Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2023 Jan 25;3(2):457–464. doi: 10.1021/acsestwater.2c00521

Virus Emissions from Toilet Flushing: Comparing Urine-Diverting to Mix Flush Toilets

Lucinda Li †, Jinyi Cai †, Joseph N S Eisenberg ‡, Heather E Goetsch †, Nancy G Love †, Krista R Wigginton †,*
PMCID: PMC9927869  PMID: 36818380

Abstract

graphic file with name ew2c00521_0006.jpg

High levels of viruses can be found in human excrement from infected individuals, a fraction of which can be emitted from toilet flushing. Unlike the common mix flush toilet (MFT), the urine-diverting toilet (UDT) separates urine from the toilet water. Specific focus on urine-associated viruses is needed because the UDT can emit different levels of urine-associated and fecal-borne viruses and urine has different properties compared to feces that can affect emission levels (e.g., protein content). In this work, we quantified emission levels of surrogate bacteriophages for urine-associated and fecal-borne viruses, MS2 and T3, from flushing a UDT and an MFT, with and without protein in the water. Emission levels of viruses in the water of the UDT were lower than that of the MFT by up to 1.2-log10 and 1.3-log10 for T3 and MS2, respectively. If urine is completely diverted in the UDT, virus emissions can be reduced by up to 4-log10. Based on these results and typical levels in urine and feces, we estimate that up to 107 and 108 gene copies of human viruses per flush can be released from the UDT and MFT, respectively. Lower emissions observed with the UDT suggest reduced exposure to viruses from flushing the UDT.

Keywords: toilet flushing, virus emissions, exposure, urine, protein

Short abstract

This work quantifies virus emissions from flushing for a urine-diverting and a mix flush toilet, with and without protein in the water, and two viruses to yield emission data that can be used to quantify the risk of infection from toilet flushing.

Introduction

Human viruses can be excreted at high levels in the vomit, feces, and urine of infected individuals. In feces, up to 1012 genome copies (gc) of viruses per wet gram can be excreted during norovirus (HuNoV), adenovirus (HAdV), and rotavirus infections.1−3 Many viruses are excreted in urine, including West Nile, Nipah, Rabies, Rubella, and Smallpox viruses.2,4 HAdV and human polyomaviruses (HPyV) JCPyV and BKPyV can be excreted at particularly high levels in urine, namely, 1010 gc mL–1.3,5,6 While virus genome copies in human excreta can be high, the fraction of these genomes that are part of infectious virions is often not known.

Flushing a toilet generates droplets and aerosols. When the toilet contains excreta with viruses, flushing creates a possible route of exposure for individuals in the restroom. This is particularly relevant for HuNoV, HAdV, and HPyV as they can be excreted at high levels and are linked to a range of illnesses including gastroenteritis, cold-like symptoms, and kidney deterioration, respectively.7−9 Viral infection risks from environmental exposure are commonly characterized using the quantitative microbial risk assessment (QMRA) framework. A limited number of QMRAs have been conducted on virus emissions from toilet flushing. They suggest that this exposure route can create a significant risk of virus infection.10−12 For context, a study on the risks of HAdV infection from inhaling contaminated bioaerosols in different settings estimated that the exposure to aerosols from toilet flushing is higher than that of a faulty drain in a sewer pipe and from working at a wastewater treatment plant and a municipal solid waste landfill.10

Mitigating virus exposure and risk of infection from toilet flushing requires an understanding of the amounts and drivers of virus emissions from toilets.13 Viruses emitted from toilets are present in large droplets that settle onto surfaces or in small droplets that evaporate to become droplet nuclei.14 The latter remains in the air for hours and travel with the air plume. Most studies have recovered infectious viruses from the air15 and on surrounding surfaces16 after flushing. Viruses on surfaces and in air may remain infective for hours to months.17,18 For example, dried hepatitis A virus and human rotavirus on surfaces were infective for more than 2 months.17 Gerba et al. detected infectious viruses in bioaerosols for up to 6 h after toilet flushing, some of which settled over a 4 h period and contaminated nearby surfaces.18 While these studies have focused on the detection of viruses in either bioaerosols or droplets, Gerba et al. captured the total droplets that reach the toilet seat.18 This approach can provide a more comprehensive understanding of the total virus emissions from flushing.

Toilet types can affect the amounts of viruses emitted. Previous studies have focused primarily on the common mix flush toilet (MFT), which flushes excrement from one compartment.15,16,18,19 In these toilets, higher flush pressures are associated with higher virus emissions.19,20 Alternative toilet technologies such as low and dual flush, composting, dry, and incinerating toilets can address the United Nations Sustainable Development Goal related to clean water and sanitation. Urine-diverting toilets (UDTs), for example, separate urine from the rest of the waste stream at the toilet so that nutrients in urine can be recovered and processed into fertilizers.21 In UDTs, urine can be collected with or without a flush, the latter likely leads to an overall decrease in virus emissions.

