Bayesian Source Attribution of Salmonella Typhimurium Isolates From Human Patients and Farm Animals in England and Wales

Bayesian evaluation of meat juice ELISA for detecting Salmonella in slaughtered pigs without specifying a cut-off

BACKGROUND

Consumption of pork and pork products is a major source of human infection with Salmonella. Salmonella is typically subclinical in pigs, making it difficult to identify infected pigs. Therefore, effective surveillance of Salmonella in pigs critically relies on good knowledge on how well the diagnostic tests used perform. A test that has been used in several countries for Salmonella monitoring is serological testing of meat juice using an ELISA (MJ ELISA) to detect antibodies against Salmonella. This MJ ELISA data could be used to estimate infection prevalence and trends. However, as the MJ ELISA output is a sample-to-positive (S/P) ratio, which is a continuous outcome rather than a binary (positive/negative) result, the interpretation of this data depends upon a chosen cut-off.

AIM

To apply Bayesian latent class models (BLCMs) to estimate diagnostic accuracy of the MJ ELISA test values in the absence of a gold standard without needing to apply a cut-off.

METHODS AND RESULTS

BLCMs were fitted to data from a UK abattoir survey carried out in 2006 in order to estimate the diagnostic accuracy of MJ ELISA with respect to the prevalence of active Salmonella infection. This survey consisted of a MJ ELISA applied in parallel with the bacteriological testing of caecal contents, carcass swabs and lymph nodes (n = 625). A BLCM was also fitted to the same data but with dichotomisation of the MJ ELISA results, in order to compare with the model using continuous outcomes. Estimates were obtained for sensitivity and specificity of the ELISA over a range of S/P values and for the bacteriological tests and were found to be similar between the models using continuous and dichotomous ELISA outcomes.

CONCLUSION

The Bayesian method without specifying a cut-off does allow prevalence to be inferred without specifying a cut-off for the ELISA. The study results will be useful for estimating infection prevalence from serological surveillance data.

Occurrence of Salmonella spp. in animal patients and the hospital environment at a veterinary academic hospital in South Africa

Background and Aims

Nosocomial infections caused by Salmonella spp. are common in veterinary facilities. The early identification of high-risk patients and sources of infection is important for mitigating the spread of infections to animal patients and humans. This study investigated the occurrence of Salmonella spp. among patients at a veterinary academic hospital in South Africa. In addition, this study describes the environmental factors that contribute to the spread of Salmonella spp. in the veterinary facility.

Materials and Methods

This study used a dataset of Salmonella-positive animals and environmental samples submitted to the bacteriology laboratory between 2012 and 2019. The occurrence of Salmonella isolates at the veterinary hospital was described based on source, month, season, year, and location. Proportions and 95% confidence intervals were calculated for each variable.

Results

A total of 715 Salmonella isolates were recorded, of which 67.6% (483/715) came from animals and the remainder (32.4%, 232/715) came from environmental samples. The highest proportion (29.2%) of Salmonella isolates was recorded in 2016 and most isolates were reported in November (17.4%). The winter season had the lowest (14.6%) proportion of isolates reported compared to spring (31.3%), summer (27.8%), and autumn (26.4%). Salmonella Typhimurium (20.0%) was the most frequently reported serotype among the samples tested, followed by Salmonella Anatum (11.2%). Among the positive animal cases, most (86.3%) came from equine clinics. Most reported isolates differed based on animal species with S. Typhimurium being common in equines and S. Anatum in bovines.

Conclusion

In this study, S. Typhimurium emerged as the predominant strain in animal and environmental samples. Equines were the most affected animals; however, Salmonella serotypes were also detected in the production animals. Environmental contamination was also a major source of Salmonella species in this study. To reduce the risk of transmission, strict infection prevention and control measures (biosecurity) must be implemented.

