COVID-19 in meat and poultry facilities: a rapid review and lay media analysis

June 4, 2020

Quentin Durand-Moreau [1]
Anil Adisesh [2]
Graham Mackenzie [3]
Jonathan Bowley [4]
Sebastian Straube [1]
Xin Hui Chan [5]
Nathan Zelyas [6]
Trisha Greenhalgh [7]

[1] Division of Preventive Medicine, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada
[2] Division of Occupational Medicine, Department of Medicine, University of Toronto and St. Michael’s Hospital, Toronto, Canada
[3] GP Specialty Trainee, NHS Education for Scotland, UK
[4] Medical student, University of Nottingham, UK
[5] Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, UK
[6] Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Alberta, Canada
[7] Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK

Series editor: Trish Greenhalgh

On behalf of the Oxford COVID-19 Evidence Service Team
Centre for Evidence-Based Medicine, Nuffield Department of Primary Care Health Sciences
University of Oxford

Correspondence to Quentin Durand-Moreau
Division of Preventive Medicine, Department of Medicine, Faculty of Medicine and Dentistry
5 – 30 University Terrace
8303 – 112 St
Edmonton AB, T6G 2T4
durandmo@ualberta.ca

---

Abstract
During the 2020 COVID-19 pandemic, multiple clusters of cases were found to be related to meat and poultry facilities in several countries. Our aim was to identify specific risk factors in relation to this industry that would explain the larger number of COVID-19 cases. We conducted two reviews in parallel. (1) We carried out a lay media analysis searching for online newspapers articles and social media content using Twitter analytics tools. (2) We conducted a rapid narrative review by searching for selected keywords without any restrictions for dates or study designs. Multiple risk factors were identified. The working environment is favorable to SARS-CoV-2 persistence with numerous metallic surfaces, low temperatures, and medium to high relative humidity rates. Organizational settings may facilitate SARS-CoV-2 transmission: crowded workplaces do not allow appropriate physical distancing in production lines or in locker rooms. Appropriate self-isolation may be compromised by the designation of meat and poultry workforce as essential workers, or by the provision of financial incentives to keep working despite having symptoms. Time pressure compromises appropriate donning and doffing of personal protective equipment. Identification of risk factors specific to meat-packing industry may help development of tailored prevention plans in this industry. Outbreaks in such facilities may easily spread outside the workplaces, to workers’ family and contacts. This illustrates the intimate relationship between occupational safety and public health.

Introduction
Numerous coronavirus disease 2019 (COVID-19) outbreaks have been described in relation to meat and poultry processing facilities in different countries. In Alberta (Canada), for example, one in four of all cases of COVID-19 have been linked to 3 meat and poultry facilities as of May 10, 2020. In the United States, the Centers for Disease Control and Prevention (CDC) have reported 4,913 cases with 20 deaths in approximately 130,000 workers of 115 meat (beef, bison, lamb, pork, and other) and poultry processing facilities (Dyal et al., 2020). The subject is attracting considerable interest from the mainstream media and on social media, which depict the industry in mostly negative terms and offer various hypotheses about the chain of causation. We wanted to better understand why there are clusters of COVID-19 in these facilities. Our review questions were:

1. In relation to the lay literature and social media:

– What hypotheses are being proposed in the lay media about COVID-19 in relation to meat and poultry facilities, and how is the issue being framed?

2. In relation to the academic evidence base:

– What is the nature of the work process in meat and poultry facilities, and how is it undertaken?

– What is the impact of the working environment on the persistence and transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in meat and poultry facilities?

– How conducive are meat and poultry facilities to airborne viruses?

Methods
The study was undertaken in May 2020. We followed two parallel research approaches.

1. An analysis of lay media press articles and posts on Twitter was conducted in order to collect information about meat and poultry facilities in which COVID-19 outbreaks have been reported. To identify such articles, we undertook a Google ‘News’ search from November 2019 to May 2020 (JB, TG, SS), and also Twitter analytics (GM), the latter principally to take us to additional newspaper and web articles. Two authors (TG and JB) used a simple framework approach in Excel to chart broad themes and quotes from the press articles and Twitter ‘memes’, as well as provenance data (author, country, source, URL). These themes were woven into a narrative summary using the constant comparative method (i.e. adding each new source to an emerging picture of the whole) until saturation was reached (i.e. until new press articles did not contain any new themes).
2. Two authors (QDM and AA) performed a rapid search of PubMed and Medline databases without date restrictions using the following keywords: meat, poultry, facility, facilities, factory, factories, bioaerosol, aerosol, COVID-19, coronavirus disease 2019, SARS, severe acute respiratory syndrome, MERS, middle east respiratory syndrome, influenza. Relevant papers suggested by co-authors with specialist expertise were also included. All study types were eligible for inclusion. The papers were analyzed thematically and interpreted by discussion with topic experts in infection control (XHC) and microbiology (NZ).

