A quantitative microbial risk assessment for touchscreen user interfaces using an asymmetric transfer gradient transmission mode.

The ubiquitous use of public touchscreen user interfaces for commercial applications has created a credible risk for fomite-mediated disease transmission. This paper presents results from a stochastic simulation designed to assess this risk. The model incorporates a queueing network to simulate people flow and touchscreen interactions. It also describes an updated model for microbial transmission using an asymmetric gradient transfer assumption that incorporates literature reviewed empirical data concerning touch-transfer efficiency between fingers and surfaces. In addition to natural decay/die-off, pathogens are removed from the system by simulated cleaning / disinfection and personal-touching rates (e.g. face, dermal, hair and clothing). The dose response is implemented with an exponential moving average filter to model the temporal dynamics of exposure. Public touchscreens were shown to pose a considerable infection risk (∼3%) using plausible default simulation parameters. Sensitivity of key model parameters, including the rate of surface disinfection is examined and discussed. A distinctive and important advancement of this simulation was its ability to distinguish between infection risk from a primary contaminated source and that due to the re-deposition of pathogens onto secondary, initially uncontaminated touchscreens from sequential use. The simulator is easily configurable and readily adapted to more general fomite-mediated transmission modelling and may provide a valuable framework for future research.

Introduction

The prevalence of shared Touchscreen User Interfaces (TUIs) has become apparent in recent years; whether it be a fast-food menu or an airport terminal self check-in machine. However, their reputation for hygiene has come under scrutiny, predominantly from sensationalised media articles [1-3]. The fact that touchscreens carry pathogens is not however in question here; what is yet to be established is if they can transmit enough pathogens to a user so as to cause infection (and if so, which disease?). Thus, the aim of this work was to carry out a Quantitative Microbial Risk Assessment (QMRA) using computer simulation methods, in an attempt to answer these questions [4].

Surface contamination

Studies examining the typical microbiome of public TUIs have shown that it is made up of largely innocuous microbes found also on the human hand [5]. However, harmful pathogens such as Enterococcus, E.coli and Staphylococcus aureus have also been found in concentrations typically ∼1 CFU/cm2 [6, 7] but as high as 60 CFU/cm2 [8]. These concentrations may be taken as a baseline contamination level.

Higher levels of contamination can occur when considering droplet deposition from someone suffering from a respiratory illness such as influenza. Viral concentrations in these droplets are on the order of ∼104 PFU/ml [9, 10]. The average volume of a cough being between 0.006—0.044 ml (x 40 for a sneeze) [11-13].

Unwashed hands, particularly after toilet use, is another concerning source of contamination. It has been estimated that 30% of individuals do not wash their hands sufficiently [14-16] and there are additional issues in lavatories with regards to using contaminated soap [17] and doorknobs [18]. Bacterial load found on hands after toilet use ranges from 0.85 ± 0.93(SD) × 105CFU for washed and dried hands to 3.64 ± 4.49(SD) × 105 CFU for unwashed hands [19]. In other experiments, 108 CFU/g has been used to approximate natural bacterial contamination levels [20]. Thus, a TUI with heavy use (and little cleaning) may accumulate a considerable concentration of pathogens. This is exacerbated by the fact that traditional TUI design would have each user touching the same locations on the screen e.g. an OK or Pay Now button, concentrating the effective area of the screen irrespective of the actual physical dimensions.

Transfer coefficients

Transfer coefficients (or rates) are used to describe the proportion of microbes transferred from a fingertip to a surface (and vice versa) after a touch event. There is a wealth of experimental evidence to show that an asymmetry exists between transfer coefficients dependant on which (finger or surface) is the source or destination of initial microbial contamination [21-23]. Moreover, they are dependant on the touch pressure, duration, surface porosity and environmental conditions such as humidity and temperature. Thus, transfer coefficients may be better described through a stochastic process.

More recent experiments have demonstrated that transfer coefficients also depend on the difference between microbial concentration on the finger and the surface being touched [24, 25]. This gradient assumption, in the absence of pathogen die-off or other sources of loss, leads to a dynamic equilibrium being reached after multiple touch events. Essentially, just as many microbes from the finger are deposited onto a surface as are picked up from the surface.

A contribution of this paper is combining both asymmetry and gradient properties of microbe transfer into a single model (Transfer Rates: Asymmetric Gradient Model). This model also gives the ability to simulate the re-deposition of microbes (i.e. cross-contamination of surfaces) and assess the relative infection risks of direct and secondary exposure.