Studies on virus emissions from toilets have primarily focused on fecal-borne viruses15,16,18,19 even though urine can contain high levels of human viruses. Urination accounts for an estimated 6 of the 7 daily toilet flushes per person;22,23 the higher frequency of urine flushes could result in higher exposure to urine-associated viruses. Virus release may be impacted by the toilet contents, such as the higher amount of protein in toilet water containing urine and feces, which can affect adsorption of viruses to surfaces24 and the formation of droplets.25 Likewise, virus properties, such as size or isoelectric point (IEP), may also affect the release of viruses because these characteristics impact sorption.19,26 Characterizing the influence of toilet water contents and virus characteristics on virus emissions is necessary to conduct informed risk assessments of toilet flushing.

This study aims to compare the total amount of virus emissions from flushing a UDT and an MFT. We used a method to capture viruses emitted from the toilet water during flushing and compared the emission levels from a commercially available UDT and MFT installed in an institutional restroom. Experiments were conducted with two surrogate viruses; bacteriophages MS2 and T3 were selected as surrogates for human viruses common in urine and feces based on their similar physicochemical properties such as size, IEP, and absense of envelope (Table 1). We also tested the effect of protein in the toilet water on virus emissions, using a protein concentration that can be found in urine. The measured emission levels of the surrogate viruses were extrapolated to estimate the emissions of human viruses HuNoV, HAdV, and HPyV when the maximum reported viral loads are excreted in feces and urine and flushed from the MFT, flushed from the UDT, and diverted from the UDT.

Table 1. Surrogate and Human Virus Properties.

virus genome type genome size (kb/kbp) size (nm) isoelectric point (IEP) max fecal concentration (gc g–1) max urine concentration (gc mL–1) references
T3 dsDNA 38 50 2.0–5.0 N/A N/A (28, 33)
human polyomavirus (HPyV) dsDNA 5 44 N/A N/A 1010 (6, 29)
adenovirus (HAdV) dsDNA 26–45 90 4.5 1011 1010 (3, 30, 33, 34)
MS2 ssRNA 3.6 27 2.2–3.9 N/A N/A (31, 33)
norovirus (HuNoV) ssRNA 7.7 27–38 5.5–6.0 1012 N/A (1, 32, 33)
Open in a new tab

Methods

Toilet Information

Two toilets on a university campus were used in this study to represent an MFT and a UDT (Figure 1). The MFT was a Kohler model 4330-0 (Kohler, Wisconsin, USA) installed in 2016 with a small flush volume of 4.8 L and a large flush volume of 6.1 L. The UDT was a Wostman Ecoflush (Wostman, Saltsjö-boo, Sweden) installed in 2016 with a small flush volume of 0.3 L and a large flush volume of 2.5 L. Only the large flushes were used in this study.

Figure 1.

Figure 1

Open in a new tab

Photos of the MFT (A) and UDT (B).

Virus Surrogates

Bacteriophages MS2 and T3 were used as surrogates for human ssRNA viruses (e.g., HuNoV) and human dsDNA viruses (e.g., HPyV and HAdV), respectively (Table 1). MS2 is a non-enveloped, ssRNA bacteriophage that is commonly used to represent enteric viruses of similar size and genome type such as HuNoV. T3 is a non-enveloped, dsDNA bacteriophage that we used to represent human dsDNA viruses such as HPyV and HAdV. MS2 (ATCC 15597 - B1) and T3 (ATCC 11303 - B3) were propagated in theirEscherichia coli hosts (ATCC 15597 and 11303). Following chloroform extraction and polyethylene glycol precipitation,27 the viruses were further concentrated by 100 kDa ultrafiltration (MilliporeSigma UFC901024) and filter sterilized with polyethersulfone 0.22 μm filters (Celltreat 229747). The concentrated virus stocks (1011 plaque forming unit mL–1) were stored at 4 °C until use.

Virus Solution

Based on the reported mean 24 h urine and feces excretion quantities and median daily urination and defecation frequencies,22 we estimated that 130–710 mL of urine and 100–300 g of feces are excreted into a toilet per event. Up to 1011 gc of HAdV34 and 1012 gc of HuNoV1 have been measured in 1 g of feces and 1010 gc of HPyV6 and HAdV3 in 1 mL of urine. Based on typical virus concentrations and excrement amounts, we simulated a virus loading event by adding 1010 plaque forming unit (pfu) of the surrogate viruses into the toilet. Specifically, 10 mL of the virus solution, containing 109 pfu mL–1 of MS2 and T3 each, in phosphate-buffered saline (PBS, Gibco 10010023) or PBS with 1% (83.3 mg L–1) bovine serum albumin (BSA, Dot Scientific DSA30075-25), was added to the toilet. To test the effects of the protein content on virus emissions, experiments were conducted with and without BSA in the toilet water. Based on previous reports of the median urination and defecation frequency, urine and feces protein content, and urine and feces excretion quantities,22 we estimated that protein levels of up to 20.8 mg L–1 and 11.25 g L–1 are present in the toilet following urination and defecation events, respectively.

