Eye infection risks from Pseudomonas aeruginosa via hand soap and eye drops

ABSTRACT

Eye infections from bacterial contamination of bulk-refillable liquid soap dispensers and artificial tear eye drops continue to occur, resulting in adverse health outcomes that include impaired vision or eye enucleation. Pseudomonas aeruginosa (P. aeruginosa), a common cause of eye infections, can grow in eye drop containers and refillable soap dispensers to high numbers. To assess the risk of eye infection, a quantitative microbial risk assessment for P. aeruginosa was conducted to predict the probability of an eye infection for two potential exposure scenarios: (i) individuals using bacteria-contaminated eye drops and (ii) contact lens wearers washing their hands with bacteria-contaminated liquid soap prior to placing the lens. The median risk of an eye infection using contaminated eye drops and hand soap for both single and multiple exposure events (per day) ranged from 10^–1 to 10⁻⁴, with contaminated eye drops having the greater risk. The concentration of P. aeruginosa was identified as the parameter contributing the greatest variance on eye infection risk; therefore, the prevalence and level of bacterial contamination of the product would have the greatest influence on health risk. Using eye drops in a single-use container or with preservatives can mitigate bacterial growth, and using non-refillable soap dispensers is recommended to reduce contamination of hand soap. Given the opportunistic nature of P. aeruginosa and its ability to thrive in unique environments, additional safeguards to mitigate bacterial growth and exposure are warranted.

IMPORTANCE

Pseudomonas aeruginosa (P. aeruginosa) is a pathogen that can persist in a variety of unusual environments and continues to pose a significant risk for public health. This quantitative microbial risk assessment (QMRA) estimates the potential human health risks, specifically for eye infections, associated with exposure to P. aeruginosa in bacteria-contaminated artificial tear eye drops and hand soap. This study applies the risk assessment framework of QMRA to evaluate eye infection risks through both consumer products. The study examines the prevalence of this pathogen in eye drops and soap, as well as the critical need to implement measures that will mitigate bacterial exposure (e.g., single-use soap dispensers and eye drops with preservatives). Additionally, limitations and challenges are discussed, including the need to incorporate data regarding consumer practices, which may improve exposure assessments and health risk estimates.

INTRODUCTION

Pseudomonas aeruginosa (P. aeruginosa) is an opportunistic pathogen that can ubiquitously occur in the environment, posing a significant threat to public health (1). This Gram-negative bacteria has been observed to grow in bulk-refillable soap dispensers and in ophthalmic solutions (e.g., dry eye, artificial tears), posing a risk of infection to users, particularly immunocompromised individuals (2, 3). In addition, P. aeruginosa infections can be acquired in medical settings (such as through water, blood, and disinfectants) and has been identified as a significant nosocomial agent, causing an estimated 32,600 infections among hospitalized patients in the United States in 2017 (4–6). Globally, an estimated 10%–15% of nosocomial infections are due to P. aeruginosa (7). Consequently, refillable soap dispensers are not recommended for healthcare facilities (8, 9), yet they are still widely used for public facilities, such as restaurants, schools, stores, and public restrooms.

Detectable concentrations of bacteria have been identified in both bulk-refillable liquid soap dispensers and artificial eye drops (10–13), and can include many types of coliform and other opportunistic pathogens (10). Yapicioglu et al. (14) documented an outbreak of P. aeruginosa in an intensive care unit linked to liquid soap with the presence of 3 to 7 $\times$ 10⁶ CFU/mL. Pinna et al. (2) identified a concentration of >10⁶ CFU/mL of P. aeruginosa in a preservative-free ophthalmic solution 24 h after the solution was inoculated with 100 CFU/mL of the bacterium. Additionally, when P. aeruginosa is introduced into the eye cornea, either by rubbing or touching the eyes with contaminated fingers, by penetrating the eye through a contaminated foreign object such as contact lenses, or by introducing contaminated medications such as eye drops, it acts with extreme virulence causing blindness in many cases and even necessitating enucleation (15).

Quantitative microbial risk assessment (QMRA) is an approach used to estimate the probability of infection and illness from exposure to pathogenic microorganisms in the environment (16, 17). Several government agencies have used the QMRA framework to establish guidelines and standards for microbial exposure to drinking water and food (18–20). This framework, which comprises hazard identification, exposure assessment, dose-response, and risk characterization (16, 17), can be applied to estimate the probability of an eye infection associated with ocular exposure to P. aeruginosa.