Development and validation of a random forest algorithm for source attribution of animal and human Salmonella Typhimurium and monophasic variants of S. Typhimurium isolates in England and Wales utilising whole genome sequencing data

Source attribution has traditionally involved combining epidemiological data with different pathogen characterisation methods, including 7-gene multi locus sequence typing (MLST) or serotyping, however, these approaches have limited resolution. In contrast, whole genome sequencing data provide an overview of the whole genome that can be used by attribution algorithms. Here, we applied a random forest (RF) algorithm to predict the primary sources of human clinical Salmonella Typhimurium (S. Typhimurium) and monophasic variants (monophasic S. Typhimurium) isolates. To this end, we utilised single nucleotide polymorphism diversity in the core genome MLST alleles obtained from 1,061 laboratory-confirmed human and animal S. Typhimurium and monophasic S. Typhimurium isolates as inputs into a RF model. The algorithm was used for supervised learning to classify 399 animal S. Typhimurium and monophasic S. Typhimurium isolates into one of eight distinct primary source classes comprising common livestock and pet animal species: cattle, pigs, sheep, other mammals (pets: mostly dogs and horses), broilers, layers, turkeys, and game birds (pheasants, quail, and pigeons). When applied to the training set animal isolates, model accuracy was 0.929 and kappa 0.905, whereas for the test set animal isolates, for which the primary source class information was withheld from the model, the accuracy was 0.779 and kappa 0.700. Subsequently, the model was applied to assign 662 human clinical cases to the eight primary source classes. In the dataset, 60/399 (15.0%) of the animal and 141/662 (21.3%) of the human isolates were associated with a known outbreak of S. Typhimurium definitive type (DT) 104. All but two of the 141 DT104 outbreak linked human isolates were correctly attributed by the model to the primary source classes identified as the origin of the DT104 outbreak. A model that was run without the clonal DT104 animal isolates produced largely congruent outputs (training set accuracy 0.989 and kappa 0.985; test set accuracy 0.781 and kappa 0.663). Overall, our results show that RF offers considerable promise as a suitable methodology for epidemiological tracking and source attribution for foodborne pathogens.

Salmonella in horses at slaughter and public health effects in Italy

The study assessed the role of equids at slaughter as faecal carriers of Salmonella enterica and the occurrence of contaminated equid carcasses during the slaughter process in Northern Italy (Emilia-Romagna Region). From June to November 2021, 152 equids (146 horses, 5 donkeys and 1 mule) were tested for Salmonella both in caecal contents and through carcass swabs. Antimicrobial resistance (AMR) of recovered strains was tested against 15 antimicrobials. Salmonella was detected in 3/152 of the caecal contents (2.0 %), while all carcass samples were negative. S. enterica serovars Enteriditis, Typhimurium and Stanleyville were identified. The only AMR isolate was S. Typhimurium with AMR profile AmCStxT. Considering the consumption of raw horse meat (i.e., minced raw meat named "pesto di cavallo" and dried and smoked strips named "sfilacci di cavallo") in different areas of Northern Italy, we also investigated the possible link between horse meat eating and salmonellosis cases in the human population in the same area. Specifically, we compared the Salmonella strains collected during the study with those routinely processed in the laboratory surveillance system for human salmonellosis in Emilia-Romagna (a region with about 4.5 million inhabitants). The comparison was based on whole genome sequencing data through core genome multi-locus sequence typing (cgMLST) used in routine surveillance. A genomic match in cgMLST was found between the strain of S. enterica serovar Enteritidis isolated from a horse caecal content and an enduring outbreak of 17 human cases in Emilia-Romagna during the study period. The consequent epidemiological investigation highlighted that a number of cases with known food history reported the consumption of horse meat and traced different batches of the consumed meat, released weeks apart from each other, to the slaughter investigated in the study. The results of the epidemiological investigation suggested the role of horses in the S. enterica serovar Enteritidis outbreak affecting raw horse meat consumers. This study shows that, despite the low prevalence on equid carcasses, S. enterica in horse meat can represent a risk to consumers. From the perspective of the slaughter activities, this highlights the need to maintain a high level of hygiene during the entire process, starting from the hygiene at lairage up to the slaughtering phase and dressing of carcasses.

