Non-typhoidal salmonella contamination along the pork value chain in a rural East African setting: a cross-sectional study

ABSTRACT Background Non-typhoidal Salmonella (NTS) is a serious foodborne pathogen that has previously been isolated from pigs presented for slaughter in a rural pork value chain in western Kenya. Methods To understand varying NTS contamination along the value chain we assessed prevalence at slaughter, transport and retail. Suspect isolates from culture were confirmed using matrix-assisted laser desorption-ionization time of flight mass spectrometry. Results Prevalence on pig carcasses, meat transportation containers, retailed raw and cooked pork and accompanying side salads was 18.1%, 23.9%, 28.0%, 1.9% and 8.6%, respectively. Conclusion NTS contamination is propagated along the pork value chain in rural western Kenya, demonstrating the need for improved hygiene measures to prevent human exposure.


Introduction
The Institute for Health Metrics and Evaluation estimated that, in 2010, enteric non-typhoidal salmonellosis accounted for 4 847 000 disability-adjusted life years lost and 81 300 deaths worldwide, with disproportionately higher share/burden observed in sub-Saharan Africa. 1 Pigs are considered among the most significant sources of infection, especially as asymptomatic carriers of non-typhoidal Salmonella (NTS), which poses challenges in preventing their entry into the food chain. 2 A recent study isolated NTS from 12.7% of pigs being presented for slaughter in rural western Kenya, including serotypes that cause human gastroenteritis. 3Previous studies in the same region have also identified NTS in diarrheal patients, with Akullian et al. reporting an isolation rate of 12.4% from stool samples collected from 2007 to 2014, and Shapiro et al. reporting a rate of 4% from stool samples collected during 1997 and 1998. 4 , 5Crump et al. 6 proposed that meat pathways are potential sources of NTS infections in East Africa, and the burden of NTS attributed to pork consumption has been found to be highest in the African sub-regions. 7onsidering the awareness of the proportion of Salmonellainfected pigs entering this value chain, the present study aimed to investigate how the prevalence of contamination changes throughout the various stages of the food system, starting from the dispatch of dressed carcasses from slaughter to the point of retail, where there is contact with pork consumers.The results are an essential prerequisite for quantifying the public health risk associated with this value chain, enabling the design of targeted and efficient mitigation strategies, particularly in anticipation of a significant increase in pork consumption within this region in the coming decades. 8

Materials and Methods
The methodology used in this study is outlined briefly here, with additional details available in the supplemental material.A crosssectional survey was conducted, and simple random sampling was employed to collect samples at each stage of the pork supply chain from slaughter to retail in Busia County, western Kenya.This region was chosen for the study as existing data are already available for contamination at slaughter 3 and because the informal pork value chain is growing rapidly in this high-density human population. 8. M. Gichuyia et al.Sterile, moist gauze swabs were used to sample 144 postslaughter carcasses (specifically from the ham, back, belly and jowl), as well as 113 meat transportation containers (sampling from three sides and the bottom of each container), each covering a total surface area of 400 cm 2 .During retail, sampling involved requesting pork retailers to package 250-g portions of raw and/or cooked pork as well as vegetable side salads using their regular equipment, as they routinely carry out for their customers.A total of 247 retail raw pork samples from 125 raw pork retail outlets, as well as 104 cooked pork samples and 81 raw vegetable salad samples from 104 cooked pork retail outlets, were collected.
Salmonella colonies were isolated by culture using selective media according to standard protocols outlined by the International Organization for Standardization (ISO 6579:2002).These were further confirmed as Salmonella spp.using the matrixassisted laser desorption-ionization time of flight mass spectrometry (MALDI-TOF MS) direct colony technique. 9o summarize the prevalence data of Salmonella at each stage, proportions with 95% CIs were computed using R version 4.1.3software (R Core Team, Vienna, Austria). 10