P. aeruginosa has been identified as one of the most common causes of eye infections among contact lens wearers (21, 22), has been isolated from bulk-refillable liquid soap dispensers (10), and has been the cause of microbial keratitis in 81 individuals across 12 states in the United States who used contaminated artificial tear eye drops (23). Ocular exposure may occur through contaminated lenses (microorganisms present on the finger are transferred when placing the lenses in the individual’s eye) or through artificial tear eye drops placed directly in the eye. This study utilized a dose-response model from previously published data to characterize the health consequences associated with P. aeruginosa exposure to the eye. The dose-response output informed a QMRA to estimate the probability of an eye infection through ocular exposure to P. aeruginosa.

RESULTS AND DISCUSSION

The beta-Poisson distribution model has been found to best fit the dose-response data for ocular exposure of P. aeruginosa (24). Utilizing this dose-response model, the risk of an eye infection from P. aeruginosa-contaminated eye drops and hand soap was estimated for the two exposure scenarios.

Eye infection risks of P. aeruginosa in both eye drops and hand soap

The median risk of an eye infection from a single exposure event was three orders of magnitude greater for the eye drop exposure (5.16 × 10⁻¹) than for an infection associated with a finger washed with hand soap touching an eye (2.52 × 10⁻⁴) (Fig. 1). Daily infection risks were described as applying eye drops four times over the duration of 1 day to a single eye and two occasions of using contaminated hand soap to touch the eye when inserting a contact lens (e.g., morning and night). The estimated median daily infection risk for eye drops in one eye (9.45 × 10⁻¹) was three orders of magnitude greater than the daily median infection risk for the hand soap exposure (5.04 × 10⁻⁴). Both exposure scenarios exceeded the daily modified infection risk threshold (1 infection per 1,000,000 individuals or 1 × 10⁻⁶) (18, 25) by at least two orders of magnitude.

Fig 1.

Single and daily infection risks associated with exposure to eye drops and contaminated hand soap. The modified daily risk of infection threshold (not shown) is 1 x 10⁻⁶ (or 1 infection per 1,000,000 individuals). Note that for the eye drop exposure pathway, the 50th and 75th percentiles are too narrow to display a visible box at this scale; therefore, the box appears condensed.

P. aeruginosa-associated eye infections are not uncommon due to the occurrence of this pathogen in the environment and the different pathways by which eye exposure may occur (4). Infection risk estimates in this assessment range from 2.5:10,000 associated with hand soap to 5:10 for eye drop exposure. As traditional for QMRAs, both scenarios assume the presence of P. aeruginosa in the contaminated product, with this assessment estimating a greater probability of infection associated with the application of contaminated eye drops. Although the likelihood of the occurrence of P. aeruginosa in hand soap and eye drops may originate during manufacturing, the influence of consumer behaviors on the chance or magnitude of exposure should not be underestimated. Specifically, behaviors related to hand washing, contact lens wearing, and applying eye drops can all contribute to the likelihood of infection.

P. aeruginosa is capable of maintaining adherence to contact lens surfaces by forming biofilms (4, 22, 26, 27), thus exacerbating host exposure if lenses are extended wear and/or are not removed for proper disinfection according to manufacturer recommendations. In addition, extended-wear contact lenses create a hypoxic environment that fosters the secretion of norepinephrine, a hormone that has been shown to enhance the pathogenesis of P. aeruginosa in mouse studies (28). Moreover, eye fluids have diminished antimicrobial properties following contact lens wear (29), thereby adding to the increased likelihood of infection. Contact lens solution may be of poor quality, so even if used as instructed, lenses may not be disinfected appropriately (30). Finally, the health consequences from exposure to P. aeruginosa and resulting infection can be severe due to the antibiotic-resistant properties of this pathogen (2, 31, 32).

Diseases transmitted by direct physical contact are best prevented by effective hand washing (33). However, using contaminated hand soap from refillable dispensers can result in Gram-negative bacteria colonizing the hands in public settings, thus increasing bacteria count on the skin (10, 34–36). It has been observed that the main entry pathway for bacterial biofilms in liquid soap pump dispensers is the pump head (9). The biofilms can enter the pump dispensers immediately through the pressure release when the pump is activated, as a result of liquid buildup in the pump head and ensuing biofilm growth. P. aeruginosa biofilms are capable of persisting and accumulating in liquid soap solution, and persisting cells have the ability to return to their growth stage and increase exposure risks (9, 37). As previously described, P. aeruginosa has antibiotic-resistant properties and is a significant contributor to nosocomial infections, including bacteremia, skin ulcers, and urinary tract infections (3, 10, 38–41).

