Bayesian techniques for comparison of the test performance of PCR and culture for the identification of Campylobacter in enriched comminuted chicken samples.

AIMS: Using Bayesian methods that do not require the definition of a gold standard, the diagnostic sensitivity and specificity of a real-time polymerase chain reaction (PCR) assay are compared to those of an enriched culture assay for detection of Campylobacter in enriched comminuted chicken samples. METHODS AND RESULTS: Food Safety and Inspection Service comminuted chicken samples were collected from production facilities across the United States. Enriched samples were examined using a commercial real-time PCR kit and plated for culture. Allowing for conditional dependence between these approaches and defining relatively uninformed prior distributions, the 'no gold standard' Bayesian methods generated estimates of the means (95% credible interval) of the posterior distributions for sensitivity and specificity of the PCR as 93% (79, 100%) and 95% (87, 100%) respectively. The estimated sensitivity implies a mean false negative frequency of 7%. The estimated means of the posterior distributions for sensitivity and specificity of the culture assay were 91% (76, 100%) and 96% (88, 100%) respectively. In this case, the mean false negative frequency is 9%. Graphical comparisons of the posterior distributions with their corresponding prior distributions suggested only subtle differences in the sensitivities of both tests, but the posterior distributions for specificities are substantially more certain than the prior distributions. CONCLUSIONS: The study suggests that the commercial real-time PCR assay is a more sensitive screening test that would provide timelier negative test results. The modest 1% reduction in specificity of this PCR assay, as compared to an enriched culture assay, is less of a concern for regulatory testing programs if a culture-based confirmatory assay is applied to all presumptive positive samples. SIGNIFICANCE AND IMPACT OF THE STUDY: The sensitivity and specificity of a PCR assay and a culture assay for Campylobacter in comminuted poultry produced in the United States were estimated. The PCR assay was shown to be an appropriate alternative screening test.

Time valuation of historical outbreak attribution data.

Human illness attribution is recognized as an important metric for prioritizing and informing food-safety decisions and for monitoring progress towards long-term food-safety goals. Inferences regarding the proportion of illnesses attributed to a specific commodity class are often based on analyses of datasets describing the number of outbreaks in a given year or combination of years. In many countries, the total number of pathogen-related outbreaks reported nationwide for an implicated food source is often fewer than 50 instances in a given year and the number of years for which data are available can be fewer than 10. Therefore, a high degree of uncertainty is associated with the estimated fraction of pathogen-related outbreaks attributed to a general food commodity. Although it is possible to make inferences using only data from the most recent year, this type of estimation strategy ignores the data collected in previous years. Thus, a strong argument exists for an estimator that could 'borrow strength' from data collected in the previous years by combining the current data with the data from previous years. While many estimators exist for combining multiple years of data, most either require more data than is currently available or lack an objective and biologically plausible theoretical basis. This study introduces an estimation strategy that progressively reduces the influence of data collected in past years in accordance with the degree of departure from a Poisson process. The methodology is applied to the estimation of the attribution fraction for Salmonella and Escherichia coli O157:H7 for common food commodities and the estimates are compared against two alternative estimators.

Fitting distributions to microbial contamination data collected with an unequal probability sampling design.

AIMS: The fitting of statistical distributions to microbial sampling data is a common application in quantitative microbiology and risk assessment applications. An underlying assumption of most fitting techniques is that data are collected with simple random sampling, which is often times not the case. This study develops a weighted maximum likelihood estimation framework that is appropriate for microbiological samples that are collected with unequal probabilities of selection. METHODS AND RESULTS: A weighted maximum likelihood estimation framework is proposed for microbiological samples that are collected with unequal probabilities of selection. Two examples, based on the collection of food samples during processing, are provided to demonstrate the method and highlight the magnitude of biases in the maximum likelihood estimator when data are inappropriately treated as a simple random sample. CONCLUSIONS: Failure to properly weight samples to account for how data are collected can introduce substantial biases into inferences drawn from the data. SIGNIFICANCE AND IMPACT OF THE STUDY: The proposed methodology will reduce or eliminate an important source of bias in inferences drawn from the analysis of microbial data. This will also make comparisons between studies and the combination of results from different studies more reliable, which is important for risk assessment applications.

Association of herd management factors with colonization of dairy cattle by Shiga toxin-positive Escherichia coli O157.

Management factors in 36 Pacific Northwest dairy herds were evaluated for their association with the prevalence of Shiga toxin-positive Escherichia coli O157 (E. coli O157) in dairy cattle. The within-herd prevalence of E. coli O157 was estimated by bacteriological culture of fecal pat samples, collected monthly for 6 months (approximately 60 per visit), from heifer cattle. During the first visit to each farm, a management questionnaire was administered that covered a broad range of animal husbandry practices. On each subsequent visit, a brief questionnaire was administered to detect changes in management practices. A significantly higher prevalence of E. coli O157 was noted in herds that fed corn silage to heifers compared to herds that did not feed corn silage. More tentative associations of E. coli O157 prevalence were observed for weaning method, protein level of calf starter, feeding of ionophores in heifer rations, feeding of grain screens to heifers, and feeding of animal by-products to cows.