Toilet Experiments

Toilet bowl surfaces were sanitized with a 70% ethanol solution and flushed before experiments. A 10 mg L–1 sodium thiosulfate solution (Fisher Scientific S474-500) was added to quench residual chlorine in the water. Total chlorine and free chlorine were measured using a Hach meter DR 900 and DPD pillows (Hach 2105669) to ensure that both were below detection limits before experiments. Experiments were performed in ambient conditions in the restroom.

In the UDT, most of the urine is diverted to the front of the toilet, but we performed our experiments to quantify virus emissions for urine and feces that are deposited into the toilet water. After adding the virus solution, the toilet water was mixed for 1 min with a sterile serological pipette. Control experiments were performed with 10 mL of water added to a second MFT that was not used in the virus experiments. Prior to flushing, a 1 mL aliquot of toilet water was collected to quantify the initial virus concentration in the toilet water. A polyethylene film (Office Depot 32007-OD) was placed over the toilet bowl area, and the toilet was flushed (Figure 2). After 1 min, the film was removed, and a sterile cotton gauze pad (Dukal 2283) that had been soaked in a PBS solution containing 1% BSA was wiped over the film to recover viruses. The gauze pad was then placed in 10 mL of 1% BSA in PBS, and the solution was vortexed at a maximum speed for 1 min. Recovery experiments suggested that 88.8% of T3 and 130% of MS2 were recovered with this approach (Supporting Information). Infectious viruses were quantified using a double overlay agar plaque assay with a limit of quantification (LOQ) of 20–250 pfu mL–1 and a limit of detection (LOD) of 10 pfu mL–1. Plaque assay negative and positive controls were conducted with each experiment to rule out contamination and problems with the assays. The fraction of viruses emitted was calculated as the total number of viruses, in pfu, recovered divided by the total number of viruses, in pfu, added to the toilet. When the amount recovered from the toilet was below the LOQ, the fraction of viruses emitted was calculated using the LOD of 10 pfu mL–1 in the numerator.

Figure 2.

Figure 2

Open in a new tab

Graphical representation of the experiment procedure: a recovery surface was placed over the toilet bowl area, and the toilet was flushed. Droplets and aerosols emitted on the recovery surface were captured with a soaked cotton gauze pad and suspended into solution by vortexing.

Particle Size Distribution of the Virus Emissions

We used toilet emission particle size distributions reported in a previous study35 and our measured fraction of viruses emitted to estimate the virus emissions in different particle size ranges. Knowlton et al. measured droplets near a hospital toilet before and 1 min after 10 flushes with fecal waste. They categorized droplets into six bin sizes, including 0.3, 0.5, 1, 3, 5, and 10 μm.35 We used WebPlotDigitizer to extract the average particle concentration in each bin presented in their publication and used the resulting data to calculate the percentage of the total emission volume emitted in each size bin. Under the assumption that the number of viruses in a particle is directly proportional to the volume of the particle, we calculated the fraction of viruses emitted for each bin size.

Data Analysis

Plaque assay data were log-transformed and analyzed using GraphPad Prism. A Shapiro–Wilk test was used to validate the lognormality of the data, and multiple unpaired, parametric student t-tests with Welch correction were performed to assess statistical significance (p < 0.05).

Results and Discussion

Emissions from the UDT Were Significantly Lower Than Those from the MFT

Of the 1010 pfu of MS2 and T3 added to the MFT water, an average of 25 × 102 pfu of MS2 and 4.3 × 102 of T3 were emitted when the toilet water was not supplemented with protein. In terms of the fractions of the total viruses added to the toilet, these values are equivalent to 9.6 × 10–7 for MS2 and 13 × 10–7 for T3. Compared to previous studies in which viruses were flushed with an MFT and then measured either in air or on surfaces, our results are slightly higher. For example, 2.4 × 103 pfu m–3 of MS2 was measured in the air after an MFT was flushed containing 1010 pfu.15 This is equivalent to a 2.4 × 10–7 fraction of the added viruses. Sassi et al. measured 1.9 × 104 pfu of MS2 on the surrounding floor and 3.4 × 105 pfu on the surface of the toilet seat after flushing a toilet, which are equivalent to 0.19 × 10–7 and 3.4 × 10–7 fractions of the total added viruses (1012 pfu of MS2), respectively.16 Compared to a study in which all droplets at the toilet seat level were measured, our results were similar—Gerba et al. measured an average of 8.6 × 102 pfu of poliovirus, which is equivalent to a 30 × 10–7 fraction of the added viruses.18 The higher fractions measured in these studies may be reflective of the total emission capture method as compared to separate air and surface sampling.