Sensitivity analysis

A sensitivity analysis was conducted in Oracle Crystal Ball (Austin, TX), which used rank correlation values to quantify the variance associated with the probability of infection due to each parameter (42) (Fig. S1 to S4; Supplementary Materials). Probabilistic values were incorporated to address variability in the risk assessment, and the concentration of P. aeruginosa (${\displaystyle C_{\mathrm{S}\mathrm{o}\mathrm{a}\mathrm{p}}}$ and ${\displaystyle C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$) was identified to be the parameter that contributed the greatest variability on the overall risk estimate. For the eye drop exposure scenario, the volume of an eye drop (${\displaystyle V_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$) contributed significantly less variability, while the transfer ratio of bacteria from liquid soap to the hand (${\displaystyle T}$) was identified to have a similar contribution to variability as that of the bacteria concentration on the hand soap scenario. Identifying which parameters contribute the greatest variability in the risk assessment can be targeted for public health mitigation.

P. aeruginosa concentration in contaminated product and infection risks

After identifying that the concentration of P. aeruginosa in both eye drops and hand soap contributed the greatest variability on the risk of infection, further analysis regarding infection risks associated with varying bacteria concentrations were evaluated. Specifically, point estimate values for all parameters were considered for this component of the study to evaluate the magnitude that bacteria concentration had on health risks given different exposure scenarios (Table S1; Supplementary Materials). Given the multiple stages from washing the hands with soap to touching an eye, a significantly smaller risk for an eye infection is associated with a larger pathogen concentration in hand soap (Fig. 2). In contrast, eye drops are applied directly to the eye; therefore, a smaller concentration of the P. aeruginosa yields a much greater infection risk (Fig. 3). For example, if the eye drop solution is contaminated with a bacteria concentration of 10³ CFU/mL, the estimated health risk is 1.72 × 10⁻² or an estimated two infections per 100 individuals. Alternatively, if hand soap is contaminated with 10³ CFU/mL of bacteria, the health risk (5.25 × 10⁻⁸) is six orders of magnitude lower, with an estimated five infections per 100,000,000 individuals.

Fig 2.

Probability of eye infection with different concentrations of P. aeruginosa on the hand (from soap) resulting from one touch of the eye.

Fig 3.

Probability of eye infection with different concentrations of P. aeruginosa in one individual contaminated eye drop.

The risk of eye infection is dependent on multiple variables, including (but not limited to) the concentration of the bacterium in the eye drops and soap, volume of individual drops, size of the hand, finger, and fingertip, the process of the contact lens being removed and reinserted, and the number of times the event occurs. This analysis assumes that a single eye drop is placed directly into the eye (with no blinking or interference on the volume of the drop reaching the eye) and that the contact lens is placed in the eye after the hands are washed. However, it is likely that some risk always exists when the eyes are rubbed after the hands are washed. Additionally, since four eye drops are recommended per eye over the course of a day and that two contact lenses are usually worn simultaneously, the risk per exposure scenario is likely greater.

The 2023 outbreak in the United States of the carbapenem-resistant P. aeruginosa strain (VIM-GES-CRPA), found in EzriCare artificial tears, resulted in severe eye infections affecting at least 81 individuals and causing four deaths (43–46). This specific strain had not been previously identified in the United States and was found to be resistant to all antibiotics previously used to treat bacterial keratitis (44). In addition, a previous study examining bacterial contamination of liquid soaps in public restrooms found that about 25% are contaminated with bacteria in concentrations above 500 CFU/mL (10). The average concentration of bacteria in these dispensers was 3.9 $\times$ 10⁶ bacteria per mL and ranged from 590 to 1.3 $\times$ 10⁷ bacteria per mL. P. aeruginosa was found to be the predominant bacteria in seven out of 30 samples in which identification was conducted (10). Thus, exposure to high numbers of bacteria from these types of dispensers is not uncommon. Additionally, P. aeruginosa was found in hand lotion traced to an outbreak in a neonatal intensive care unit (47).