For both analysis, texts in English, French and German were considered for analysis, reflecting the language skills of the research team members. The data were drawn together iteratively using narrative synthesis, progressively building an understanding of the multiple interacting factors at human, material, organizational and system level that could account for the high incidence of COVID-19 in relation to meat and poultry facilities. Findings from the media analysis fed into further searches of the academic literature and vice versa, and into refinement of the overall synthesis.

Given the nature of this research work, which is a review, we did not seek ethics approval.

Results
Hypotheses proposed in the lay media about COVID-19 in relation to meat and poultry facilities.

The lay media search identified a sample of 28 articles from 8 countries (USA, Canada, Ireland, Germany, Australia, France, Brazil, Spain); after this point, any additional articles identified did not contain any new themes. Whilst this was an incomplete selection of press coverage, it provided a useful overview of the kinds of cases and how they were reported in the press. Many of the reports were short descriptive accounts in local newspapers which focused on temporary closures, re-openings, illness among local workers, or (rarely) deaths in workers or their relatives. There were some longer reviews and commentaries in the national press in which causal hypotheses were explored and the perspectives of policymakers, industry spokespeople and workers were contrasted. Both local and national press tended to depict the stories negatively. Twitter analytics identified additional themes and are described elsewhere (Greenhalgh and Mackenzie, 2020).  Our overall analysis revealed the following broad themes:

– A high-risk industry for COVID-19 spread. Meat packing plants, abattoirs and slaughterhouses were depicted as major sources of local outbreaks, and sometimes as a key source of a national upsurge in cases when the disease was otherwise under control.  The largest reported cluster in our sample was 1500 cases linked to a single meat-packing plant in Alberta, Canada. A short article from the US Food and Environment Reporting Network linked to a regularly-updated interactive map of the United States showing – when accessed on May 15, 2020 – 213 separate outbreaks (9 farms and the remainder meat-packing plants) and 15,689 cases of COVID-19 (Douglas, 2020).

– Business pressures. The meat sector has been under intense competitive pressure for years, with small plants tending to close and be replaced by large plants owned by multinational companies achieving economies of scale. The industry was described as lacking the resilience to withstand an external shock.

In contrast, reports in trade magazines tended to depict a showcased local business successfully reopening while doing its best to protect workers, and the sector as a whole to be coping moderately well with the crisis.

– Prioritization of the supply chain. Some articles depicted the meat industry as under pressure to maintain production because its output was a necessary and “just-in-time” staple (i.e. the animals have been fattened to be killed at a particular time, and food would be wasted if the slaughter were deferred). Farmers were described as having to “euthanize” their animals. Industry officials expressed grave concerns about lack of meat on the shelves, and some articles expressed concern about the need to bring in “plant-based meat”. Some company officials depicted workers’ sickness absence as “absenteeism” which whilst strictly accurate has a negative connotation.

– A vulnerable workforce. Many articles described how meat workers are typically immigrants from several countries (e.g. Filipinos, Africans and Mexicans in the USA and Canada; Romanians and Bosnians in Germany; Filipinos, Africans, Romanians and Bulgarians in Ireland), sometimes with undocumented or uncertain resident status, rarely fluent in the local language and with limited health literacy and little or no understanding of their employment rights. Many would have feared losing their jobs, and so may have accepted low pay and poor conditions without protest.

– Inadequate health and safety measures. Many articles pointed out an absence of adequate infection control measures on the factory floor and indeed the impossibility of social distancing in the cramped conditions of some plants.  In some cases, masks were supplied but workers allegedly told they did not have to wear them.  Some articles described perverse incentives to workers to keep coming to work even when unwell (lack of sick pay, risk of redundancy and even financial bonuses for not missing shifts when sick). A few articles described extensive infection control measures in some facilities including widespread hand sanitizer, social distancing (sometimes with fiberglass or plastic barriers), staggered breaks, compulsory masking and handwashing, infrared scanners to detect fevers, worker education and supervision.

– Poor housing. Several articles (from Canada, USA, Ireland and Germany) described overcrowded living conditions, sometimes with immigrant workers living in “barrack-style” dormitories and travelling on crowded buses or vans to and from work. These conditions were sometimes depicted as potentially contributing to the spread of the virus but outwith the responsibility of the industry.

– Regulatory oversight. Measures to implement food standards regulations and align these with pandemic response measures such as contact tracing were described in broadly reassuring terms. A report from Germany described an unsuccessful attempt from a meat manufacturer to appeal against a closure order, but one from Brazil described a successful appeal.

– Political involvement. Especially in the United States, maintenance of the meat supply chain was depicted as a political as well as business priority. On 28th April 2020, the US Government declared the beef, chicken and pork industries as “essential infrastructure” and mandated that they stay open.  Social media threads talked of “the piece of chicken on the President’s plate” and alluded to the influence of political lobbying from large multinational meat suppliers.