Effectiveness of cleaning & disinfection

Given the uniformly smooth non-porous surface of a TUI one might assume that traditional wiping with a cleaning and disinfectant agents would prove highly effective at controlling surface microbes.

Traditional chemical based sprays and wipes have been shown to be up to 98% effective at removing pathogens [26]. Alternatives to chemicals are UV light [27-29] and the emerging technology of self-cleaning antimicrobial surface coatings, many of which are commercially available e.g. for tablets/smart-phones [30-32].

The effectiveness of all of these methods is reduced when administered inappropriately or by insufficiently trained staff. Also, they may not be suitable for controlling all pathogen types. Any cleaning regimen requires ongoing monitoring and measurement of bioburden to truly ascertain its effectiveness [33]. In a network of people and public TUI use, timing is a key factor that is explored and discussed in Fig 7 and Decontamination.

Dose response

The traditional paradigm of fomite-mediated transmission assumes that a person will at some point touch their face (∼20 times per hour) and transfer pathogens from their fingers to their mucosal membranes (mouth, nasal passages and/or eyes) i.e. self-inoculate [34, 35]. Mucus plays an important role in the immune response and can also be a source for pathogen spread [36, 37]. The period of time over which an individual is inoculated has a significant impact on infection risk. This dynamic dose response is explored in [38, 39]. Essentially, it is assumed that pathogens experience a form of exponential decay once inside a mucus membrane due to natural die-off / inactivation in conjunction with immune system activity. This paper uses a dynamic dose response while accounting for the sporadic influx of pathogens from TUI use and self-inoculation (Dose Response).

Previous work

In a initial attempt at a QMRA for touch screens [40], a complex network of TUI interactions was simulated. It suggested that parameters such as pathogen infectivity and survival time as well as the rate of surface decontamination play an important role in determining risk. One shortcoming of this work was that the model did not include a gradient transfer model and so could not model the effects of re-deposition of pathogens / cross-contamination of surfaces. Building upon this initial work, this paper aims to overcome these deficiencies.

Summary

The remainder of this paper will describe a stochastic simulation model with particular attention to the calculation of microbial transfer (Transfer Rates: Asymmetric Gradient Model & Additional sources of pathogen removal) and dynamic dose response (Dose Response). Simulation results will be presented from a hypothetical scenario involving customers making use of two public TUIs in a commercial retail setting (Results). The risk of infection from TUI use as well as a discussion of the sensitivity of key simulation parameters are presented in Discussion.

Methods

The core of the fomite-mediated transmission paradigm used in this paper is summarised in Fig 1. The key features (highlighted in bold) are described separately in the subsequent sections. For each parameter used in the simulation, default values are listed in Table 1 of the Appendix D in S1 Appendix.

Fig 1

Disease transmission via TUI.

A basic illustration of key model features and chain of events.

Queueing network

An important feature of this QMRA is its ability to simulate the flow of people and their TUI interaction. This is accomplished using a system of first-order Markov Chains and FIFO (first-in-first-out) queue structures at each zone / location. Details of the implementation are found in Fig 2 and Appendix A in S1 Appendix. The reader is directed to [41] for further theoretical background.

Fig 2

Queueing Network and Markov Chains.

The flow of people from one zone to another can be controlled by a first-order Markov Chain (see. Eq A.1 in S1 Appendix) Upon arriving at a location, individuals will queue to use one (of possibly many) identical TUIs (e.g. a bank of ATM machines). After interacting with the TUI, they join a ‘departure’ queue (effectively hanging around at that location) until finally rejoining the Markov Chain and moving on to a new zone/location. A pair of rate parameters associated with a particular TUI type (∼Pois(λ)) and zone (∼Pois(λ)) determine how long they queue in the ‘arrival’ and ‘departure’ queues respectively.