Fewer viruses were emitted when the same number of viruses added to the toilet water was flushed with the UDT compared to the MFT. Specifically, the mean fractions of viruses emitted from the UDT were 1.6 × 10–8 for MS2 and 2.3 × 10–8 for T3 (Figure 3). On average, the fraction of viruses emitted from the MFT was greater than that of the UDT by 1.2-log10 (p = 0.02) and 1.3-log10 (p = 0.02) for T3 and MS2, respectively, when not amended with protein. The MFT emitted more viruses than the UDT by as much as 2.0-log10 for T3 and 2.3-log10 for MS2 (Figure 4). The same trends between the toilets were observed when protein was added to the toilet water (Figures 3 and 4). It is worth noting that in several of the UDT experiments, T3 or MS2 levels recovered following the flush were below the detection limit, whereas T3 and MS2 were always above the LOD following flushes with the MFT (Figure 3). These results demonstrate that the toilet type affects the number of viruses emitted from flushing a toilet.

Figure 3.

Figure 3

Open in a new tab

Virus emissions from toilet flushing. The log fraction emitted is the average fraction of viruses that was captured on the recovery surface (pfu), normalized by the number of viruses added into the toilet water (pfu), and then log10 transformed. Error bars indicate the standard deviation for each set of experiments (N = 6). The letter “a” above the bars indicates statistically significant differences at p < 0.05 using unpaired, parametric student t-tests with Welch correction. A down arrow is used to represent mean bars that include data below the LOD.

Figure 4.

Figure 4

Open in a new tab

Comparing the log difference of fraction emitted for all conditions. Log difference above zero indicates that the fraction emitted was greater for the UDT. Log difference below zero indicates that the fraction emitted was greater for the MFT. For all experimental conditions, the MFT emitted more viruses. Error bars indicate the standard deviation for each set of experiments (N = 6).

The differences in the virus emissions between the two toilet types can be driven by the different flush pressures as well as the different toilet water volumes. The flush volume of the MFT is larger than that of the UDT by 1.3 L, which may generate more water droplets. Flush pressure is a toilet characteristic that has been studied more extensively in previous research.19,20,36 Like many residential toilets, the UDT has a water tank attached to the toilet bowl that provides water pressure to flush the toilet. The MFT is a commercial toilet that utilizes a flushometer to draw water pressure from the water supply line. Typically, toilets with flushometers have a higher flush pressure than water tank toilets.20 In a previous study, Lai et al. found that flushometer toilets resulted in higher bacterial emission levels than those from a water tank toilet. Additionally, they found statistically significant greater emission levels in a flushometer toilet with a 400 kPa flush than with a 200 kPa flush.19 A similar observation was made in water tank toilets.36 Combined, our current results with viruses and previous studies with bacteria suggest that the type of toilet and specifically the flush pressure are important for the emissions of viruses from toilet flushing.

Protein in the Toilet Resulted in Slightly Greater Emissions

We tested the effect of protein added to the toilet water along with the viruses, as urine and feces result in elevated toilet water protein levels. The experiments with protein added to the toilet water consistently yielded greater virus emissions than the experiments without added protein (Figure 3). In the MFT, T3 and MS2 emissions in the experiments with added protein were greater than those in the experiments without added protein by 0.17-log10 (p = 0.70) and 0.21-log10 (p = 0.55), respectively. These differences were more pronounced with the UDT toilet, with T3 and MS2 emissions in the protein experiments greater by 0.49-log10 (p = 0.09) and 0.93-log10 (p = 0.05), respectively.

Protein content can affect virus adsorption to surfaces and aerosolization of the toilet contents. For example, viruses adsorb to toilet surfaces,18 and the extent of sorption can be affected by the protein content in urine and feces.24 In addition to competing for adsorption sites, protein in the toilet water can affect how droplets are formed during a flush.25 Namely, the presence of protein can reduce the surface tension of droplets,37 subsequently reducing their size.38 It remains unclear, however, whether the total volume of all emissions is affected by the presence of protein. In our experiments, the addition of BSA generally resulted in higher emissions from the toilet, but the differences were not statistically significant.

Protein in the Toilet Resulted in Differences between the Two Toilets

The difference between viruses emitted from the MFT and the UDT was greater when the toilet water was not supplemented with protein. Specifically, for MS2, the difference in the fraction emitted between the toilets was 0.53-log10 with protein present and 1.3-log10 without protein present. For T3, the difference was 0.83-log10 with protein present and 1.2-log10 without protein present (Figure 4). While all experiments in the MFT and all experiments with added protein were above the LOQ, five and two out of six replicates in the UDT without added protein were below the LOD for T3 and MS2, respectively (Table S1). One possible explanation for our observed impacts from protein and toilet type is that there are compounding effects of protein, toilet flush pressure, flush volume, and virus type. Lai et al. previously correlated variables such as pathogen size and flush pressure to pathogen emission levels,19 although they did not consider the protein content or viruses. Future work should explore correlations between the protein content and additional matrix properties (e.g., the presence of feces) with virus emissions from toilet flushing.