In sum, recent newspaper articles depict an industry dominated by large multinational companies, working to “just-in-time” supply chain pressures and oriented to maximizing efficiency. The workforce is depicted as extremely vulnerable and the nature of the work as making it difficult or impossible to implement and follow high standards of infection control. Some plants appear to have seen extremely high levels of transmission. Poor-quality and at times overcrowded communal living and travelling conditions are depicted as exacerbating the risks. The tension between regulatory controls (which often require prolonged closures) and lobbying by powerful industry voices (which push for remaining open) plays out differently in different countries.

The above themes resonate strongly with a Human Rights Watch report published in September 2019 on workers’ rights in the meat industry (Humans Rights Watch, 2019), and with a recent report from the US Centers for Disease Control and Prevention, which included qualitative risk assessments in affected sites (Dyal et al., 2020). The CDC report states, for example, that:

 “Facility challenges included structural and operational practices that made it difficult to maintain a 6-foot (2-meter) distance while working, especially on production lines, and in nonproduction settings during breaks and while entering and exiting facilities. The pace and physical demands of processing work made adherence to face covering recommendations difficult, with some workers observed covering only their mouths and frequently readjusting their face coverings while working. Some sites were also observed to have difficulty adhering to the heightened cleaning and disinfection guidance recommended for all worksites to reduce SARS-CoV-2 transmission.”

Below, we present the findings from our academic analysis which explores the issues raised by the lay media and tests specific hypotheses raised by journalists and citizens.

Description of the work process in meat and poultry facilities

The following description is a general overview to help understand the overall process and identify the steps at higher risks of COVID-19 transmission.

The beef slaughtering process has several steps (International Labour Organization, 2011a): cattle receiving and holding, stunning, bleeding, head and shank removal, skinning, evisceration, splitting of carcass, final wash, chilling and processing. The first steps of the process are conducted with higher temperatures: e.g. the kill floor is especially hot and humid, the hair removal is done by passing the carcass through tanks of water heated to 58°C, and the final wash may be done with hot water or steam. The chilling is a cold step of the process, usually around 2°C. Freezers may generate temperatures as low as -40°C. The processing can be done either in the same facility or in dedicated processing facilities. Processing areas are kept around 4°C.

Specific personal protective equipment (PPE) used for this process includes metal mesh protective gloves to operate powered saws. Ventilation systems are critical to prevent heat stress to the workers. The working environment is considered to be wet overall. Slippery floors are identified as a usual occupational hazard. Biologic hazards include brucellosis, erysipeloid, leptospirosis, dermatophytosis, viral warts, tuberculosis and Q fever (International Labour Organization, 2011a). Verruca vulgaris can be spread by workers who handle contaminated tools and objects such as towels, knives, or worktables.

Poultry processing (which covers chicken and turkey) differs somewhat from the above description (International Labour Organization, 2011b). The steps of the process are receiving, live-hang, stunning, dressing, rehung, evisceration, chilling, cut-up, processing, overwrap and chilling. Dressing also requires the conveyor of birds to pass through a series of tanks of circulating hot water called scalers. The water used is usually chlorinated. Evisceration may imply the use of an automatic venting machine to push up the abdomen to cut the carcass properly. The chilling step, up to an hour, helps to lower the bird’s body temperature. Levels of free chlorine released may be high so workers may experience symptoms of skin, eye and respiratory irritation. After cut-up, some products may be packed in cartons covered with ice. Some products may go through a step of deboning. Temperature in the chiller may be as low as -2°C. An efficient exhaust ventilation system is an important part of the installation to reduce the impact of airborne particles. Knife sharpening may constitute up to 11% of the working time in poultry factories (Vézina et al., 2000).

Of relevance for our research question, both the above processes include steps with low and high temperatures, include working environments with high levels of relative humidity and requiring efficient ventilation systems (in theory). Also noteworthy is that both working environments are typically noisy.

Impact of the working environment on the persistence of SARS-CoV-2
SARS-CoV-2 survives longer in lower temperatures. A study by Chin et al. (2020) has shown that at 22°C, the virus titer decreases by 3 logs in 7 days and is undetectable in 14 days; whereas at 4°C, the virus titer only decreases by 0.4 logs after 14 days. It also persists for longer on materials that are common in meat and poultry facilities such as plastic or stainless steel (undetectable after 7 days) compared to paper (undetectable after 3 hours) or cloth (undetectable after 2 days).

It has been shown that high (>60%) and low relative humidity (<40%) allow viability of influenza virus in droplets, whereas with relative humidity levels between 40% and 60%, the virus becomes inactivated (Moriyama et al., 2020). Stability of the coronavirus 229E (HCoV-229E) was best at low temperatures with medium relative humidity (Ijaz et al., 1985) (Table 1).

