Salmonella testing data from U.S. Department of Agriculture Food Safety and Inspection Service's regulatory programs, we developed a point-of-processing foodborne illness attribution model by adapting the Hald Salmonella Bayesian source attribution model. Key model outputs include estimates of the relative proportions of domestically acquired sporadic human Salmonella infections resulting from contamination of raw meat, poultry, and egg products processed in the United States from 1998 through 2003. The current model estimates the relative contribution of chicken (48%), ground beef (28%), turkey (17%), egg products (6%), intact beef (1%), and pork (<1%) across 109 Salmonella serotypes found in food commodities at point of processing. While interpretation of the attribution estimates is constrained by data inputs, the adapted model shows promise and may serve as a basis for a common approach to attribution of human salmonellosis and food safety decision-making in more than one country.

Salmonella Enteritidis in meat, poultry, and pasteurized egg products regulated by the U.S. Food Safety and Inspection Service, 1998 through 2003.

The U.S. Food Safety and Inspection Service (FSIS) tests for Salmonella in meat, poultry, and egg products through three regulatory testing programs: the Pathogen Reduction-Hazard Analysis and Critical Control Point (PR-HACCP) program, the ready-to-eat program for meat and poultry products, and the pasteurized egg products program. From 1998 through 2003, 293,938 samples collected for these testing programs were analyzed for the presence of Salmonella enterica serotypes. Of these samples, 12,699 (4.3%) were positive for Salmonella, and 167 (1.3%) of the positive samples (0.06% of all samples) contained Salmonella Enteritidis. The highest incidence of Salmonella Enteritidis was observed in ground chicken PR-HACCP samples (8 of 1,722 samples, 0.46%), and the lowest was found in steer-heifer PR-HACCP samples (0 of 12,835 samples). Salmonella Enteritidis isolates were characterized by phage type, pulsed-field gel electrophoretic pattern, and antimicrobial susceptibility. Phage typing of 94 Salmonella Enteritidis isolates identified PT13 (39 isolates) and PT8 (36 isolates) as the most common types. One isolate from a ready-to-eat ham product was characterized as PT4. Electrophoretic analysis of 148 Salmonella Enteritidis isolates indicated genetic diversity among the isolates, with 28 unique XbaI electrophoretic patterns identified. Of these 148 isolates, 136 (92%) were susceptible to each of 16 antimicrobials tested. Two isolates were resistant to ampicillin alone, and 10 isolates were resistant to two or more antimicrobials. Isolation of Salmonella Enteritidis from FSIS-regulated products emphasizes the need for continued consumer education on proper food handling and cooking practices and continued work to decrease the prevalence of Salmonella in meat, poultry, and pasteurized egg products.

U.S. Food safety and Inspection Service testing for Salmonella in selected raw meat and poultry products in the United States, 1998 through 2003: an establishment-level analysis.

The U.S. Food Safety and Inspection Service (FSIS) pathogen reduction-hazard analysis critical control point systems final rule, published in 1996, established Salmonella performance standards for broiler chicken, cow and bull, market hog, and steer and heifer carcasses and for ground beef, chicken, and turkey meat. In 1998, the FSIS began testing to verify that establishments are meeting performance standards. Samples are collected in sets in which the number of samples is defined but varies according to product class. A sample set fails when the number of positive Salmonella samples exceeds the maximum number of positive samples allowed under the performance standard. Salmonella sample sets collected at 1,584 establishments from 1998 through 2003 were examined to identify factors associated with failure of one or more sets. Overall, 1,282 (80.9%) of establishments never had failed sets. In establishments that did experience set failure(s), generally the failed sets were collected early in the establishment testing history, with the exception of broiler establishments where failure(s) occurred both early and late in the course of testing. Small establishments were more likely to have experienced a set failure than were large or very small establishments, and broiler establishments were more likely to have failed than were ground beef, market hog, or steer-heifer establishments. Agency response to failed Salmonella sample sets in the form of in-depth verification reviews and related establishment-initiated corrective actions have likely contributed to declines in the number of establishments that failed sets. A focus on food safety measures in small establishments and broiler processing establishments should further reduce the number of sample sets that fail to meet the Salmonella performance standard.

