Human Health Risks posed by High Tech Fomites

Problem Statement

GCI is interested in quantification of the risk posed to their customers and employees at their product stores. As the potential for widespread transmission of the pathogen is relatively large – especially during the winter season – the risk might be rather serious.

Therefore, we have calculated this risk for the Influenza virus, which is considered to be one of the most widespread pathogens. With the results we have defined a financially viable and non-intrusive risk management strategy to reduce the infection risk. The stakeholders comprise GCI, its employees, customers, and close relations of the employees and customers (via secondary transmission of the pathogens). As GCI is a worldwide operating company, the results will have global influence.

Hazard Identification

Fomites refer to objects that are likely to contain large reservoirs of viruses and bacteria. Examples of fomites are mobile phones, computer keyboards, and door handles. Hartmann et al. (2003) have performed microbial analyses from samples in hospital rooms, yielding 8.9 % of the samples to contain up to 26 potentially pathogenic microorganisms. Ulger et al. (2009) have performed a sampling of the hands and the mobile phones of 200 healthcare workers. Of these 200 phones, 94.5 % was contaminated with different types of bacteria, and methicillin-resistant Staphylococcus aureus (MRSA) was isolated in 13 % of the cases.

One of the most widespread diseases of relevance is the Influenza virus, which infects the respiratory system. Influenza is mainly transmitted via three primary ways:

1. Airborne transmission of the pathogen after a cough or sneeze.

2. Deposition of the pathogen on the product from the air or from droplets after a cough or sneeze.

3. Deposition of the pathogen on the products from contaminated hands of an infected person.

The Influenza virus infects the respiratory system, which may lead to serious symptoms and even possible mortality among especially elderly people and people with a chronic disease. The regular symptoms comprise headache, fever, coughing, sore throat, sore muscles, a cold, a feeling of weakness, and loss of appetite (CDC^[1]). Secondary complications comprise among others pneumonia, bronchitis and secondary bacterial infections (Haas, 1999). Besides, the annual rate of Influenza related death in the U.S. was 1.4 – 16.7 deaths out of every 100,000 persons (CDC^[2]).

The annual averaged U.S. prevalence is estimated to be 2.72 % (CDC^[3]), and the annual costs attributed to this virus run up to 1 billion dollars (Haas 1999). Combined with the fact that the Influenza virus can remain viable and infectious up to 24 hours on hard and smooth surfaces (Kramer, 2006), and up to 4 – 6 hours on porous surfaces (Bean 1982; Haas, 1999), the risk of infection of every customer is expected to be not negligible.

Exposure Assessment

For the risk assessment, we have simulated the Influenza contamination in one computer store. In all scenarios, it is assumed that one person – either a customer or an employee – is infected by the Influenza virus. This person may infect other (healthy) people by initially aerosolising the virus after a cough. The aerosolised pathogens may be inhaled by the healthy people, or they may deposit on the surface of the products. The calculations were performed with a dose-response model. The calculated risk is extrapolated to all GCI stores in the U.S. in order to calculate an overall risk of illness or mortality.

Exposure Pathway

We have defined six different scenarios for contamination by one infected person (Figure 1). In all scenarios, this person infects one other person. Scenarios 1 – 3 comprise an infected customer (customer A), who will remain in the store for 45 minutes (Apple^[4]) and who will cough 6x during this time period (Loudon, 1997). In scenario 1, this customer infects customer B, as this person inhales the contaminated air. In scenario 2, customer C becomes infected after having touched one of the contaminated products. In scenario 3, customer D becomes infected after both having inhaled the contaminated air and after having touched one of the contaminated products. Scenarios 4 – 6 comprise an infected employee, who will be in the store for 3 hours before he is actually leaving. During this period this employee coughs 22x (Loudon, 1997). For the rest, these scenarios are similar to scenarios 1 – 3. Table 1 shows a list of all the values that were used and all other assumptions that were made for the model.

Figure 1. Schematic overview of the risk calculation scenarios. The parenthesised numbers indicate the scenario number.

In addition, the following assumptions are defined:

Table 1. Model values and assumptions

Exposure Model

Our developed model considers various states of exposures, which all terminate in various ways. States are any physical location where the pathogens are at any given point. The individual rates of removal of the Influenza virus from their respective states are denoted as λ I. (see figure 2). State 1 represents the air in which the (infected) person emits the pathogens. This air is circulating the store and it can terminate either by inhalation of the healthy person (state 2) or in removal out of the building (state 9). Additionally, the pathogen can deposit on surfaces: glass (state 3), plastic (state 4), aluminium (state 5), or non-fomite surfaces (state 6). Non-fomite surfaces are considered to be a terminal endpoint. Glass, plastic and aluminium are typical high-tech materials, which are often touched by the customers. Glass is found on the smartphones and tablet surfaces. Plastic is found on the keyboards of the notebooks and aluminium is the fomite that is in the area around the keyboard. The pathogens in states 3, 4, and 5 can either continue on to the hands of the customers or employees (state 7), or can terminate as decay on the surface itself (State 10). From the hands at state 7, the exposure terminates in state 8, on the face, where the pathogen is ingested.