	Virus half-life (hours)
Temperature (°C)	Low Relative Humidity (%) 30 ± 5%	Medium Relative Humidity (%) 50 ± 5%	High Relative Humidity (%) 80 ± 5%
6 ± 1	34.46 ± 3.21	102.53 ± 9.38	86.01 ± 5.28
20 ± 1	26.76 ± 6.21	67.33 ± 8.24	3.34 ± 0.16

Table 1: Half-life of aerosolized coronavirus 229E in experimental conditions reported by Ijaz, Brunner, Sattar, Nair and Johnson-Lussenburg in 1985

Transmission of influenza virus was more efficient in guinea pigs at 5°C, compared to 20°C (Moriyama et al., 2020). Low temperatures allow transmission even with high relative humidity levels (80%) but such transmission does not appear to occur at high temperatures (20°C). This may be explained by the reduced mucociliary clearance and improved stability of influenza virions at 5°C. The absolute humidity level may also play a role. At a maintained relative humidity level of 80%, the absolute level at 5°C is less than at 20°C (5.5 g/m³ vs. 14 g/m³). In mice, it seems that the best relative humidity level to prevent aerosol respiratory viral transmission is between 40% and 60%. Some authors hypothesize that a high humidity and low temperature environment would promote the viability of SARS-CoV-2 in droplets (Moriyama et al. 2020).

Personal protective equipment properly worn and removed is likely to reduce transmission of the virus in this environment. However, there may be barriers to the proper and consistent wearing of such equipment, even when it is provided.  Personal protective equipment may not be tailored for long-term use or use in hot and wet working environments. Wearing surgical masks may cause skin irritation or pain over the ears (Kashyap  et al., 2020). Workers may be tempted to develop strategies to help them wear PPE items over a long day of work (e.g. using cardboard to reduce the pressure of elastic cords of masks over the ears or deciding to self-assess the risk and wearing items of PPE only when they feel the risk is high). Although such strategies may help the worker to comply with wearing PPE, they may also reduce its efficacy and safety. In particular, wet environments may increase the need for replacing masks more often, as they become wet. This may not be possible or prioritized, either because of shortages in PPE or because of the loss of productive time associated with switching to a new item of PPE to comply with hygiene standards. Furthermore, some of the highest levels of SARS-CoV-2 in healthcare have been measured in areas where PPE is removed. This doffing stage is also a high risk for physical contamination of unprotected skin and clothing. It is questionable whether adequate training in PPE removal is always provided to workers in meat plants, where the focus tends to be on product safety.

Another aspect of the working environment is noise. Meat and poultry factories are recognized as noise-exposed environments with levels over 100dBA recorded from the slaughter line and from power saws, 90dBA from chopper bowls and 85-90dBA from packing machinery (Iulietto et al., 2018). A laser light-scattering experiment has recently helped visualize droplets produced when a person is speaking (Anfinrud et al., 2020). Each droplet produced one flash as it passed through the light sheet. The interesting finding from this experiment is that the study authors reported that the number of flashes increased with the loudness of speech. However, it should be noted that the experiment placed emphasis on the visualization of droplets and does not allow direct conclusions to be drawn on the transmission of viruses.

Previous studies have shown that the emission of particles during normal speech increases with the sound intensity of the speech, from 1 to 50 particles per second and 0.06 to 3 particles per cm³ (Gehanno et al., 2020). Despite noise, workers in meat and poultry facilities still need to communicate and talk to each other and will be likely to shout where the noise level is 85dBA or more.  This may increase the aerosolization of SARS-CoV-2 by infected workers. This hypothesis needs further research.

Airborne transmission of viruses in meat and poultry factories
Meat and poultry factories are known to have high levels of aerosols. Bioaerosols in such facilities contain varying levels of dust, proteins, endotoxin, bacteria (Just et al., 2011), archaea (Just et al., 2013) and fungi (Gibbs et al., 2004). There is existing literature about contamination of meat plant aerosols by human bacteria. Many bacteria found in poultry environments may cause infections in both animals and humans (Just et al., 2012). Shale et al. (2006) have studied airborne concentration of staphylococci in deboning rooms of red-meat abattoirs in South Africa. They identified Staphylococcus species (Staphylococcus capitis and Staphylococcus epidermidis) that suggested a large contribution of human-borne bacteria to the bioaerosol population. The risk of transmission is unlikely to be equal in all working areas of abattoirs. Meat and poultry factories contain wet working environments with aerosols in some areas, including areas in which the initial steps of meat processing take place (slaughter and cutting). Aerosols from animals may be generated on the hide-removal floors and bacteria may attach to carcasses (Jericho et al., 2000).

The possibility of airborne transmission of SARS-CoV-2 is now raised (Gehanno et al., 2020). Animal-shed aerosol particles and indoor air transmission of infectious diseases from animals to human has been documented with bacteria (brucellosis, Q fever), or fungi (histoplasmosis) (Spendlove and Fannin, 1983). In 1999 an outbreak of Nipah virus occurred amongst pig abattoir workers in Singapore, the mode of transmission is not fully understood but respiratory secretions are a possible route (Paton et al., 1999). The role of such shedding is hypothetical as direct transmission of COVID-19 from animals to humans has not been reported to date (Anses, 2020). However, some authors (Opriessnig and Huang, 2020) postulate that transmission from pigs to humans may nonetheless be a possibility. A study has found SARS-CoV (the virus responsible for Severe Acute Respiratory Syndrome, SARS) RNA in a pig, but that pigs do not play a role in virus amplification (Weingartl et al., 2004). However, it has been found that pigs may amplify MERS-CoV (Vergara-Alert et al., 2017). SARS-CoV-2 may use the porcine Angiotensin-Converting Enzyme 2 to enter the cell in vitro. This possibility of zoonotic transmission should be monitored, although it is not currently considered to be likely.