Results and Discussion
Table 1 summarizes the prevalence of Salmonella along the value chain nodes.
The prevalence of Salmonella on dressed pig carcasses at the slaughterhouses is comparable with that reported by Wilson et al., who examined mesenteric lymph nodes and intestinal contents from slaughter pigs entering this value chain. 3This is consistent with studies indicating that infected slaughter pigs raise the likelihood of carcass contamination. 2 Thus, this emphasizes the need to implement interventions that reduce the entry of infected pigs into the slaughter process, minimize carcass contamination during slaughter and enable decontamination of carcasses before they leave the slaughterhouses. 2he contamination of nearly one-quarter of the meat transportation containers before loading freshly slaughtered carcasses underscores the potential risk of cross-contamination of carcasses with Salmonella at this particular stage.The highest levels of contamination were observed at the raw pork retail outlets, highlighting this node as a potential source of infection to con-sumers purchasing pork for household consumption, especially because some pork consumers in this population previously indicated that they frequently eat undercooked pork at home. 11The lowest prevalence was observed from cooked pork samples and can be attributed to cooking, which reduces Salmonella prevalence via thermal inactivation.The contaminated raw vegetable salads, however, are an illustration of cross-contamination during food preparation, and hygiene messaging could be targeted to cooked pork retailers.
These findings have significant public health implications.In Wilson et al.'s study, 3 13 NTS serovars were identified in slaughter pigs within this value chain, including Enteritidis, Hadar, Heidelberg and Uganda, which have been associated with severe gastroenteritis outbreaks worldwide, resulting in illness and even death. 2 , 12The assumption made in this study is that these same serovars are carried along the value chain to the consumer level, thereby making the observed contamination levels a cause for concern.
The increasing demand for pork in this region is outpacing the development of public health infrastructure required to support this growth. 8Without adequate mitigation strategies, pork may become an increasingly significant source of NTS infections.Further comprehensive investigations are required to quantify the actual risk posed to pork consumers in this region, assess the contribution of each stage of production to this risk and analyze the behavior of the individuals involved at each stage of the value chain.Such studies are essential for designing targeted interventions aimed at reducing the risk to consumers, particularly considering that this value chain operates in one of Kenya's most densely populated and low-income regions characterized by a high burden of diarrheal diseases and co-endemic pathogens. 13

Conclusion
The findings of this study demonstrate the propagation of NTS contamination along the pork value chain in rural western Kenya, thereby highlighting the potential risk to pork consumers and necessitating additional investigations to evaluate this risk and identify influential factors at each stage.
Furthermore, these findings contribute to filling the data gap regarding the incidence of significant foodborne disease hazards in animal source food value chains in developing countries.More Transactions of the Royal Society of Tropical Medicine and Hygiene of these data are necessary to improve our estimation of the global foodborne disease burden.
Authors' contributions: CMG, LFT and EMF designed the study.CMG and CM implemented the study.LO designed and implemented the MALDI-TOF protocol and interpreted the assays.CMG analyzed the data and drafted the manuscript; all the authors revised, read and approved the final manuscript.JOO, PBG, LFT and EMF provided supervision for the design and implementation, and EMF acquired the funds for this work.CMG and EMF are the guarantors of the paper.CMG is a PhD student registered at the University of Nairobi (UoN), Kenya and attached to ILRI as a graduate fellow.She was supervised by JOO and PBG at UoN and by LFT and EMF at ILRI.EMF acquired the funds for this work under the Zoonoses in Livestock in Kenya project (ZooLinK).

Acknowledgements:
The authors would like to sincerely thank the Busia pork value chain stakeholders for their participation as well as the veterinary interns attached to the Zoonoses in Livestock in Kenya project who assisted in field data collection and laboratory analysis as part of their training.).The committee is licensed by the National Commission for Science, Technology, and Innovation (NACOSTI) in Kenya and also approved by the Federal wide Assurance for the Protection of Human Subjects in the USA.Additionally, informed and signed consent was obtained from all participants prior to sample collection from their slaughter facilities, meat transportation containers and pork retail outlets.

Funding:
This work was supported by the Biotechnology and Biological Sciences Research Council, the Department for International Development, the Economic & Social Research Council, the Medical Research Council, the Natural Environment Research Council, and the Defense Science & Technology Laboratory, under the Zoonoses and Emerging Livestock Systems (ZELS) programme [grant reference BB/L019019/1].This study also received support from the CGIAR One Health initiative 'Protecting Human Health Through a One Health Approach', which was supported by contributors to the CGIAR Trust Fund ( https://www.cgiar.org/funders/ ).Competing interests: None.Ethical approval: Data were collected after approval by the ILRI Institutional Research Ethics Committee (ref.ILRI-IREC2018-13/1

Table 1 .
Salmonella prevalence across the different value chain nodes