Limitations and challenges

As with any QMRA, assumptions must be made where data are lacking to estimate health risks. Uncertainty from these assumptions includes inferring dose response from rabbits to human eye infections via contact lenses. Studies have shown that dose response for bacteria addressed for humans and animals are comparable (48, 49). However, quantifying the amount of liquid realistically contacting the eye when applying drops, along with the concentration of P. aeruginosa in the exposure, contributes both uncertainty and variability to this assessment. Data on how often contact lenses are removed and replaced within different age groups during a day and after different activities (sporting activity) or afflictions (such as due to allergies and respiratory infections) would inform a more specific risk model. The type of contact lens could also influence the potential risk for infection, as different types of contact lenses have different incidences of P. aeruginosa-associated infections (21). Finally, the size of the hand, finger, and fingertip will have some influence as it varies with age, gender, and ethnicity (50).

Conclusions

This QMRA study examined the eye infection risks associated with exposure to consumer products that have been reported to be contaminated with the opportunistic pathogen, P. aeruginosa. The estimated health risks ranged several orders of magnitude but indicated that both eye drop and contact lens users are at increased risk of eye infections from this pathogen. Single-use eye drop bottles or eye drops with preservatives can mitigate bacterial growth and exposure. The use of non-refillable sealed soap dispensers is recommended to reduce the risk of microbial contamination of soap, given the sanitary design of the dispensing systems (34). While the study describes the potential for adverse health risks associated with these products, further research should aim to characterize the chemical composition and prevalence of contamination of these products, as well as other consumer items that may pose a risk for public health. Future work should examine specific behaviors that may increase the potential for bacterial contamination of the product or transmission of the pathogen to the eyes and develop mitigation strategies that can be included in product design or practiced by the consumer. The pervasive nature of P. aeruginosa in the environment and in both eye drops and hand soap emphasizes the need to address this emerging public health concern.

MATERIALS AND METHODS

Dose-response model selection

To assess the risk of an eye infection from exposure to P. aeruginosa, a previously developed dose-response model using data reported in a study that examined keratitis in rabbits was utilized (24). In the referenced study, 30 white New Zealand rabbits wore perfilcon A (71% H₂O ionic polymers) extended-wear soft contact lenses for 7 days with complete lid closure to mimic contact lens over use. After 7 days, conjunctival cultures showed no signs of infection. The lenses were removed and then incubated with various concentrations of a single-strain P. aeruginosa (10⁷, 10⁶, 10⁵, 10⁴, and 10² CFU/mL and a saline control; n = 5 per group). They were then replaced on their respective corneas for an additional 48 h. By day 9, corneal thickness had increased significantly, and P. aeruginosa keratitis had developed in 13 of 25 exposed eyes (and in none of the five control eyes). The experiment was repeated twice, and the data pooled to inform the development of a dose-response model. Additional information regarding the study design and methods utilized are reported in the published literature (24).

Most microbial dose-response data fit an exponential or beta-Poisson model (17). The QMRA Wiki repository suggests that the beta-Poisson dose-response model provides the best fit for the dose-response data that corresponds to a reported health outcome of a corneal ulceration (24, 51). The best fit parameters for the model include α $=0.19$ and $N_{50}=\mathrm{18,500}$. Figure 4 displays the plot of observed vs. expected probabilities of response using the beta-Poisson model.

Fig 4.

Beta-Poisson dose-response curve to estimate the probability of a P. aeruginosa eye infection.

Through the QMRA framework, this dose-response model was used to estimate the probability of an eye infection resulting from exposure to varying doses of P. aeruginosa. Specifically, the probabilities of infection for two exposure scenarios—(i) using contaminated artificial tear eye drops and (ii) washing hands using contaminated soap from a refillable liquid soap dispenser—were calculated using equation 1.

$${\displaystyle P\left(\mathrm{I}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\right)=1-{\left(1+\mathrm{D}\mathrm{o}\mathrm{s}\mathrm{e}{\displaystyle \frac{\left(2^{\left(\frac{1}{0.19}\right)}-1\right)}{18,500}}\right)}^{-\left(0.19\right)}}$$ (1)

Estimating the risk of developing a P. aeruginosa eye infection

The dose (D) of viable P. aeruginosa that comes into contact with the eye for both exposure scenarios was determined from equations 2 and 3. For equation 2, the estimated dose is the product of the total concentration of P. aeruginosa in the contaminated eye drop solution (${\displaystyle C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$) and the volume of solution applied per drop (${\displaystyle V_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$).

$${\displaystyle D=C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}\times V_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$$ (2)

In which,

$D$: dose or total concentration of viable P. aeruginosa that comes in contact with the eye (CFU);

${\displaystyle C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$ : total concentration of P. aeruginosa measured in the eye drop solution (CFU/mL);