U.S. Food Safety and Inspection Service testing for Salmonella in selected raw meat and poultry products in the United States, 1998 through 2003: analysis of set results.

The U.S. Food Safety and Inspection Service (FSIS) tests sets of samples of selected raw meat and poultry products for Salmonella to ensure that federally inspected establishments meet performance standards defined in the pathogen reduction-hazard analysis and critical control point system (PR-HACCP) final rule. In the present report, sample set results are described and associations between set failure and set and establishment characteristics are identified for 4,607 sample sets collected from 1998 through 2003. Sample sets were obtained from seven product classes: broiler chicken carcasses (n = 1,010), cow and bull carcasses (n = 240), market hog carcasses (n = 560), steer and heifer carcasses (n = 123), ground beef (n = 2,527), ground chicken (n = 31), and ground turkey (n = 116). Of these 4,607 sample sets, 92% (4,255) were collected as part of random testing efforts (A sets), and 93% (4,166) passed. However, the percentage of positive samples relative to the maximum number of positive results allowable in a set increased over time for broilers but decreased or stayed the same for the other product classes. Three factors associated with set failure were identified: establishment size, product class, and year. Set failures were more likely early in the testing program (relative to 2003). Small and very small establishments were more likely to fail than large ones. Set failure was less likely in ground beef than in other product classes. Despite an overall decline in set failures through 2003, these results highlight the need for continued vigilance to reduce Salmonella contamination in broiler chicken and continued implementation of programs designed to assist small and very small establishments with PR-HACCP compliance issues.

Salmonella enteritidis in broiler chickens, United States, 2000-2005.

US Department of Agriculture Food Safety and Inspection Service (FSIS) data on Salmonella enterica serotype Enteritidis in broiler chicken carcass rinses collected from 2000 through 2005 showed the annual number of isolates increased >4-fold and the proportion of establishments with Salmonella Enteritidis-positive rinses increased nearly 3-fold (test for trend, p<0.0001). The number of states with Salmonella Enteritidis in broiler rinses increased from 14 to 24. The predominant phage types (PT) were PT 13 and PT 8, 2 strains that a recent Foodborne Diseases Active Surveillance Network (FoodNet) case-control study associated with eating chicken. FSIS is directing more sampling resources toward plants with marginal Salmonella control to reduce prevalence in products including broilers. The policy targets establishments with common Salmonella serotypes of human illness, including Salmonella Enteritidis. Voluntary interventions should be implemented by industry.

Food safety and inspection service regulatory testing program for Escherichia coli O157:H7 in raw ground beef.

We analyzed raw ground beef testing data to determine whether a decrease in the rate of Escherichia coli O157:H7-positive raw ground beef samples has occurred since the inception of Food Safety and Inspection Service (U.S. Department of Agriculture) regulatory actions and microbiological testing concerning this commodity and pathogen. A main effects log-linear Poisson regression model was constructed to evaluate the association between fiscal year and the rate of E. coli O157:H7-positive raw ground beef samples while controlling for the effect of season for the subset of test results obtained from fiscal year (FY)2000 through FY2003. Rate ratios were used to compare the rate of E. coli O157:H7-positive raw ground beef samples between sequential years to identify year-to-year differences. Of the 26,521 raw ground beef samples tested from FY2000 through FY2003, 189 (0.71%) tested positive for E. coli O157:H7. Year-to-year comparisons identified a 50% reduction in the rate of positive ground beef samples from FY2002 to FY2003 when controlling for season (95% CI, 10 to 72% decrease; P = 0.02). This decrease was the only significant year-to-year change in the rate of E. coli O157:H7-positive raw ground beef samples but was consistent in samples obtained from both federally inspected establishments and retail outlets. We believe this decrease is attributed to specific regulatory actions by Food Safety and Inspection Service and subsequent actions implemented by the industry, with the goal of reducing E. coli O157:H7 adulteration of raw ground beef. Continued monitoring is necessary to confirm that the decrease in the rate of E. coli O157:H7 in raw ground beef samples we observed here represents the beginning of a sustained trend.