MS2 and T3 Were Emitted at Similar Levels

In all conditions except the UDT flushes amended with protein, T3 was emitted at slightly greater fractions than MS2. The differences in fractions emitted between MS2 and T3 were between 0.02-log10 and 0.28-log10 across all conditions (0.34 < p < 0.95). The similarity between MS2 and T3 emissions suggests that the differences in MS2 and T3 properties, namely, their size and IEPs, did not affect their emissions in a considerable manner. T3 is approximately 1.9-fold larger than MS2, and the two viruses have similar IEPs (∼3.5 for MS2 and 2–5 for viruses similar to T3).33 Other studies have suggested that smaller pathogens have greater emissions. For example, Lai et al. observed that the smallerStaphylococcus epidermidis had emission levels 21 times greater than those of E. coli and found a statistically significant correlation between the bacterial size and the amount of bacteria emitted from toilet flushing.19 Likewise, MS2 viruses were emitted at higher levels than E. coli, and this difference was attributed to the different sizes of the organisms.18 The fact that MS2 and T3 are more similar in size than S. epidermidis vs E. coli and E. coli vs MS2 may explain why we did not observe more emissions for the smaller virus. MS2 and T3 are similar in size and have similar IEPs to many human viruses that are excreted in urine and feces (e.g., HuNoV: 5.5–6, HAdV: 4.5, and HPyV: N/A); our results suggest that the emission behaviors of these viruses may be similar to the surrogate viruses measured here. More research is necessary to understand the role that the virus size and IEP have in emissions. Likewise, the impact of lipid envelopes on some viruses should be studied to understand the emission behaviors of viruses like coronaviruses, Influenza virus, and Ebola virus. These types of virus characteristics could affect how viruses partition to the toilet bowl or fecal matter present in the toilet water.

Estimated Human Virus Emissions Can Exceed Infectious Levels

We applied the fractions of the surrogate viruses emitted in our experiments, adjusted by the recovery experiments (Supporting Information) (calculated with pfu), and the levels of human viruses in feces and urine reported in literature (as gc) to estimate the range of human viruses that could be emitted from those deposited in the toilet water (Table 2). We first conducted control experiments to confirm that the ratios of gc to pfu in the virus solutions added into the toilet were not significantly different than those of the recovered samples from the toilet flush (Tables S2 and S3).

Table 2. Estimated Virus Emissions from Each Toilet for HAdV, HPyV, and HuNoV Using the Minimum and Maximum Fractions Emitted from Experiments with Surrogate Virusesa.

  MFT
 UDT
  minimum maximum minimum maximum
HPyV and HAdV in urine (gc/event) 4.6 × 104 1.7 × 107 1.8 × 104* 5.9 × 105
HAdV in feces (gc/event) 2.1 × 105 7.5 × 107 8.1 × 104* 2.7 × 106
HuNoV in feces (gc/event) 3.9 × 106 3.9 × 108 2.8 × 105* 6.7 × 107
Open in a new tab
a

Data with * indicate that the minimum fraction emitted that was used in the calculation was below the LOD.

In feces, HuNoV and HAdV can be excreted into and emitted from toilets at a range of levels. HuNoV levels in human feces, for example, have been reported in the range of 104–1012 gc per wet gram.1,39 An infected individual excretes an average of 128 g of feces per day and defecates, on average, 1.2 times per day, averaging 107 g of feces per event.22 The estimated viral loading into a toilet per event for HuNoV is therefore 106–1014 gc. Based on the emitted fractions of MS2, the surrogate virus we used for HuNoV, we estimated that HuNoV can be emitted at up to 39 × 107 and 6.7 × 107 gc per flushing event from the MFT and UDT, respectively. Similarly, HAdV is excreted at a reported maximum of 1011 gc per gram of feces.34 Based on our results for T3, the surrogate virus we used for HAdV, we estimated that HAdV can be emitted at up to 7.5 × 107 gc per event from the MFT and 0.27 × 107 gc from the UDT (Table 2).

Virus emissions from flushing after urination events can be as high as those from defecation events. We used the maximum concentrations reported for HAdV and HPyV in urine (1010 gc mL–1)3,5,6 and the average volume per urination event (237 mL calculated from the reported daily volume and frequency)22 to estimate the amounts of these viruses that are present in the toilet following urination (1012 gc). Based on the emitted fractions for T3 measured in our experiments, we estimated that HAdV and HPyV can be emitted at up to 166 × 105 gc per urination event in the MFT and 6 × 105 gc in the UDT (Table 2). While we report the maximum emissions for HAdV and HPyV from the UDT here, the toilet bowl of the UDT is designed to physically separate urine from the toilet water; consequently, the number of viruses emitted from the UDT after urination is affected by the efficiency of the UDT at separating urine. If all urine was separated by the UDT, the emissions for viruses excreted in urine would be zero, and up to 4-log10 fewer infectious viruses and 7-log10 fewer gc from urine would be emitted with the UDT compared to the MFT.