Transfer rates: Asymmetric gradient model

The data concerning transfer coefficient asymmetry and gradient behaviour can be combined and incorporated into a single model (Eqs 1 and 2). Let the initial concentrations (number of microbes / unit area) on a surface or finger be denoted as C and C respectively. For TUI interaction, it is assumed the same effective area of the screen is being contacted with the same finger. Consequently, one can reinterpret C and C simply as being proportional to the number of microbes. Further, one can normalise these values by the total number of microbes such that C+ C ≡ 1. Consider a time interval in which n repeated identical touch events take place; C and C can be updated as follows: where α deposit rate (from finger to surface), β is the deposit rate (from surface to finger) and d and d are the pathogen survival rates (% survival / unit time) on fingers and the surface, respectively. Consider the case of either a clean finger touching a contaminated surface (C = 0)) or a contaminated finger touching a clean surface (C = 0)). From these scenarios it becomes evident that α and β can be estimated from empirical results of transfer asymmetry experiments already mentioned in [21-23].

It can also be shown from Eqs 1 and 2, that in lossless conditions (i.e. d = d = 1), a dynamic equilibrium is reached between finger and surface pathogen levels. The equilibrium point is found by noting: which results in:

By Incorporating pathogen decay (i.e. 0 < d, d < 1), it is evident that all pathogen levels will eventually go to 0 as n → ∞.

This can be seen by simulating repeated touch events between a single person and a touchscreen (Fig 3). Default parameters are used throughout this simulation for α and β for a finger in contact with a non-porous surface (e.g. glass, ceramic tile) [22-24]. Note: when ‘averaging’ the net effect of random transfer coefficients the geometric mean is used; similar to interest rates, transfer coefficients compound in a multiplicative sense.

Fig 3

Dynamic equilibrium of surface bioburden.

A single individual (with clean hands) continuously touches a surface at the same location(s) e.g. the ‘buttons’ on a TUI. The geometric mean and standard deviation (GSD) of surface bioburden is depicted with three variations of the pickup/deposit rate model: fixed parameters (α = 0.05, β = 0.27) with (green) and without (red) loss due to pathogen decay/die-off, and (blue) random parameters α ∼ f(0.05, 0.3, 0, 1) (median average 0.22) and β ∼ f(0.27, 0.3, 0, 1) (median average 0.34) where f(μ, σ, a, b) is a truncated normal distribution on the domain [a, b]. The plots are consistent with Eq 4, which predicts an equilibrium point of 0.15 for the fixed parameter case and ≈0.40 for random parameters. Once pathogen loss is incorporated, all levels tend to 0 over time.

Additional sources of pathogen removal

Loss to environment/clothing

An additional source of pathogen loss in this system comes from an individual touching other surfaces, namely their skin, hair and clothing. This rate of personal-touching can be as high as 200 per hour [42]. In our simulation, all individuals are assumed healthy; so any microbes picked up from personal-touching would correspond only to innocuous human microbiota. For the purposes of a QMRA this can be ignored. Also, in the case of clothing the rate of transfer from fabric back to the finger is very low (<1%) [43], particularly when both surfaces are dry. When combined with the fact that any of these surfaces are unlikely to be touched repeatedly in exactly the same spot personal-touch events can be treated as a ‘sink’ for finger pathogens. Mathematically, one simply ignores Eq 2 and set C ≡ 0 ∀n in Eq 1.

The deposit rate (finger to clothing) (α from Eq 1 that can be denoted as α to distinguish it from TUI related parameters) is a parameter that is explored later in the simulation; default values are listed Table 1 in S1 Appendix. With regards to face touching events, it is assumed to be a random transfer coefficient, ∼f(0.35, 0.1, 0, 0.6) [24, 35].

Cleaning efficiency

Based on chemical wipes, cleaning effectiveness is modelled using a log-normal distribution with modal value of 94% [44]. It should be noted that this is likely an overestimate as typical alcohol based gels may be hindered by other factors such as the presence of mucus (with survival exceeding 10%) [45]. Cross-contamination from improper cleaning methods is not modelled. In the context of this simulation, chemical cleaning and alternatives like UV light are largely comparable in terms of effectiveness. It is more the rate at which cleaning is carried out on TUIs that is a parameter of interest for this simulation. On the other hand, anti-microbial surface coverings have different characteristics that will be discussed in Decontamination.

The human ID50 is defined as the infective dose with 50% probability of infection. For example, respiratory viruses typically require a relatively large dose for infection (103 − 104 PFU) [11, 46, 47]. For many types of bacteria and enteric viruses this can be as low as 10—100 CFU (or PFU) respectively [48, 49]. If one were to interpret these values as estimates for ID50, provided one sticks with the same units, pathogen levels can effectively be normalised and made into a generic type for the purposes of mathematical simulation.