Infected workers may unwittingly contaminate working environments with meat and poultry factories having specific characteristics that may increase airborne transmission of aerosols. Some authors consider that food handlers are the primary sources of indoor bioaerosols in the food industry. Employees may carry and distribute micro-organisms through their clothes, by contact, coughing or sneezing (Shale et al., 2006). Poor ventilation systems may play a significant role as well as improper maintenance and sanitation.

There are several publications about airborne transmission of influenza viruses in poultry markets. A study by Chen et al. (2010) measured ambient influenza and avian influenza virus RNA transported by Asian dust storms in a wet poultry market in Taiwan using real-time quantitative PCR (qPCR). Another study from Bui et al. (2018) detected cultivable influenza A virus (not H5 or H7) in 4-hour air samples in Hanoi’s (Vietnam) largest live poultry market. Several publications have studied influenza viruses in poultry production facilities and in premises. Zoonotic viruses (such as avian influenza viruses) have been detected in dust generated in poultry production. A study by Schofield et al. (2005) was conducted in three chicken farms in Abbotsford, British Columbia (Canada), in 2004 concerning the avian influenza virus, H7N3. This virus has been reported to have infected two poultry workers previously. The purpose of the study was to determine whether the virus could spread by wind. Detection of the virus was performed using real-time reverse transcriptase-PCR (RT-PCR) and viability was assessed by cell culture. These authors faced technical issues: they hypothesized that collection of the virus during daylight hours may have reduced the chances in finding active virus in the sample, as any viable virus would probably have been inactivated by sunlight. However, they measured a viral load of 292 viral doses / m³ of barn air. Highly pathogenic avian influenza (influenza type A, carrying the hemagglutinin H5 or H7), including the H5N1 influenza virus has been studied experimentally in chicken facilities in the Netherlands (Spekreijse et al., 2012). Airborne transmission of H5N1 turkey influenza virus has been documented using identification of viral RNA in air samples, and with the infection of naive chickens exposed to air emitted by infected chickens. The rate of indirect transmission was 20-fold lower than the rate of direct transmission between chickens housed in the same cage. Indirect transmission can occur between chickens housed in two separated rooms, with ventilation forcing the air from the room with infected animals to the other room. However, this might be less efficient. These authors hypothesize that the amount of viable virus in their air samples was lower than measured in RT-qPCR because this method also detects inactivated virus particles. However, viable virus was detected, but they were not able to measure the virus titer.

Data about other viruses are scarcer. Metagenomic analysis of air samples has revealed the DNA of WU polyomavirus and human papillomavirus 120 in slaughterhouses in New Zealand, but such analysis has considerable limitations as it is uncertain whether the sample directly measured the workers’ own exhalation rather than viruses present in the workplace air environment (Hall et al., 2013). High throughput genomic sequencing of bioaerosols in broiler chicken production facilities in the United States showed the presence of viral nucleic acid (O’Brien et al., 2016). Viruses were the second most abundant domain in the inhalable litter sampling (25%). Some of the viruses identified belonged to orders that include viruses pathogenic to humans (Picornavirales, Mononegavirales and Herpesvirales). These authors do not report the presence of any Nidovirales virus, which is the order to which the Coronaviridae family belongs. This study did not report any results on viral cultures.

A cross-sectional study in Sarawak (Malaysia) was conducted among pig farms, abattoirs and animal markets in 2017 to determine whether some viruses (including coronaviruses) are aerosolized at the animal interface and if workers are carrying those viruses (Borkenhagen et al., 2018). Bioaerosol sampling was conducted using a National Institute for Occupational Safety and Health (NIOSH) sampler. Nucleic acid amplification techniques were used to detect pathogenic adenovirus, coronavirus, encephalomyocarditis virus, enterovirus, influenza A – D, porcine rotaviruses A and C, and porcine circovirus 2. All of the farms and the markets were open air. Coronaviruses were detected in 2 workers (2.6%) but not in any of the aerosol samples. However, results for the porcine circovirus 2 (PCV2) are interesting.  These authors showed that PCV2 was found in 5.1% of the workers’ nasal wash samples, in 38.2% of the pig fecal samples and in 23.1% of the farm/abattoir bioaerosol samples. The positive bioaerosol samples were in the >4 μm particle size fraction, and the size fraction below 4 μm were all negative. Workers involved in wean-to-finish production had a higher risk of testing positive for any of the viruses sought (OR 9.33, 95% CI [1.12;78.97]).