We estimated that HuNoV can be emitted at up to 108 gc from the MFT and 107 gc from the UDT per flush in the worst-case scenario. The worst-case scenario was calculated using the maximum reported viral loads in feces and urine and the highest fraction emitted for the surrogate virus from our experiments. Given that the probability of infection from human challenge experiments is 0.1 for a dose of 103 gc HuNoV to 0.7 for a dose of 108 gc,40 our estimates suggest that the amounts of HuNoV emitted from flushing are within the range of the infectious doses. Infectious doses of HAdV from ingestion are in the range of 10–500 TCID50.41 The minimum infectious dose from inhalation is lower, at approximately 0.5 TCID50.42 Assuming a gc to infectious virus ratio of 1 × 10–3 for HAdV,43 up to 8 × 104 and 2 × 104 infectious HAdV viruses are emitted with a flush of feces or urine in the MFT, respectively; these values are within the range of infectious doses. Infectious doses for HPyV are not available at the time of this study, so we cannot compare the estimated amount of HPyV emitted to its infectious dose.

Our emission results combined with the literature on virus excretion and infectious doses suggest that some viruses may be emitted from toilet flushing at levels that approach or exceed infectious doses. It is highly unlikely, however, that people in a restroom would be exposed to the total number of viruses emitted from a flush. Emissions that are smaller than 5 μm evaporate quickly and travel with the air plume.14 Environmental factors such as humidity, temperature, and the air exchange rate will impact the density of infectious viruses in the restroom after a flush, as well as user-dependent factors such as the inhalation rate and time spent in the restroom. Emissions greater than 5 μm settle onto surfaces near the toilet. Humidity, temperature, and surface material will affect the rate at which these viruses are inactivated, and user behaviors such as contact with surfaces, handwashing, and time spent in the restroom can affect exposure to infectious viruses.

Most studies measured either droplets or aerosols generated from flushing, whereas we captured the total number of infectious viruses emitted from the toilet, similar to a method used by Gerba et al.18 The total emission approach bypasses the challenge of choosing an appropriate sampling location and the assumption that the flush emissions are uniformly distributed in the air and on surfaces surrounding the toilet. The emissions we measured are estimates of the total virus emissions directly from the flush, leading to more comprehensive data for exposure and risk assessments. A QMRA using our emission levels can be used to quantify the risk of virus infection from toilet flushing and inform toilet use and maintenance behaviors. Because QMRAs are unique to specific viruses and exposure routes, namely, inhalation and ingestion, we estimated the fraction of viruses emitted in different particle size ranges using the particle size distribution of flush emissions from a previous study35 and our average fraction emitted for the surrogate viruses when protein was added to the toilet (Table 3). More viruses are emitted in the larger particle size ranges due to the larger volume. In future work, these data can be coupled with virus loading into the toilet, dose–response data, exposure time, and contact and inhalation frequency to quantify an individual’s exposure to infectious virus emissions from toilet flushing and their risk of infection.

Table 3. Log10 Fraction Emitted in Different Particle Sizes from Toilet Flushing.

  dsDNA viruses
 ssRNA viruses
particle size (μm) MFT UDT MFT UDT
0.3 –7.5 –8.3 –7.7 –8.2
0.5 –7.4 –8.3 –7.7 –8.2
1 –7.2 –8.1 –7.5 –7.9
3 –6.6 –7.5 –6.9 –7.4
5 –6.4 –7.3 –6.7 –7.1
10 –6.7 –7.6 –7.0 –7.5
Open in a new tab

Conclusions and Future Work

This work expands on toilet flushing as a source of exposure to viruses and compares emission levels of fecal-borne and urine-associated viruses from flushing an MFT and a UDT to inform future QMRAs. Like previous studies, we report that virus emissions were lower from the toilet with a lower flush pressure—we found that UDTs emitted fewer viruses than MFTs. For fecal-borne viruses, flush emissions were reduced by up to 2.3-log10 per flush in the UDT. Because UDTs can collect urine from the toilet without a flush, up to 4-log10 fewer infectious viruses and 7-log10 fewer gc from urine can be emitted with the UDT compared to the MFT. In MFTs, specific focus on urine-associated viruses is warranted as they are excreted at high levels into toilets, higher protein levels in urine can increase their emission levels, and urine accounts for a majority of toilet flushes. In particular, HPyV can be emitted at high levels, but infectious virus excretion in urine and dose–response data are necessary to quantify the risk of transmission from urine during toilet flushing. MS2 and T3 were emitted at similar levels, but more work is needed to confirm that the virus structure does not affect emissions, including the presence of a lipid envelope. Although we used T3 as a surrogate virus for HAdV because they are both dsDNA viruses that can be excreted in urine, HAdV is a larger virus (Table 1), and this difference may result in lower emissions. While our results demonstrate that the toilet type had the greatest effect on virus emission levels of the factors we studied, future work should evaluate higher protein levels, incorporation in feces (e.g., sorption to organic material and emissions of fecal particles containing viruses), and a range of flush pressures on virus emissions. A systematic approach to evaluating these properties is important for understanding how toilet design and environmental controls can affect human exposure to viruses from toilet flushing. Finally, the emission data we gathered in this study can be used in future QMRAs to quantify an individual’s risk of infection from viruses emitted during toilet flushing.