It is customary to model a dose response using an exponential cumulative distribution function (CDF) [12]. The probability of infection, P, at any time, t, will depend on the effective accumulated pathogen load in an individuals mucus membranes, D(t). D(t) will vary continuously as an individual interacts with TUIs and self-inoculates. Let each self-inoculation event at time t result in an instantaneous dose . In order to capture the dynamic characteristics of this exposure, an exponential moving average filter characterised with a time-constant parameter γ is used. where Δt is the time-step (resolution) of the simulation e.g. 1 minute (γ should be defined using the same units). It should be clear that Eq 6 implies an exponential decay (die-off) of pathogen in mucus. γ can be referred to as the inoculation period as it describes a time-interval over which pathogens can contribute to infection risk. An illustration of the effects of different values for γ can be found in Appendix C (Fig 8) in S1 Appendix.

Simulation

For the main simulation the aim is to provide a simple yet plausible scenario that would allow for closer examination of the effects of each governing parameter. Moreover, in order to examine the effects of re-deposition of pathogens i.e. cross-contamination it would seem essential to include at least two TUIs.

One example of sequential TUI use would be airport self-checkin and bag-drop machines. A more generic case (used in this simulation) is in retail outlet centres (shopping malls) where customers may venture from one shop to another making use of a TUI to browse and complete transactions (Fig 4).

Fig 4

Simulated scenario.

Consider two shops (A & B) with customers (n = 32) arriving at each randomly, with equal probability. Some customers (2/3) at A will move on to B afterwards (and vice-versa). The remaining 1/3 will visit only one of the shops (it can be assumed that they then move on to other locations outside the simulation’s focus). The TUI at shop A has already been contaminated with an initial bioburden. There is a particular interest in Alice and Bob; Alice will visit shop A exclusively while Bob will visit shop B exclusively. Therefore, Bob can only be infected from secondary transfer of pathogens from shop A (even though he will never visit A). Alice will receive ‘first-hand’ exposure to the contaminated screen. How does the risk of infection differ for the these two individuals?.

Everyone enters the simualtion from an initial population pool at an average rate ∼Pois(0.25) i.e. once every 4 minutes. Thus, the simulation was expected to take approximately 2 hours. Alice and Bob are programmed to enter the simualtion ∼Pois(0.0167) i.e. once per hour; they arrive as if from the middle of the initial population queue (on average).

Touch rates

As with transfer rates, the number of touches, n, expected to complete a transaction or menu selection can be modelled using a truncated normal distribution. For example, an ATM pin pad requires a minimum of 5 touches (4 pin numbers + OK), i.e. both mode and minimum can be set to 5. However, cancelled transactions, re-attempts, correcting invalid input etc. means that one can expect a small variance in touch numbers. The simulation should ideally account for additional touches associated with (for example) browsing for products. The default simulation parameters f(μ, σ, a, b) → f(12, 3.5, 8, 40) incorporate this random behaviour.

Another important assumption in this simulation is that people will interact with a TUI using just one finger e.g. index finger of the dominant hand. As a consequence, a somewhat crude simplification is to scale down by a factor of 10 all touch rates associated with face/ personal-touching quoted in the literature.

Simulation code

The simulator was implemented in the C programming language. Custom configuration files were used to set and load simulation parameters. Simulation outputs were exported to.csv files, ready for plotting and further analysis. The code is available as open-source at https://gitlab.com/fomite-simulator/fomite_sim_v2. A readme file provides an introduction for compilation and use, along with sample configuration files.

Results

Each simulation result is averaged over N = 40000 realisations in order to achieve a suitable level of confidence in the output (Appendix B in S1 Appendix). The simulation had a time-step (resolution) of 1 minute. In all figures, the light orange curve is associated with the infection risk exclusive to location A (i.e. infection risk to Alice); infection from the primary source of contamination. The purple curve is associated with second-hand exposure at location B (i.e. infection risk to Bob).

Surface cross-contamination

In Fig 5 the (geometric) mean average bioburden on TUI A and B is plotted over time.

Fig 5

Bioburden.

Geometric mean bioburden (log10 scale and normalised) and standard deviation (GSD) on TUIs A and B. A is initially contaminated. Over time microbes are transferred to B (0.1% avg. at 20 minutes) through sequential use. The eventual decay of both curves can be explained by pathogen die-off, personal-touching and the fact that 1/3 of people will visit only one location before leaving the simulation (mitigating further cross-contamination between A and B).