V _Drop: volume of a single eye drop administered to the eye (mL)

Equation 2 describes the potential dose of P. aeruginosa that would be directly applied to the eye. The values for ${\displaystyle C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$ were retrieved from a study assessing P. aeruginosa growth in the preservative-free ophthalmic solution, Refresh Plus (which is used to treat dry-eye symptoms). Samples of the solution were inoculated with three different isolates of the opportunistic pathogen for a target concentration of ~100 CFU/mL (2). Samples were enumerated at 6, 12, and 24 h to determine the growth of the bacteria population. After 6 h of incubation, sample concentrations (for each isolate) ranged from 2,000 to 6,480 CFU/mL. By 24 h, all samples exceeded 1,000,000 CFU/mL (2). For this risk assessment, a uniform distribution with a minimum value of 2,000 CFU/mL and a maximum value of 1,000,000 CFU/mL was assumed for the ${\displaystyle C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}}$ parameter. A study assessing eye drop volumes discussed that while volumes of 0.005 to 0.015 mL are ideal for biopharmaceutical and economic needs, eye drop volumes administered to the eye typically range from 0.025 to 0.070 mL (V _Drop) (52). Therefore, the parameter, V _Drop, was assumed to follow a uniform distribution with a minimum value of 0.025 mL and a maximum value of 0.070 mL.

Equation 3 describes the estimated dose a contact lens user would be exposed to when placing a contact lens in their eye after washing hands with contaminated hand soap. The dose is the product of the ratio of the concentration of P. aeruginosa on one hand from hand soap (${\displaystyle C_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}$) and the average area of the palm side of one hand (${\displaystyle A_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}$), the fraction of one hand that is one finger (${\displaystyle F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}}}$), the fraction of one finger that is the fingertip (${\displaystyle F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{p}}}$), the hand-to-eye transfer efficiency ($f_{\mathrm{2,3}}$), and the average exposed surface area of the eye (${\displaystyle A_{\mathrm{E}\mathrm{y}\mathrm{e}}}$).

$${\displaystyle D={\displaystyle \frac{C_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}{A_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}}\times F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}}\times F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{p}}\times f_{2,3}\times A_{\mathrm{E}\mathrm{y}\mathrm{e}}}$$ (3)

In which,

D: dose or total concentration of viable P. aeruginosa that comes in contact with the eye (CFU);

${\displaystyle C_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}$ : total concentration of P. aeruginosa left on one hand after washing with liquid soap from refillable soap dispensers, (CFU);

A_Hand: average area of palm side of one hand (cm²);

${\displaystyle F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}}}$ : the fraction of one hand that is the finger (decimal, dimensionless);

${\displaystyle F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{p}}}$ : the fraction of one finger that is the fingertip (decimal, dimensionless);

f_2,3: hand-to-eye transfer efficiency of viable P. aeruginosa (decimal, dimensionless);

A_Eye: average exposed surface area of the eye (cm²)

The concentration of P. aeruginosa remaining on the hand after washing both hands with contaminated refillable liquid soap was determined from previous work (34). Bulk-refillable liquid soap dispensers were inoculated with $3.24\times {10}^7$ CFU/mL of the Gram-negative bacteria, Serratia marcescens, in the first study and with $3.24\times {10}^4$ CFU/mL of the same Gram-negative bacteria in the second study. Subjects in both studies were instructed to use 1.5 mL of soap, wash both hands for 10 s, and rinse with water for 10 s. Post-wash bacteria recoveries were enumerated and averaged $1.91\times {10}^5$ CFU/hand for the first study and averaged $5.01\times {10}^1$ CFU/hand for the second study. Gram-negative bacterial transfer ratio (T) was defined as the concentration of bacteria recovered from the palm of one hand to the concentration of bacteria inoculated into the liquid soap, resulting in T = 0.00589 for the first study and T = 0.00155 for the second study (34). In this present study, T was assumed to follow a triangular distribution, with minimum, likeliest (average), and maximum values of 0.00155, 0.00372, and 0.00589, respectively. The parameter, ${\displaystyle C_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}$, as shown in equation 4, considers the bacterial concentration in the liquid soap is CFU/0.375 mL or CFU/event (${\displaystyle C_{\mathrm{S}\mathrm{o}\mathrm{a}\mathrm{p}}}$), with an event defined as using 1.5 mL of contaminated liquid soap to wash both hands (four sides of two hands) for 10 s, and rinsing them with water for 10 s (resulting in 0.375 mL of contaminated liquid soap per one side of one hand).

