Occupational biological risk assessment is particularly difficult because of the wide range of pathogenic agents, different types of exposure, the stochastic nature of infections, the presence of workers with different levels of susceptibility to risk, and the lack of epidemiological data that would permit the establishment of certain limits of exposure. Biological agents can pose serious risks in many work settings, mainly when they are transmitted through bioaerosols, defined as airborne particles with a biological origin [1
Bioaerosols include bacteria, fungi, pollen, viruses, their fragments and byproducts (e.g., endotoxins and mycotoxins), and products or fragments from living beings, such as animal allergens [2
]. Health effects that are caused by exposure to bioaerosols and can severely impact public health include infectious diseases, acute toxic effects, allergies, and cancer. The most widely studied and probably most important bioaerosol-associated health effects [3
] are respiratory symptoms and impairment in lung function.
Despite recognizing the importance of bioaerosol exposure with regard to human health, the assessment of occupational biological risk is generally limited to evaluating “potential exposure”, without any quantitative estimate. This evaluation is based on the concept that biological risk is stochastic thus even a single microbial infectious particle can theoretically cause infection in an exposed worker. Nevertheless, the probability of this event depends on the infectivity of the pathogen, its concentration in the air, and exposure time. The risk assessment output is then usually expressed as a level (low, medium, high) and represented in risk matrices combining magnitude and probability scores [4
When risk comes from environmental contamination, such as exposure to bioaerosols, microbiological monitoring can help define levels of risk and the sources and spread of biological agents [5
]. Usually, this monitoring considers a generic contamination, measured as total bacterial count and total fungal count (TBC and TFC, respectively). Less frequently, and only in situations where faecal contamination occurs, an Escherichia coli
-coliform count (ECC) is used. These indicators are not representative of all pathogens and cannot be correlated with the real risk, but rather only with general hygienic conditions [6
In some cases, monitoring includes pathogens; however, owing to the lack of information on dose-response relationships and threshold limit values, risk management is based on the “precautionary principle”, and prescribes the adoption of preventive measures independent of the dose of the exposure. Occupational exposure limits (OELs) are rarely recommended in legislation, but several countries have proposed limits, mainly for total number of bacteria, gram-negative bacteria, bacterial endotoxin, and fungi [4
Based on experience from biological risk assessments for drinking water and food [9
], one possible way of estimating the risk from infectious agents is to use the Quantitative Microbial Risk Assessment (QMRA) methodology. This is a process of estimating the risk from exposure to pathogenic microorganisms (or a medium in which pathogens occur) by combining dose-response information for the infectious agent with information on the distribution of environmental contamination [13
]. This methodology could also be useful for occupational risk assessment by allowing risk management to be planned on the basis of a defined acceptable risk and the simulation of prevention measures.
QMRA was derived from the chemical risk assessment paradigm that was set forth by the National Research Council (NRC) in 1983 [14
], consisting of four basic stages: (i
) hazard identification; (ii
) exposure assessment; (iii
) effect assessment (dose-response relationship); and (iv
) risk characterization. In occupational settings, these stages should take into account the worker’s activity and identify the transmission chain, routes of exposure, and matrices involved. It is also essential to choose one index pathogen (or several) representative of the routes of exposure and infection and that is well known in terms of dose-response, resistance, pathogenicity, ease of detection and enumeration in environmental matrices. Although the measurement of pathogens in the air is more difficult than measurements in water or food, biomolecular methods represent an important technical progress for a rapid and easy detection of specific microorganisms in “difficult” matrices [16
One possible innovative approach for risk assessment in occupational settings that are characterized by bioaerosol exposure is monitoring the air for a pathogen, which can be further used for QMRA. The choice of an index pathogen depends heavily on the sources of aerosol, environmental spread, resistance, and routes of exposure and infection. In many cases, the bioaerosol comes from water or dust contaminated by faecal or respiratory secretions, as in wastewater treatment [17
], waste processing [18
] or, more simply, from toilets [7
Human adenoviruses (HAdVs) are the agents for numerous symptomatic and asymptomatic infections affecting the respiratory tract, the eyes, and the gastrointestinal tract. They can be excreted in the feces, urine, and respiratory secretions and transmitted via contact with the eyes, the fecal-oral route, or inhalation. HAdVs have a number of features that justify their use as index pathogens for air in occupational settings possibly contaminated by faecally-excreted pathogens. In fact, HAdVs have been suggested as virological markers for water quality because of their high concentrations in environmental waters and resistance to disinfection [19
]. HAdVs can be pathogenic by either a respiratory or faecal-oral route of exposure, and the relationship between dose and infection in these types of exposures has been studied [13
]. Moreover, methods of culture and biomolecular research and quantification of HAdVs are relatively simple and standardized.