Simulation parameter sensitivity

Presented here are the effects of several key model parameters on infection risk.

Fig 6(a) depicts the effect of initial bioburden on A over a range of ID50. Fig 6(b) and 6(c) show the infection risk associated pathogen decay (half-life) and inoculation period (γ), respectively. Note: pathogen half-life, 11/2, is related to parameter d from Eq 1 by t1/2 = −ln2/lnd.

Fig 6

Parameter sensitivity on infection risks.

Average infection risk (n = 40000 realisations) with 95% CI. (a) Initial bioburden seeded on TUI A (over a range of ID50) (95% CI omitted for clarity). Bioburdens range from baseline (∼101) to ‘heavy’ contamination (105). (b) Pathogen surface half-life (minutes). (c) Inoculation period (time-constant γ of the dynamic dose response model). (d) Average number of touch interactions, n required for TUI use. (e) Number of TUIs available for use at B. (f) Deposit rate associated with personal-touching events α.

Fig 6(d) depicts the infection risk as a function of the average number of touches required for TUI interaction. Fig 6(e) illustrates the effect of increasing the number of available touchscreens for use at B. In order to establish the importance of pathogen loss due to personal-touching events, various deposit rate coefficients are considered Fig 6(f).

It can be seen from Fig 6 that infection rates for Alice is approximately 2.5% to 3% at default parameter settings (see Appendix D in S1 Appendix). The risk to Bob is typically half that of Alice.

Effects of surface cleaning

Fig 7 show the effects of increasing the rate of cleaning/disinfection at each location and the associated risks to A and B.

Fig 7

Cleaning/disinfection rates.

Average Infection risk (n = 40000 realisations) with 95% CI. Solid lines depict infection risk when cleaning rates applied at location B only. Dashed lines are associated with cleaning at A only. Dotted lines correspond to cleaning rates applied both at A and B. Decontamination at B alone has little impact on the risk to Alice (though a slight decrease is observed at higher cleaning rates as pathogens are prevented coming full-circle back to A from B). Decontamination at A alone has slightly better outcomes for Bob than cleaning at B alone. This suggests the risk to customers at one location can be subject to hygiene standards and cleaning regimens at another independent location.

Discussion

Suitability of simulated scenario

The rationale behind the simulation scenario was to provide a risk assessment of TUI use from a generic pathogen type under plausible circumstances. Comparing risk from primary and second-hand exposure was an important and distinctive property of the asymmetric gradient transmission model. The model allowed for estimation of an individual’s finger bioburden over time and subsequent cross-contamination of surfaces. Two individuals (Alice and Bob)were used to represent the personal risk of using a TUI at two locations (A and B) respectively. By having a proportion of the population move on after one visit/interaction a wider network of shops and TUIs was effectively incorporated into the model; the effect was a more rapid removal of pathogens from the overall system.

As a simplifying assumption, the total bioburden in each simulation was fixed. Another option would be to include additional time-dependant sources such as infectious individuals coughing or sneezing in the vicinity of a TUI. It is reasonable to deduce that an introduction of more pathogens would lead to higher infection rates than those presented in the results of this simulation.

Contamination levels

It is clear from Fig 6(a) that in order to reach a significant infection risk to Alice or Bob the ratio of pathogen contamination number (bioburden) to ID50 needs to be large. When considering baseline levels that have been associated with public TUIs (∼1 CFU/cm2), a pathogen would need to be highly infectious to be viable for disease transmission. More significant risk is associated with higher initial contaminations levels; a result of possibly poor hygiene practices from customers and/or cleaning regimens at the establishment.

From Fig 6(c) it appears that the risk associated with an increasing inoculation period (γ) begins to plateau. This may be largely due to the total bioburden being fixed; individuals will reach some expected maximum limit of pathogen numbers in which they can self-inoculate with. When considering a strong innate immune response (associated with ever decreasing γ) infection risk is reduced by ∼30% for both Alice and Bob.