Discussion
Meat and poultry facilities exhibit several risk factors that may increase the spread of COVID-19 and the likely development of clusters:

Essential workers. Since these industries were considered as essential in most cases, most meat and poultry facilities workers have not discontinued their activities.

Language barriers in a dependent relationship: Many workers are immigrants, especially in North America or in Germany; they may have trouble understanding recommendations provided by health authorities when not in their first language. They may comply with the conflicting pressures given by their superiors (e.g. going to work despite having symptoms of COVID-19), being under pressure to continue working.

Precarious work. Some workers are low-paid and have minimal or no benefits. They may be reluctant to properly self-isolate if they are promised bonuses for staying in their post. Depending on local policies, providing bonuses for staying at work may contravene recommended self-isolation by health authorities, especially for low-paid workers.

Overcrowded housing and transportation. Workers may share crowded dwellings or transportation means that may help spread SARS-CoV-2.

Time pressure. Work pace and physical demands may be so high that it may be difficult for workers to comply with proper PPE use or with upgraded hygiene standards.

Inability to practice proper physical distancing. The spatial settings of production lines, break rooms and locker rooms may make proper physical distancing difficult.

Temperature and humidity rates. The low temperatures (4 to 6°C) and medium to high relative humidity rates (RH >50%) that occur at some steps of the process in meat and poultry facilities may help coronaviruses remain stable longer.

Transmission routes. SARS-CoV-2 RNA has been found to be transmitted airborne in a hospital in Wuhan (Liu et al., 2020). It is possible that infected workers can release SARS-CoV-2 into the air of the working environment in the meat and poultry facilities. It is also a possibility that some animals, such as pigs, may constitute a reservoir for SARS-CoV-2, although not proven to date. Ventilation systems may influence the spread of viruses shed in aerosols. Speaking loudly in a noisy environment increases the release of droplets, and this may facilitate the spread of SARS-CoV-2.

Surfaces. It has been shown that SARS-CoV-2 persists longer on some metallic or plastic surfaces than on cloth or paper surfaces. Meat and poultry factories usually comprise many stainless-steel surfaces, from hooks to hang birds to mesh metal gloves as PPE when splitting the carcasses.

Personal Protective Equipment. Items of PPE may be subject to shortage during the COVID-19 pandemic, as in other industries. Changing items of PPE appropriately requires a high level of training and may be challenging as masks may become wet quite quickly, and there is a tension between compliance with hygiene protocols and production activities.

This narrative review using combined methods presents a certain number of limitations. Research on COVID-19 currently generates a very high number of publications with variable degrees of quality. It is thus very likely that we are not able to capture all of the evidence available. We also reported several publications about Influenza viruses, which may be considered as a proxy in this context. However, we do believe that it is helpful to share this rapid narrative review to help stakeholders build prevention plans rapidly. Further systematic reviews will certainly be helpful by addressing different and narrower research questions (Greenhalgh et al., 2018).

The confluence of several diverse and interacting risk factors means that prevention plans should be designed accordingly. Providing only items of PPE (i.e. limiting the prevention strategy to the individual level) cannot effectively reduce the overall risk if environmental, organizational, and administrative factors are not targeted as well. It is worth emphasizing the importance of social factors common in precarious work such as low wages and poor living conditions. The situation is a threat to the life and health of meat plant workers, public health, food security, and employment security. Valuing meat plant workers seems to be a central theme, such that enlightened employment practices that support workers and reduce vulnerability would likely enhance production, product safety, and product value. As recently as Human Rights Watch (2019) reported, “extremely difficult working conditions, including instances where workers said they were pushed to work past their physical and mental limits. Some workers also reported difficulties in accessing adequate health care, at times, waiting weeks or even months before being referred to physicians after visits to plant health facilities”. Against such a background it is perhaps unsurprising that COVID-19 has found fertile ground for rapid growth.

The COVID-19 crisis could be an opportunity to remember that the International Labour Organization has adopted the resolution on “advancing social justice through decent work” in 2016 (International Labour Organisation, 2016). The COVID-19 pandemic in this industrial sector as in others highlights the intersectionality of infection requiring the combined approaches of occupational health and public health.

Funding and links of interest: The authors report that there was no funding source for the work that resulted in the article or the preparation of the article.

Dr. Straube reports grants/grants pending from the Workers’ Compensation Board of Alberta, fees for lectures from the Workers’ Compensation Board of Alberta, and honoraria from the Canadian Board of Occupational Medicine and WorkSafeBC; all outside the submitted work. Dr. Durand-Moreau reports grands from the Workers’ Compensation Board of Alberta (as a Co-Investigator), outside of the submitted work.

References

Anfinrud P, Stadnytskyi V, Bax CE, Bax A. (2020) Visualizing Speech-Generated Oral Fluid Droplets with Laser Light Scattering. New England Journal of Medicine;382(21):2061-2063. doi:10.1056/nejmc2007800.