Multiple sources of Escherichia coli O157 in feedlots and dairy farms in the northwestern USA.

Samples from cattle, other domestic and wild animals, flies, feeds, and water-troughs were collected from 12 cattle farms and tested for Escherichia coli O157. E. coli O157 was isolated from bovine fecal samples on all 12 farms with a within herd prevalence ranging from 1.1% to 6.1%. E. coli O157 was also found in 1 of 90 (1.1%) equine fecal samples, 2 of 65 (3.1%) canine fecal samples, 1 of 200 pooled bird samples (0.5%), 2 of 60 pooled fly samples (3.3%), and 10 of 320 (3.1%) water-trough sample sets (biofilm and water). No E. coli O157 were isolated from 300 rodents, 33 cats, 34 assorted wildlife, or 335 cattle feed samples. Indistinguishable pulsed-field gel electrophoresis patterns of XbaI digested chromosomal DNA and Shiga toxin types were observed for bovine and water-trough isolates from two farms and for one equine and two bovine isolates from one farm.

Commingled grazing as a risk factor for trichomonosis in beef herds.

OBJECTIVE: To evaluate commingled grazing on public lands as a risk factor for Tritrichomonas foetus infection in beef herd bulls. DESIGN: Case-control study. SAMPLE POPULATION: Based on 1994 data from the mandatory Idaho bull testing program, all 65 infected herds (case herds), 78 randomly selected test-negative herds that tested < 10 nonvirgin bulls, and 81 randomly selected test-negative herds that tested > or = 10 nonvirgin bulls (control herds). PROCEDURE: Managers of government-owned grazing lands in Idaho identified producers who had grazing permits for an allotment under their jurisdiction in 1993 and for that allotment recorded the number of animals the producer was permitted to graze on the allotment, the number of herds with grazing permits, the total number of animals permitted, and the dates on which grazing began and ended. The number of bulls tested, number of times tested, and test results were collated from the testing database. RESULTS: The relative sensitivity of bacterial culture of preputial smegma was 81%. The attributable fraction of T foetus infection associated with commingled grazing was 33%, and the odds ratio of infection was 9.0 for herds commingled with > or = 14 other herds. The total number of animals permitted on an allotment, the use of public lands, and type of public lands used were not significantly associated with infection status. CLINICAL IMPLICATIONS: These results suggest that to control the spread of trichomonosis, the number of herds commingling on a grazing allotment should be minimized and commingled herds should be managed collectively.

Epidemiologic findings from an outbreak of cysticercosis in feedlot cattle.

An outbreak of cysticercosis in a south-central Idaho custom feedlot reached a peak prevalence of 11% in January 1993 and extended from October 1992 through March 1993. Of 5,164 cattle slaughtered from this feedlot during the outbreak, 457 (9%) were cysticercosis infected. Total discounts on the infected cattle at slaughter cost the feedlot $154,400. Most evidence was suggestive of feed-borne transmission of Taenia saginata eggs to the cattle in the feedlot. By use of logistic regression analysis of feedlot records, significant (P = 0.004) association of cysticercosis prevalence at slaughter with days on feed was revealed. Similarly, a decline in cysticercosis prevalence was significantly (P < 0.001) related to the number of days cattle were fed a ration not containing potato byproduct. Although sources other than potato byproduct were systematically evaluated during the investigation, findings suggested that potato byproduct fed in this feedlot was contaminated with T saginata eggs.

Occurrence of Salmonella enteritidis in unpasteurized liquid egg in the United States.

In order to gain a greater understanding of the occurrence and distribution of Salmonella enteritidis in the United States, a survey of unpasteurized liquid egg (collected at 20 egg-breaking plants across the United States) was conducted over a 52-week period. Weekly liquid egg samples were submitted for Salmonella culturing at the National Veterinary Services Laboratories. Group D positive salmonellae were serotyped and phage-typed. On a regional basis, the Northern Region of the United States had the highest S. enteritidis recovery, with 20% of the samples submitted from plants in that region culture-positive for this serotype. Salmonella enteritidis positives from the Southeast, Central, and Western regions were 10%, 15%, and 6% of the samples submitted, respectively.

Occurrence of Salmonella enteritidis in the U.S. commercial egg industry: report on a national spent hen survey.

In order to estimate the prevalence and distribution of Salmonella enteritidis in U.S. commercial egg-production flocks, a survey of spent laying hens was conducted over a 3-month period. Seven of the 10 largest spent-hen processing plants in the United States participated. Ceca were sampled twice weekly from birds presented for slaughter at these plants. Samples were cultured for Salmonella and S. enteritidis, and S. enteritidis isolates were phage-typed. Overall, 23,431 pooled cecal samples were collected from a total of 406 layer houses. Salmonella (any serotype) and S. enteritidis were recovered from 24% and 3% of the pooled samples, respectively. The distribution of S. enteritidis phage types was consistent with data reported by others. Regionally, the estimated prevalence of S. enteritidis-positive houses (i.e., at least one positive sample found in a house) for the Northern, Southeastern, and Central/Western regions was 45%, 3%, and 17%, respectively. Overall, the prevalence of Salmonella-positive houses was 86%.