TUI design

As the number of touches required to complete a menu selection or transaction increases one would intuitively expect the infection risk to increase (Fig 6(d)). However, this is not the case. There are several factors to consider: (1) the gradient transmission model ultimately limits bioburdens on fingers and surfaces via dynamic equilibrium (2) pathogens are removed through die-off but also personal-touching and self-inoculation events. The more pathogens picked up off surfaces, the more are removed from the system altogether. This has a significant impact namely because of (3) fixed total initial bioburden on A. Essentially, both Alice and Bob were shielded from further risk because everyone else was carrying pathogens away with them.

Having more TUIs to choose from at B meant a de-concentration of surface bioburdens and a reduced infection risk for Bob (Fig 6(e)). A rather perplexing exception is the slight rise in risk for 2 to 4 TUIs. One explanation could be related to the flow of people: both Alice and Bob were programmed to arrive at their destination such that they were (on average) in the middle of the population queue of 32 people (based on 1 TUI per location). By altering the number of TUIs the throughput at B is changed; Bob arrives sooner (possibly when pathogen deposits are ‘fresher’). After a certain threshold, the de-concentration effect takes over. This highlights the complexities that may arise within queued networks.

An interesting approach to hygienic TUI design suggests to shift the layout of the interface as it appears on screen, thus removing the constraint of forcing individuals to touch the exact same area [50]. This would effectively create multiple virtual TUIs to choose from as in Fig 6(e) (without actually increasing throughput).

Personal-touching

In Fig 6(f) the average deposit rate coefficient associated with personal-touch events was altered. Though not dramatically so, higher rates do lead to reduced risk as more pathogens are transferred to clothing etc. and removed from the system. The importance of personal-touch events in modelling fomite-mediated transmission is apparent but remains a complex behaviour to completely replicate.

Decontamination

Throughout the simulation it was assumed that individuals did not deliberately wipe/wash their hands between TUI. Such habits would, of course, drastically reduce infection risk associated with fomites. However, enforcing stringent hand-washing policies is not an attractive option for retail businesses and may not be feasible in all circumstances.

An alternative approach that places the responsibility and control with the TUI owners/operators is enhanced sterilization regimens. From Fig 7 it is apparent that cleaning can have a significant impact on infection risk. However, if the goal is to virtually eliminate the risk (≪ 1%), disinfection would have to take place every few minutes to keep up with constant cross-contamination from other sources.

An interesting result in Fig 7 is from comparing Bob’s infection risk when cleaning is carried out at A exclusively and B exclusively. The infection risk is lower when the original source of contamination (i.e. A) is cleaned. This suggests the risk to customers at one location can be subject to hygiene standards and cleaning regimens at another independent location. Ultimately, the infection risk to Alice was consistently higher than that of Bob suggesting that second-hand exposure can only succeed that of the primary source of contamination.

Anti-microbial coatings are designed to reduce pathogen survival rates on surfaces. For this reason, Fig 6(b) best illustrates their possible benefits. It is clear that a well-designed coating can have significant impact on infection risk; driving it virtually to 0 in the extreme case. However, further developments still need to be made to meet this performance level across a sufficiently wide range of pathogens types [33, 51].

Touchscreens can be designed or retrofitted to employ automated cleaning cycles using UV light. An automated cycle after each use would greatly reduce surface bioburdens and infection risk. However, there would be an associated added cost for the equipment and increased energy consumption. Dependant on power output, the UV cycle may take several seconds to operate which could adversely affect user throughput. Additionally, the effectiveness of UV light is hindered by dirt and other fluids; therefore TUIs would still require a traditional cleaning. UV itself may also poses potential health risks (though an automated system would least reduce exposure risks to staff employed to clean) [52]. Yet another risk posed by decontamination methods, including UV, is the inadvertent selection for microbial resistance [53].

An alternative to decontamination altogether is to re-develop or retrofit existing TUIs with hand tracking cameras to provide a touch-less interface. This nullifies surface microbial transfer and eliminates infection risk (Fig 6(d)). Whether or not businesses and venues choose to implement such alternatives will likely depend on the cost of cleaning, replacement or conversion of existing TUIs.

Risk assessment

In these simulations, infection occurred (for Alice) as much as ∼3% of the time. It is worth discussing whether or not this poses a significant risk. Firstly, infection risk in an epidemiological context really refers to the likelihood or probability of infection. Risk, in a heath and safety context, is often defined as a product of the likelihood and the severity of an adverse event. While this simulation was capable of computing the former, there was no function implemented to assess the cost to health, long or short term, nor any subsequent economic losses due to disease spread.