Anses (2020). Avis de l’Agence nationale de sécurité sanitaire de l’alimentation, de l’environnement et du travail relatif à une demande urgente sur certains risques liés au COVID-19, Accessible at: https://www.anses.fr/en/system/files/SABA2020SA0037-1.pdf . Accessed May 14, 2020.

Borkenhagen LK, Mallinson KA, Tsao RW, et al. (2018) Surveillance for respiratory and diarrheal pathogens at the human-pig interface in Sarawak, Malaysia. PLoS ONE;13(7):e0201295. doi:10.1371/journal.pone.0201295.

Bui VN, Nguyen TT, Nguyen-Viet H, et al. (2018) Bioaerosol Sampling to Detect Avian Influenza Virus in Hanoi’s Largest Live Poultry Market. Clinical Infectious Diseases. 2018;68(6):972-975. doi:10.1093/cid/ciy583.

Chen P-S, Tsai FT, Lin CK, et al. (2010) Ambient Influenza and Avian Influenza Virus during Dust Storm Days and Background Days. Environmental Health Perspectives;118(9):1211-1216. doi:10.1289/ehp.0901782.

Chin AWH, Chu JTS, Perera MRA, et al. (2020) Stability of SARS-CoV-2 in different environmental conditions. The Lancet Microbe;1(1):e10. doi:10.1016/s2666-5247(20)30003-3.

Douglas L. (2020). Mapping Covid-19 outbreaks in the food system, Food and Environment Reporting Network. Food and Environment Reporting Network. Available at: https://thefern.org/2020/04/mapping-covid-19-in-meat-and-food-processing-plants/. Accessed 20 July 2020.

Dyal JW, Grant MP, Broadwater K, et al. (2020) COVID-19 Among Workers in Meat and Poultry Processing Facilities ― 19 States, April 2020. MMWR Morbidity and Mortality Weekly Report;69(18). doi:10.15585/mmwr.mm6918e3

Gehanno J, Bonneterre V, Andujar P, et al. (2020) Arguments pour une possible transmission par voie aérienne du SARS-CoV-2 dans la crise COVID-19. Archives des Maladies Professionnelles et de l’Environnement. In Press. doi:10.1016/j.admp.2020.04.018.

Gibbs SG, Green CF, Tarwater PM, Scarpino PV. Airborne Antibiotic Resistant and Nonresistant Bacteria and Fungi Recovered from Two Swine Herd Confined Animal Feeding Operations. Journal of Occupational and Environmental Hygiene. 2004;1(11):699-706. doi:10.1080/15459620490515824.

Greenhalgh T, Mackenzie G. (2020) Learning about covid-19 outbreaks from non-viral tweets, The BMJ Opinion, Available at: https://blogs.bmj.com/bmj/2020/05/29/learning-about-covid-19-outbreaks-from-non-viral-tweets/ Accessed June 8, 2020

Greenhalgh T, Thorne S, Malterud K. (2018) Time to challenge the spurious hierarchy of systematic over narrative reviews? Eur J Clin Invest;48:e12931. https://doi.org/10.1111/eci.12931

Hall RJ, Leblanc-Maridor M, Wang J, et al. (2013) Metagenomic Detection of Viruses in Aerosol Samples from Workers in Animal Slaughterhouses. PLoS ONE;8(8):e72226. doi:10.1371/journal.pone.0072226.

Human Rights Watch (2019). “When We’re Dead and Buried, Our Bones Will Keep Hurting”: Workers’ Rights Under Threat in US Meat and Poultry Plants. Available at: https://www.hrw.org/report/2019/09/04/when-were-dead-and-buried-our-bones-will-keep-hurting/workers-rights-under-threat, Accessed June 2, 2020

Ijaz MK, Brunner AH, Sattar SA, Nair RC, Johnson-Lussenburg CM. (1985) Survival Characteristics of Airborne Human Coronavirus 229E. Journal of General Virology;66(12):2743-2748. doi:10.1099/0022-1317-66-12-2743.

International Labour Organisation. (2016) Resolution on Advancing Social Justice through Decent Work. 105th Session of the International Labour Conference. Available at: https://www.ilo.org/ilc/ILCSessions/previous-sessions/105/texts-adopted/WCMS_497583/lang–en/index.htm Accessed May 15, 2020.

International Labour Organization (2011a). Meatpacking/Processing. Encyclopaedia of Occupational Health and Safety. 4^th Edition. Accessible at: https://iloencyclopaedia.org/part-x-96841/food-industry/food-processing-sectors/item/860-meatpacking/processing . Accessed May 14, 2020

International Labour Organization (2011b). Poultry Processing. Encyclopaedia of Occupational Health and Safety. 4^th Edition. Available online: https://iloencyclopaedia.org/part-x-96841/food-industry/food-processing-sectors/item/861-poultry-processing Accessed May 14, 2020.