Using SNP addresses for Salmonella Typhimurium DT104 in routine veterinary outbreak detection

SNP addresses are a pathogen typing method based on whole-genome sequences (WGSs), assigning groups at seven different levels of genetic similarity. Public health surveillance uses it for several gastro-intestinal infections; this work trialled its use in veterinary surveillance for salmonella outbreak detection. Comparisons were made between temporal and spatio-temporal cluster detection models that either defined cases by their SNP address or by phage type, using historical data sets. Clusters of SNP incidents were effectively detected by both methods, but spatio-temporal models consistently detected these clusters earlier than the corresponding temporal models. Unlike phage type, SNP addresses appeared spatially and temporally limited, which facilitated the differentiation of novel, stable, or expanding clusters in spatio-temporal models. Furthermore, these models flagged spatio-temporal clusters containing only two to three cases at first detection, compared with a median of seven cases in phage-type models. The large number of SNP addresses will require automated methods to implement these detection models routinely. Further work is required to explore how temporal changes and different host species may impact the sensitivity and specificity of cluster detection. In conclusion, given validation with more sequencing data, SNP addresses are likely to be a valuable addition to early warning systems in veterinary surveillance.

Relationship between the Cycle Threshold Value (Ct) of a Salmonella spp. qPCR Performed on Feces and Clinical Signs and Outcome in Horses

The objective of this retrospective study was to evaluate the clinical significance of fecal quantitative real-time polymerase chain reaction (qPCR) Salmonella results when taking the cycle threshold values (Ct) into account. The study included 120 Salmonella qPCR-positive fecal samples obtained from 88 hospitalized horses over a 2-year period. The mean Ct of the qPCR test was evaluated in regard to (1) clinical outcome and (2) systemic inflammatory response syndrome (SIRS) status (no SIRS, moderate SIRS, or severe SIRS) of the sampled horses. An ROC analysis was performed to establish the optimal cut-off Ct values associated with severe SIRS. The mean ± SD Ct value was significantly lower in samples (1) from horses with a fatal issue (27.87 ± 5.15 cycles) than in surviving horses (31.75 ± 3.60 cycles), and (2) from horses with severe SIRS (27.87 ± 2.78 cycles) than from horses with no (32.51 ± 3.59 cycles) or moderate (31.54 ± 3.02 cycles) SIRS. In the ROC analysis, the optimal cut-off value of Ct associated with a severe SIRS was 30.40 cycles, with an AUC value of 0.84 [95% confidence interval 0.76-0.91] and an OR of 0.64 [0.51-0.79]. Results suggest that including the Ct value in the interpretation of fecal qPCR results could improve the diagnostic value of this test for clinical salmonellosis in horses.

Salmonella enterica in soybean production chain: Occurrence, characterization, and survival during soybean storage

This study aimed to determine Salmonella enterica occurrence along the soybean meal production chain (raw material, in-processing samples, final products, and in the environment of five processing plants), characterize the isolates, and assess the survival of Salmonella Senftenberg 775W in soybeans stored under different temperature conditions. Among 713 samples analyzed, 12.9% (n = 92) were positive for Salmonella enterica. Dust collected inside and outside processing plants (n = 148) comprised the samples with the highest positivity for Salmonella enterica, 47.3%. The occurrence of Salmonella enterica varied among the different processing plants. Twenty-nine (n = 29) Salmonella serotypes were isolated, with S. Mbandaka as the most frequent serotype, whereas S. Typhimurium was mainly linked to final product samples (soybean meal). S. Senftenberg 775W did not survive for a long time in soybean stored at 20-37 °C, but at 20 °C, cells were viable for more than 60 days. This study suggests that soybean meal may harbor Salmonella serotypes related to foodborne disease outbreaks in humans and can be responsible for Salmonella introduction into livestock and, consequently, in foods of animal origin. This study provides crucial data on contamination pathways of Salmonella in the soybean production chain, contributing to the understanding of Salmonella epidemiology which is strategic for the development of preventive and control measures to reduce the burden of salmonellosis linked to products of animal origin.