Acknowledgments

This work was supported by a University of Michigan Rackham Merit Fellowship to L.L., a University of Michigan MCubed grant, and National Science Foundation INFEWS T3 Grant No. 1639244.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsestwater.2c00521.

  • Recovery experiment methods and data, qPCR methods and data, and gene copy to pfu ratios (PDF)

Author Contributions

CRediT: Lucinda Li conceptualization (equal), formal analysis (equal), methodology (equal), writing-original draft (equal), writing-review & editing (equal); Jinyi Cai formal analysis (supporting), investigation (equal), methodology (equal); Joseph N.S. Eisenberg supervision (supporting), writing-review & editing (supporting); Heather E Goetsch conceptualization (equal), writing-review & editing (supporting); Nancy G. Love conceptualization (equal), formal analysis (equal), funding acquisition (equal), supervision (equal), writing-original draft (equal), writing-review & editing (equal); Krista R. Wigginton conceptualization (equal), formal analysis (equal), funding acquisition (equal), supervision (equal), writing-original draft (equal), writing-review & editing (equal).

The authors declare no competing financial interest.

Supplementary Material

ew2c00521_si_001.pdf (149.2KB, pdf)

References

  1. Atmar R. L.; Opekun A. R.; Gilger M. A.; Estes M. K.; Crawford S. E.; Neill F. H.; Graham D. Y. Norwalk Virus Shedding after Experimental Human Infection. Emerg. Infect. Dis. 2008, 14, 1553. 10.3201/EID1410.080117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Abney S. E.; Bright K. R.; McKinney J.; Ijaz M. K.; Gerba C. P. Toilet Hygiene—Review and Research Needs. J. Appl. Microbiol. 2021, 131, 2705–2714. 10.1111/JAM.15121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Hanaoka N.; Ito S.; Konagaya M.; Nojiri N.; Yasuda M.; Fujimoto T.; Deguchi T. Infectious Human Adenoviruses Are Shed in Urine Even after Disappearance of Urethral Symptoms. PLoS One 2019, 14, e0212434 10.1371/JOURNAL.PONE.0212434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barzon L.; Pacenti M.; Franchin E.; Pagni S.; Martello T.; Cattai M.; Cusinato R.; Palù G. Excretion of West Nile Virus in Urine During Acute Infection. J. Infect. Dis. 2013, 208, 1086–1092. 10.1093/INFDIS/JIT290. [DOI] [PubMed] [Google Scholar]
  5. Randhawa P.; Ho A.; Shapiro R.; Vats A.; Swalsky P.; Finkelstein S.; Uhrmacher J.; Weck K. Correlates of Quantitative Measurement of BK Polyomavirus (BKV) DNA with Clinical Course of BKV Infection in Renal Transplant Patients. J. Clin. Microbiol. 2004, 42, 1176–1180. 10.1128/JCM.42.3.1176-1180.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Urbano P. R. P.; Oliveira R. R.; Romano C. M.; Pannuti C. S.; da Fink M. C. D. S. Occurrence, genotypic characterization, and patterns of shedding of human polyomavirus JCPyV and BKPyV in urine samples of healthy individuals in São Paulo, Brazil. Brazil. J. Med. Virol. 2016, 88, 153–158. 10.1002/JMV.24318. [DOI] [PubMed] [Google Scholar]
  7. Adenovirus | CDC. https://www.cdc.gov/adenovirus/index.html (accessed 2021-11-30).
  8. Ahsan N.; Shah K. Polyomaviruses and Human Diseases. Adv. Exp. Med. Biol. 2006, 577, 1–18. 10.1007/0-387-32957-9_1. [DOI] [PubMed] [Google Scholar]
  9. Minnesota Department of Health . Norovirus Fact Sheet - Minnesota Dept. of Health. https://www.health.state.mn.us/diseases/norovirus/noro.html (accessed 2021-11-30).