Iulietto MF, Sechi P, Mansi Gaudenzi C, et al. (2018) Noise assessment in slaughterhouses by means of a smartphone app. Italian Journal of Food Safety;7(2). doi:10.4081/ijfs.2018.7053.

Jericho KWF, Ho J, Kozub GC. (2000) Aerobiology of a High-Line Speed Cattle Abattoir. Journal of Food Protection;63(11):1523-1528. doi:10.4315/0362-028x-63.11.1523.

Just N, Blais Lecours P, Marcoux-Voiselle M, et al. (2013) Archaeal characterization of bioaerosols from cage-housed and floor-housed poultry operations. Canadian Journal of Microbiology;59(1):46-50. doi:10.1139/cjm-2012-0305.

Just N, Kirychuck S, Gilbert Y et al. (2011) Bacterial Diversity Characterization of Bioaerosols From Cage-Housed and Floor-Housed Poultry Operations, Environ Res;111(4):492-8 doi: 10.1016/j.envres.2011.01.009

Just NA, Létourneau V, Kirychuk SP et al.(2012), Potentially Pathogenic Bacteria and Antimicrobial Resistance in Bioaerosols from Cage-Housed and Floor-Housed Poultry Operations, Annals of work exposure and health;56(4):440-449 doi:10.1093/annhyg/mer105.

Kashyap A, Singh K, Sabat D, Maini L. (2020) Fast and economic cardboard cutout use to increase compliance of face mask wear during COVID-19 pandemic. Journal of Clinical Orthopaedics and Trauma;11:S298-S300 doi:10.1016/j.jcot.2020.04.032.

Liu Y, Ning Z, Chen Y, et al. (2020) Aerodynamic analysis of SARS-CoV-2 in two Wuhan hospitals. Nature; 582(7813):557-560. doi:10.1038/s41586-020-2271-3.

Moriyama M, Hugentobler WJ, Iwasaki A. (2020) Seasonality of Respiratory Viral Infections. Annual Review of Virology;7(1). doi:10.1146/annurev-virology-012420-022445.

O’Brien KM, Chimenti MS, Farnell M, et al. (2016) High throughput genomic sequencing of bioaerosols in broiler chicken production facilities. Microbial Biotechnology;9(6):782-791. doi:10.1111/1751-7915.12380.

Opriessnig T, Huang Y. (2020) Coronavirus disease 2019 (COVID‐19) outbreak: Could pigs be vectors for human infections? Xenotransplantation;27(2). doi:10.1111/xen.12591.

Paton NI, Leo YS, Zaki SR, et al. (1999) Outbreak of Nipah-virus infection among abattoir workers in Singapore. The Lancet.;354(9186):1253-1256. doi:10.1016/s0140-6736(99)04379-2.

Schofield L, Ho J, Kournikakis B, Booth T. (2005) Avian Influenza Sampling Campaign in the British Columbia Fraser Valley, 9–19 April 2004: Sampling of rare biological events. DRDC Suffield TR 2005-032. Defence R&D Canada – Suffield

Shale K, Lues JE, Venter P, Buys EM. (2006) The distribution of Staphylococci in bioaerosols from red-meat abattoirs. J Environ Health;69(4):25‐32.

Spekreijse D, Bouma A, Koch G, Stegeman A. (2012) Quantification of dust-borne transmission of highly pathogenic avian influenza virus between chickens. Influenza and Other Respiratory Viruses;7(2):132-138. doi:10.1111/j.1750-2659.2012.00362.x.

Spendlove JC, Fannin KF. (1983) Source, significance, and control of indoor microbial aerosols: human health aspects. Public Health Rep;98(3):229‐244.

Vergara-Alert J, van den Brand JMA, Widagdo W, et al. (2017) Livestock Susceptibility to Infection with Middle East Respiratory Syndrome Coronavirus. Emerging Infectious Diseases;23(2):232-240. doi:10.3201/eid2302.161239.

Vézina N, Prévost M, Lajoie A. (2000) Élaboration d’une formation à l’affûtage des couteaux dans six usines d’abattage et de transformation du porc : une étude ergonomique. Rapport de l’IRSST R-243, Accessible at: https://www.irsst.qc.ca/media/documents/PubIRSST/R-243.pdf?v=2020-05-19 Accessed May 19, 2020.

Weingartl HM, Copps J, Drebot MA, et al. (2004) Susceptibility of Pigs and Chickens to SARS Coronavirus. Emerging Infectious Diseases;10(2):179-184. doi:10.3201/eid1002.030677.

Authors’ contributions: QDM and TG conceptualized and designed the work. GM, JB, SS and TG worked on lay media analysis, GM provided a systematic review of Twitter postings, TG summarized and drafted the section of the manuscript dedicated to lay media analysis. QDM, AA, XHC worked on the rapid review per se. QDM summarized and drafted the section of the manuscript dedicated to the rapid review. NZ provided expert review as a medical microbiologist of the overall manuscript. All authors have reviewed, and corrected the manuscript.