Risk factor analysis for Salmonella contamination of broiler chicken (Gallus gallus) hatcheries in Great Britain

Salmonellosis is the second most commonly reported zoonosis in the European Union and contaminated meat from broiler chickens (Gallus gallus) is an important source of human infection. In Great Britain (GB), prevalence of Salmonella enterica in broiler flocks is low, having declined considerably since the introduction of the Salmonella National Control Programme in 2010. However, this decreasing trend has stabilised in recent years and serovars with known ability to persistently colonise hatcheries have been isolated from broiler flocks with increasing frequency, indicating that further controls on hatchery contamination are required. The broiler industry in GB has changed dramatically over the last 15 years, with greater intensification and dominance by a small number of very large companies which rely on relatively few hatcheries. An investigation of risk factors for Salmonella contamination in GB broiler hatcheries was therefore carried out so that relevant up-to-date advice on Salmonella control can be provided. Twenty-two hatcheries, representing most commercial scale GB broiler hatcheries, were visited between 2015 and 2018. Salmonella contamination was comprehensively investigated at each hatchery by collecting between 108 and 421 environmental swab samples per hatchery (6990 samples in total from all hatcheries). An in-depth questionnaire on hatchery operations was completed for each hatchery, and results were incorporated into a risk factor analysis (univariable followed by multivariable mixed effects logistic regression) to identify factors associated with Salmonella occurrence. Overall, 6.0 % (416/6990) of environmental samples were Salmonella-positive and Salmonella was isolated from 17/22 hatcheries. Ten different serovars were isolated, the most common being S. Senftenberg and S. Mbandaka which are known hatchery colonisers. Sixty-four risk factor variables were investigated. Twenty-two of these were initially retained based on univariable analyses (p ≤ 0.25) and six were ultimately left in the final multivariable model (p ≤ 0.05). Salmonella detection was positively associated with having ≥30 hatchers in regular use compared to fewer (Odds ratio [OR] 23.7, 95 % confidence interval [CI] 6.7-84.2), storing trays in process rooms (OR 28.8, CI 7.8-106.3), drying set-up trolleys in corridors (OR 15.6, CI 5.9-41.4) and having skips located in enclosed areas (OR 8.99, CI 5.89-41.35). Using a closed waste disposal system was negatively associated with Salmonella detection (OR 0.08, CI 0.04-0.18) and the odds of detecting Salmonella in hatcheries with 31-60 total workers was lower compared to hatcheries with ≤30 staff (OR 0.16, CI 0.06-0.40). Despite the complexities of hatchery enterprises, changes to a relatively small number of features may significantly reduce the occurrence of hatchery contamination.

Identification of a Recently Dominant Sublineage in Salmonella 4,[5],12:i:- Sequence Type 34 Isolated From Food Animals in Japan

Salmonella enterica subsp. enterica serovar Typhimurium sequence type 34 (ST34) and its monophasic variant (Salmonella 4,[5],12:i:-) are among the most frequently isolated clones from both humans and animals worldwide. Our previous study demonstrated that Salmonella Typhimurium/4,[5],12:i:- strains isolated in Japan could be classified into nine clades and that clade 9 consisted of ST34 strains. In Japan, ST34/clade 9 was first found in the 1990s and has become predominant among food animals in recent years. In the present study, we analyzed the whole genome-based phylogenetic relationships and temporal information of 214 Salmonella Typhimurium/4,[5],12:i:- ST34/clade 9 strains isolated from 1998 to 2017 in Japan. The 214 strains were classified into two sublineages: the newly identified clade 9–2 diverged from clade 9 in the early 2000s and has predominated in recent years. Clonally expanding subclades in clades 9–1 or 9–2 lacked Gifsy-1 or HP1 prophages, respectively, and some strains in these subclades acquired plasmids encoding antimicrobial resistance genes. Additional genome reduction around the fljB gene encoding the phase 2-H antigen was generated by an IS26-mediated deletion adjacent to the transposon in clade 9–2. Although most of the clade 9 strains were isolated from cattle in Japan, the clonally expanding subclades in clade 9–2 (i.e., all and 24% strains of subclades 9–2a and 9–2b, respectively) were isolated from swine. The spread of clade 9 in recent years among food animals in Japan was responsible for the emergence of multiple host-adapted sublineages involving the clonally expanding subclades generated by mobile genetic element-mediated microevolution.