For comparison, another study to asses the role of fomite transmission in a simulated norovirus outbreak predicted a likelihood of infection ranging widely from 5% to 57% [54]. Similar values were also found in an assessment of fomite transmission in a real world SARS outbreak [55]. However, in the case of SARS, other transmission routes e.g. airborne played a dominant role.

It seems possible then that TUIs may play a small contributing role in larger outbreaks. The reason why they do not gain more attention or blame in scenarios where touch is the dominant form of transmission may have something to do with the severity and timing of illness. Consider, in this simulation, if Alice or Bob were to eventually develop mild to moderate gastroenteritis the following evening. Would they attribute it to a single interaction with a TUI? Would their illness eventually be reported and counted by the local health authority? Without this data, it is easy to see why epidemiologists still strive to understand the nuances of fomite mediated transmission.

Pathogen survival (Fig 6(b)) and infectiousness (Fig 6(a)) are key factors in determining which pathogens are likely to be responsible for TUI disease transmission. It is worth noting that bacteria have the capacity, theoretically, to survive and even grow on surfaces given the right conditions. Viruses, on the other hand, can only diminish. It is interesting to note in Fig 6(b) that infection rates level-off when pathogen half-life exceeds ∼100 minutes; after which pathogen removal is dominated by personal-touching (Fig 5). Therefore, when considering pathogen survival it should be brought into the context of TUIs; it is not necessary for a pathogen to survive long, merely long enough given the rate of TUI use.

The default survival rates used in this simulation where derived from available laboratory reports of a select group of pathogens on glass/ceramic surfaces. Data from actual TUIs could reveal different results and could, for example, take into account any effects from the light / heat emanating from the screen. Moreover, techniques for quantifying bacteria and viral particles is prone to wide margins of error. Therefore, the simulation is only as reliable as the data put into it.

Model limitations

An important limitation of the model was to assume all interaction took place with a single dominant finger. A further consequent simplification was that the area on the TUI was effectively homogenized. This allowed for the transfer coefficients to be modelled as simple individual random variables (see Appendix D in S1 Appendix).

In order to model more complex interactions using multiple fingers, it would be necessary to consider not only an extended set of transfer coefficients but also partition the TUI surface into regions where each finger is likely to touch. There could also be cross-contamination between an individual’s fingers. Essentially, Eqs 1 and 2 would need to be implemented for each finger and for each possible touch interaction. This would add a significant level of complexity to the model but may be useful in assessing risk associated with a specific TUI design/layout. The advantage of the current simplified model is that it offers generalised insights into how TUI and networks of people interact to spread disease.

Simulator framework

The implementation of the simulator was developed in C which offers good performance for computationally intensive operations (see Code availability). The simulator can be operated using a (relatively) easy-to-use configuration file system, making it straightforward for a researcher to simulate complex queueing networks and TUI types without re-compiling code. It is hoped that this may provide a valuable framework for future research endeavours in fomite-mediated disease transmission modelling.

(PDF)

Click here for additional data file.

26 Nov 2021

A quantitative microbial risk assessment for touchscreen user interfaces using an asymmetric transfer gradient transmission mode

PLOS ONE

Dear Dr. DiBattista,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please see below the comments and suggested MAJOR revisions made by the individual(s) who reviewed your manuscript.  If provided, the referee's report(s) indicate the revisions that need to be made before it can be accepted for publication.

Please submit your revised manuscript by Jan 10 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Ricardo Santos

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

3. Thank you for stating the following financial disclosure:

“ADB is an employee of Ultraleap Ltd. As sponsors, Ultraleap played no role in the study design, data collection and analysis or preparation of the manuscript. The intention to publish this work and to provide open-source data/code, was arranged with Ultraleap prior to the start of this project. Ultraleap has funded the publication costs of this paper.”

We note that one or more of the authors is affiliated with the funding organization, indicating the funder may have had some role in the design, data collection, analysis or preparation of your manuscript for publication; in other words, the funder played an indirect role through the participation of the co-authors. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please do the following:

a. Review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. These amendments should be made in the online form.

b. Confirm in your cover letter that you agree with the following statement, and we will change the online submission form on your behalf:

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”"

4. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“The author would like to acknowledge the support of Ultraleap Ltd. and Orestis 329

Georgiou for the realisation of this research project.”