$${\displaystyle C_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}=T\times C_{\mathrm{S}\mathrm{o}\mathrm{a}\mathrm{p}}}$$ (4)

The area of one side of one hand was calculated by dividing 830 cm² by 4 resulting in 207.5 cm² (53). Previously reported guidance suggests that the palmar surface area of one finger is estimated to be 10% of the total palmar surface area, while one fingertip represents 30% of the total finger surface area (54). The hand-to-eye transfer efficiency (f_2,3) assumed is 0.339, as determined by previous work by Rusin et al. (55). The average exposed surface area of the eye (A_Eye) was estimated to be 2.0 cm² (56). All input parameters used to estimate the risk of an eye infection are defined in Table 1.

TABLE 1.

Input parameters for the QMRA model to estimate the risk of an eye infection

Parameter	Value	Unit	Reference
Ta	0.00155, 0.00372, 0.00589	Unitless	(34)
C Soap b , c	3–7 × 106	CFU/mL	(14)
C Drop b	2 × 103–1 × 106	CFU/mL	(2)
A Hand	207.5	cm2	(53)
F Finger	0.30	Unitless	(54)
F Fingertip	0.10	Unitless	(54)
f 2,3	0.339	Unitless	(55)
A Eye	2.0	cm2	(56)
V Drop b	0.025–0.070	mL	(52)
N	4 (eye drops) 2 (hand soap)	Unitless	(57) Assumed

^a The parameter ${\displaystyle T}$ follows an assumed triangular distribution [minimum, likeliest (average), maximum].

^b The parameter follows an assumed uniform distribution (minimum, maximum).

^c The bacterial count in hand soap per event is calculated as ${\displaystyle C_{\mathrm{S}\mathrm{o}\mathrm{a}\mathrm{p}}}$ * 0.375 mL; all other variables are point estimates.

Therefore, to calculate the probability of infection for both exposure scenarios—contaminated eye drops (equation 5) and contaminated hand soap (equation 6)—the dose equations were substituted in equation 1, resulting in equations 5 and 6.

$${\displaystyle P(\mathrm{I}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n})=1-{\left(1+\left(C_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}\times V_{\mathrm{D}\mathrm{r}\mathrm{o}\mathrm{p}}\right){\displaystyle \frac{\left(2^{\left(\frac{1}{0.19}\right)}-1\right)}{18,500}}\right)}^{-(0.19)}}$$ (5)

$${\displaystyle P(\mathrm{I}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n})=1-{\left(1+\left({\displaystyle \frac{T\times C_{\mathrm{S}\mathrm{o}\mathrm{a}\mathrm{p}}}{A_{\mathrm{H}\mathrm{a}\mathrm{n}\mathrm{d}}}}\times F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}}\times F_{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{p}}\times f_{2,3}\times A_{\mathrm{E}\mathrm{y}\mathrm{e}}\right)\frac{\left(2^{\left(\frac{1}{0.19}\right)}-1\right)}{18,500}\right)}^{-(0.19)}}$$ (6)

The probability of infection that can be calculated using equations 2 through 6 estimates the risk of an eye infection from a one-time event of applying eye drops (e.g., artificial tears) and washing hands (n = 1). The probability of infection associated with multiple exposure events (N) was calculated using equation 7 (Table 1). Daily application of eye drops, up to four times a day, is recommended to improve symptoms of dry eye disease (57). A handwashing event to place/remove contacts from eyes likely occurs two times per day.

$${\displaystyle P(\mathrm{I}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{e}\mathrm{t}\mathrm{o}N\mathrm{E}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{s})=1-{\left\{1-P(\mathrm{I}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{e}\mathrm{t}\mathrm{o}1\mathrm{E}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t})\right\}}^N}$$ (7)

Risk estimates were assessed using both point and probabilistic exposure and dose-response values to evaluate variability and sensitivity of specific parameters on risk outputs (Table 1). Monte Carlo simulations (10,000 simulations per exposure pathway) were conducted using Crystal Ball Software (Oracle Corp., Austin, TX, USA). Daily risks of infection were compared to the modified suggested risk threshold for waterborne pathogens of one infection per 1,000,000 persons (18, 25). The QMRA framework presented in this study only estimates the individual risk of exposure in a static model and therefore does not consider secondary transmission or immunity (58).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.02119-23.

Figures S1 to S4, Table S1. aem.02119-23-s0001.docx.

Supplemental figures and table.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.