Figure 2. Model state exposures; all states are connected via different transfer rates, which are a measure for the number of particles transferring from one box to the other.

Loss Rates

The loss rates λ_i were derived by considering how the pathogen was removed in various states. Therefore, λ_i varies as the transition between states is also varied. The first loss rate to be considered is λ_1,2, where pathogen is in state 1 and is inhaled by a healthy person:

\begin{align*} \lambda_{1,2} = Q_{resp}/V \end{align*}

where Q is the volumetric flow rate of the respiration [m3.s-1], and V is the volume of the store [m3]. The loss rate from State 1 in the air to any of the glass, plastic or aluminium fomites is:

\begin{align*} \lambda_{(1,[3,4,5])} = v_{ts}/H*A_F/(A_S-A_F) \end{align*}

where v_ts is the terminal settling velocity [m.s-1], H is the height [m], and AF and AS are the surface area of the store and fomite respectively [m2]. Furthermore, the loss rate from state 1 to any non-fomite surface is:

\begin{align*} \lambda_{1,6} = v_{ts}/H*(A_{room}-(A_G + A_P +A_A))/A_S \end{align*}

with AG, AP, and AA the surface areas of the glass, plastic and aluminium parts respectively [m2]. The loss rate from state 1 to state 9 (exhaust) is described below, where Q is the volumetric flow rate of the air in the store [m3.s-1], and V is the volume of the store [m3]:

\begin{align*} \lambda_{1,9} = Q/V \end{align*}

Markov Chain Matrix

From this model sketch, we can move on to the next step, which is to find the probabilities that the Influenza virus will stay in state i during a time step Δt:

\begin{align*} p_{ii}=exp(-\lambda_i*\Delta t) \end{align*}

with p_ii the probability of remaining in state i. The probabilities of the virus moving from state i to state j is then given by:

\begin{align*} p_{ij}=\lambda_{ij}/\lambda_i * [1-p_{ii}] \end{align*}

The probabilities were then used to build a Markov matrix, which we created as a 10x10 matrix that represents the ten states of the model. A Markov chain is a system that models the transitions from one state to another through a chain-like process, in which the next state depends only on the current state and not the entire past.

We assumed that the pathways are undirectional, meaning that once the pathogen has continued into another state, it does not backtrack to the previous one. Therefore, p₃₁ will be 0, p₃₃ is the probability that the pathogen is in state 3 and will remain in state 3, and p_3n is the probability that the pathogen is in state 3 and will continue to state n (thus, n might be 7 or 10).

Monte Carlo Simulation

After the completion of the Markov Chain, we used Crystal Ball® to create stochastic models to estimate the risk from either only breathing, only touching the fomites, and a combination of breathing and touching the fomites for both the 45 minute and 3 hour exposure scenarios. We utilized the data generated by the Markov Chain, which was correlated with the final cough experienced by the person exposed. We only used the data from the final cough since it represented a cumulative exposure over the entire 45 minute or 3 hour time of exposure. The equation that accounts for the dose received by a person breathing contaminated air is:

DoseBreathing=[#viral particles/VAir]x [IR] x [time]

This formula includes information regarding the number of viral particles in the room, which is set up as a distribution, the volume of air in the room, the inhalation rate for a person participating in light activity, as well as the specific amount of time the individual was exposed. For specific values please refer to the assumption table located in Chapter 3.1. The equation that accounts for the dose received by a person touching each individual fomite is:

dose= 〖viral particles〗_fomite*A_finger/A_fomite *λ_t*time*η_hand*η_fomite*contact rate

This includes information regarding the number of viral particles on each fomite (plastic, glass, or aluminium), the surface area of a person’s finger, the total surface area of each fomite, the frequency of times a person touches the fomite, the transfer rates for both fomite to hand, and hand to eye, as well as the frequency of times a person comes in contact with their eyes. Distributions were set for the number of viral particles on the fomite, time in contact with the fomite, as well as the contact rate. 

Table 2 shows the doses for customers and employees for inhalation and touching the surfaces of the fomites. As indicated in the table, the largest doses are obtained by inhaling the Influenza virus.