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“ADB is an employee of Ultraleap Ltd. As sponsors, Ultraleap played no role in the study design, data collection and analysis or preparation of the manuscript. The intention to publish this work and to provide open-source data/code, was arranged with Ultraleap prior to the start of this project. Ultraleap has funded the publication costs of this paper.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

5. Thank you for stating the following in the Competing Interests section:

“I have read the journal's policy and the authors of this manuscript have the following competing interests:

This was a collaborative study funded by Ultraleap Ltd. Ultraleap is a manufacturer and developer of human-computer interface technology. ADB is a paid employee of Ultraleap. Ultraleap's role was limited to funding these costs as well as any fees associated with publication.”

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: ""This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: N/A

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The paper PONE-D-21-23569 entitled " A quantitative microbial risk assessment for touchscreen user interfaces using an asymmetric transfer gradient transmission mode" discusses an important issue which is the quantitative microbial risk assessment. In particular, the author wished to know if touchscreens can transmit enough pathogens to a user to cause an infection.

My major concerns regarding the study are the lack of validation of the results obtained. Although it would be difficult to determine the number of infections, it should be at least possible to determine if the results presented in Fig. 6 are close to reality or how they compare with other studies with infection risk in the order of 1-2.5%. Models are important to simulate scenarios but it should be possible to assess how close to reality are the results of each scenario.

One thing that is not clear to the reader is if the risk determined is high or not. How can 1-2.5% infection risk be considered? How can this be compared to other situations?

The author also discusses the use of disinfectants and UV light to decrease the risk of transmission. Could this favour the risk of selecting more resistant strains?

Specific comments are listed below:

- The Methods section should allow the repetition of the study by other colleagues. Information about the software and language used to make the simulations/model should be given here. The author added information about the simulator framework only at the end of the paper, but more information is required in the Methods section.

- The discussion section also contains results that should be presented in the results section (e.g. contamination levels, dose response, etc). A general discussion about what the impact of the different assumptions and choices made is lacking. The author presents the discussion in different sub-sections and states the assumptions made but does not provide what would be the result if another choice was made. With the computational model, the impact of some assumptions could be tested to provide support to the decisions made.

- A general conclusion to the questions presented as aims of the study is not provided.

- English should be checked throughout the text. Examples of phrases that should be improved follow:

+ “A more concerning source of contamination if from a user(s) with unwashed hands, particularly after toilet use.”

+ “All of these methods suffer when administered inappropriately or by insufficiently

trained staff.” (Methods do not suffer..:)

+ “Each simulation results is averaged (…)”

Furthermore, the author should not use the word “we” in the text. Usually scientific studies are written without referring to who did the work. In this particular case, only one author is listed and the use of "we" should be avoided.

- “Another important assumption in this simulation is that people will interact with a TUI using just one finger e.g. index finger of the dominant hand.” – This may be true for older people but younger people use more fingers in touchscreens. In the discussion, decisions such as this should be discussed and their impact assessed.

Reviewer #2: Specific comments:

1. The authors conducted an excellent study and provided an appropriate amount of literature review on the topic.

2. The parameters tested in the QMRA i.e., queueing network, transfer rates-asymmetric gradient model, dynamic equilibrium of surface bioburden and the additional sources of pathogen removal provides a thorough approach validating the simulation.

3. Introducing the coding language C in the methods section and taking a step-by-step approach of explaining how the simulation was performed would be helpful for readers to utilize the simulation model. Excellent idea to provide the link to the code used in the study and the appendix information.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Reviewer comments-PONE-D-21-23569.pdf

Click here for additional data file.

7 Jan 2022

Please see Response to Reviewers document submitted.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

4 Mar 2022

A quantitative microbial risk assessment for touchscreen user interfaces using an asymmetric transfer gradient transmission mode

PONE-D-21-23569R1

Dear Dr. DiBattista,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Ricardo Santos

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The author of manuscript PONE-D-21-23569R1 improved significantly the manuscript following the reviewers’ comments. He answered satisfactorily all questions I raised. In particular, he added a link to the simulation code and added sections discussing risk assessment and model limitations, which were may main concerns I had when reviewing the original submission.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

16 Mar 2022

PONE-D-21-23569R1

A quantitative microbial risk assessment for touchscreen user interfaces using an asymmetric transfer gradient transmission mode

Dear Dr. Di Battista:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Ricardo Santos

Academic Editor

PLOS ONE