For these reasons, we decided to use HAdVs as the reference pathogen for our study, the objective of which was to develop a preliminary QMRA model to assess the microbial risk associated with the inhalation of contaminated bioaerosols. This model was then applied to data on air contamination from different settings (e.g., wastewater treatment plants, solid waste landfills, toilets in offices and hospitals) and with various exposure times.
Directive 2000/54/EC of the European Parliament and Council of 18 September 2000 [38
] on the protection of workers from risks related to exposure to biological agents at the workplace deals mainly with the risk of infectious agents and gives guidance on health surveillance and containment levels. However, exposure limits are not given for either infectious or non-infectious biological agents, thus implying that simply the potential presence of a pathogen in the air should require the use of respiratory personal protection equipment (RPPE), regardless of exposure time. This practice can be difficult to accomplish in many occupational settings (e.g., outdoor plants for wastewater treatment and landfills) where air contamination differs between areas. Additionally, wearing RPPE can be very uncomfortable. In many cases, workers may not comply with the regulations. The application of QMRA frameworks and models could be useful to give operational guidance for risk management and could lead to a better understanding of health risks from exposure in workplaces.
The use of QMRA to estimate risk associated with drinking water, food, bioaerosols, and fomites was proposed in several studies as a valuable tool to evaluate regulatory standards, bridge information gaps, and assist risk managers in making informed decisions. For example, QMRA has been used [39
] to assess the risk associated with enteric and skin pathogens via exposure to contaminated fomites and evaluate risk reductions that are needed to achieve a safety goal of 1 × 10−6
. The use of such models to predict potentially critical human exposure to Legionella
has been examined in various studies [40
] where, unlike our study, the concentration in air was estimated based on levels in the water. Moreover, in a study that was performed in southern Italy [43
], the QMRA approach was used to assess the risk from a sewage treatment plant for a nearby population. Rotavirus, Campylobacter
, and Cryptosporidium
were selected as index pathogens. Similarly, in another study, QMRA was used to quantify the associated public health risk relative to the distance downwind from the manure application area [44
]. Some of the settings we considered can also be contaminated by microbial metabolites (for instance endotoxins and gliotoxin) that could increase the viral infectivity. This was the case of solid waste processing facilities, where endotoxins were sometimes found to be at high levels [18
]. The synergistic effect of these substances could also be taken into account in a QMRA, if they could be quantifiable, but at the present time these data are unavailable.
In the present study, we applied QMRA methodology to workplace settings in order to determine the risk of infection that is caused by inhalation exposure to HAdV, which was chosen as a reference pathogen because of its wide dispersion, resistance, and infectivity [46
]. Our data confirmed the wide diffusion of this virus and also found considerable differences between indoor and outdoor settings. Our findings reveal an opportunity to reduce indoor contamination with adequate ventilation systems.
The simulations that employed empirical data showed that going to an office toilet for 3 min may be associated with a higher HAdV infection risk compared with working for 15 min at the entrance of a wastewater treatment plant. These results suggest the implausible need to wear RPPE to go to the toilet, more than when working in wastewater treatment plant areas. Such a finding could derive from the index pathogen that was chosen (HAdV). All adenovirus genomic copies were indeed included in the assessment to yield the maximal estimate of risk, although only a sub-portion of the 51 adenovirus serotypes are known to cause respiratory illnesses [47
]. Of the 51 known adenovirus serotypes, only one-third are associated with human disease, and infections by other serotypes are asymptomatic [48
]. The dose-response of adenovirus serotype 4 should not be strictly applied to all adenoviruses. Moreover, the presence of other pathogenic agents (e.g., E. coli
, reovirus, enterovirus, norovirus, hepatitis A virus, and rotavirus) could be associated with a higher risk of infection in settings where aerosols are contaminated by sewage [17
]. Finally, although the objective of the study was limited to assessing the infectious risk due to viral inhalation in occupational settings, we are aware that other contaminated matrices, such as surfaces, can be related to possible exposures. Nevertheless, specific data on dose-response relationships for the exposure to contaminated surfaces are unavailable, owing to the multiple ways of infection related to them: not only skin contact, but also hands, food, and tool contamination. In order to take into consideration multiple pathogens and multiple exposures, the QMRA model should be more complex and should use dose-response data specific for each agent and route of transmission.