Estimate of illnesses from Salmonella enteritidis in eggs, United States, 2000.

Results from our model suggest that eating Salmonella enterica serovar Enteritidis-contaminated shell eggs caused 182,060 illnesses in the United States during 2000. Uncertainty about the estimate ranged from 81,535 (5th percentile) to 276,500 illnesses (95th percentile). Our model provides but 1 approach for estimating foodborne illness and quantifying estimate uncertainty.

Comparison of management practices between Ohio, USA dairy farms participating in whole-herd Johne's disease testing programs and those not participating.

The purpose of our mail survey was to compare the adoption of management practices recommended for Johne's disease (JD) control between herds involved in whole-herd testing programs versus those that do not routinely test the entire herd for JD. A questionnaire consisted of 38 closed-ended questions that inquired about: general herd characteristics; management practices related to JD control; changes that occurred within the last 5 years regarding management practices recommended for the control of JD; producer knowledge of JD; the perceived infection status of the herd by the producer; and herd JD-testing history. The questionnaire was mailed to 810 Ohio dairy producers in September 2002; 266 questionnaires were returned (32.8% response). We used univariable logistic-regression models to assess the relationship between whole-herd testing status (TESTING versus NON-TESTING) and each management practice, each change in management practice and producer knowledge about JD. Because it is conceivable that only producers who believe their herds to be infected would be motivated to adopt the management practices recommended for control of JD, the comparisons were repeated with models that controlled for producer-perceived infection status. Of the 20 management practices recommended for JD control that we evaluated, 7 differed between TESTING and NON-TESTING herds. Additionally, TESTING herds more-frequently reported adopting changes within the past 5 years relative to NON-TESTING herds with respect to 7 of 9 management practices evaluated. Producers with TESTING herds also reported greater familiarity with JD than those with NON-TESTING herds.

Sensitivity of test strategies used in the Voluntary Johne's Disease Herd Status Program for detection of Mycobacterium paratuberculosis infection in dairy cattle herds.

OBJECTIVE: To evaluate sensitivities at the herd level of test strategies used in the Voluntary Johne's Disease Herd Status Program (VJDHSP) and alternative test strategies for detecting dairy cattle herds infected with Mycobacterium paratuberculosis. DESIGN: Nonrandom cross-sectional study. SAMPLE POPULATION: 64 dairy herds from Pennsylvania, Minnesota, Colorado, Ohio, and Wisconsin. Fifty-six herds had at least 1 cow shedding M. paratuberculosis in feces; the other 8 herds were free from paratuberculosis. PROCEDURE: For all adult cows in each herd, serum samples were tested for antibodies to M. paratuberculosis with an ELISA, and fecal samples were submitted for bacterial culture for M. paratuberculosis. Sensitivities at the herd level (probability of detecting infected herd) of various testing strategies were then evaluated. RESULTS: Sensitivity at the herd level of the testing strategy used in level 1 of the VJDHSP (use of the ELISA to test samples from 30 cows followed by confirmatory bacterial culture of feces from cows with positive ELISA result) ranged from 33 to 84% for infected herds, depending on percentage of cows in the herd with positive bacterial culture results. If follow-up bacterial culture was not used to confirm positive ELISA results, sensitivity ranged from 70 to 93%, but probability of identifying uninfected herds as infected was 89%. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that the testing strategy used in the VJDHSP will fail to identify as infected most dairy herds with a low prevalence of paratuberculosis. A higher percentage of infected herds was detected if follow-up bacterial culture was not used, but this test strategy was associated with a high probability of misclassifying uninfected herds.