  10. Carducci A.; Donzelli G.; Cioni L.; Verani M. Quantitative Microbial Risk Assessment in Occupational Settings Applied to the Airborne Human Adenovirus Infection. Int. J. Environ. Res. Public Health 2016, 13, 733. 10.3390/IJERPH13070733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Overbey K. N.; Hamra G. B.; Nachman K. E.; Rock C.; Schwab K. J. Quantitative Microbial Risk Assessment of Human Norovirus Infection in Environmental Service Workers Due to Healthcare-Associated Fomites. J. Hosp. Infect. 2021, 117, 52–64. 10.1016/J.JHIN.2021.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Shi K. W.; Huang Y. H.; Quon H.; Ou-Yang Z. L.; Wang C.; Jiang S. C. Quantifying the Risk of Indoor Drainage System in Multi-Unit Apartment Building as a Transmission Route of SARS-CoV-2. Sci. Total Environ. 2021, 762, 143056 10.1016/J.SCITOTENV.2020.143056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Johnson D. L.; Mead K. R.; Lynch R. A.; Hirst D. V. L. Lifting the Lid on Toilet Plume Aerosol: A Literature Review with Suggestions for Future Research. Am. J. Infect. Control 2013, 41, 254–258. 10.1016/J.AJIC.2012.04.330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. WHO . Infection Prevention and Control of Epidemic-and Pandemic Prone Acute Respiratory Infections in Health Care; WHO, 2014. [PubMed] [Google Scholar]
  15. Barker J.; Jones M. V. The Potential Spread of Infection Caused by Aerosol Contamination of Surfaces after Flushing a Domestic Toilet. J. Appl. Microbiol. 2005, 99, 339–347. 10.1111/j.1365-2672.2005.02610.x. [DOI] [PubMed] [Google Scholar]
  16. Sassi H. P.; Reynolds K. A.; Pepper I. L.; Gerba C. P. Evaluation of Hospital-Grade Disinfectants on Viral Deposition on Surfaces after Toilet Flushing. Am. J. Infect. Control 2018, 46, 507–511. 10.1016/J.AJIC.2017.11.005. [DOI] [PubMed] [Google Scholar]
  17. Abad F. X.; Pinto R. M.; Bosch A. Survival of Enteric Viruses on Environmental Fomites. Appl. Environ. Microbiol. 1994, 60, 3704–3710. 10.1128/AEM.60.10.3704-3710.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gerba C. P.; Wallis C.; Melnick J. L. Microbiological Hazards of Household Toilets: Droplet Production and the Fate of Residual Organisms. Appl. Microbiol. 1975, 30, 229. 10.1128/am.30.2.229-237.1975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lai A. C. K.; Tan T. F.; Li W. S.; Ip D. K. M. Emission Strength of Airborne Pathogens during Toilet Flushing. Indoor Air 2018, 28, 73–79. 10.1111/INA.12406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Johnson D.; Lynch R.; Marshall C.; Mead K.; Hirst D. Aerosol Generation by Modern Flush Toilets. Aerosol Sci. Technol. 2013, 47, 1047–1057. 10.1080/02786826.2013.814911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Udert K. M.; Wächter M. Complete Nutrient Recovery from Source-Separated Urine by Nitrification and Distillation. Water Res. 2012, 46, 453–464. 10.1016/J.WATRES.2011.11.020. [DOI] [PubMed] [Google Scholar]
  22. Rose C.; Parker A.; Jefferson B.; Cartmell E. The Characterization of Feces and Urine: A Review of the Literature to Inform Advanced Treatment Technology. Crit. Rev. Environ. Sci. Technol. 2015, 45, 1827–1879. 10.1080/10643389.2014.1000761. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Silverman D. T.; Alguacil J.; Rothman N.; Real F. X.; Garcia-Closas M.; Cantor K. P.; Malats N.; Tardon A.; Serra C.; Garcia-Closas R.; Carrato A.; Lloreta J.; Samanic C.; Dosemeci M.; Kogevinas M. Does Increased Urination Frequency Protect against Bladder Cancer?. Int. J. Cancer 2008, 123, 1644–1648. 10.1002/IJC.23572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jeyachandran Y. L.; Mielczarski J. A.; Mielczarski E.; Rai B. Efficiency of Blocking of Non-Specific Interaction of Different Proteins by BSA Adsorbed on Hydrophobic and Hydrophilic Surfaces. J. Colloid Interface Sci. 2010, 341, 136–142. 10.1016/J.JCIS.2009.09.007. [DOI] [PubMed] [Google Scholar]
  25. Okubo T.; Kobayashi K. Surface Tension of Biological Polyelectrolyte Solutions. J. Colloid Interface Sci. 1998, 205, 433–442. 10.1006/JCIS.1998.5632. [DOI] [PubMed] [Google Scholar]
  26. Gerba C. P. Applied and Theoretical Aspects of Virus Adsorption to Surfaces. Adv. Appl. Microbiol. 1984, 30, 133–168. 10.1016/S0065-2164(08)70054-6. [DOI] [PubMed] [Google Scholar]
  27. EPA . Method 1601: Male-Specific (F + ) and Somatic Coliphage in Water by Two-Step Enrichment Procedure; EPA, 2001. [Google Scholar]
  28. Kroger D. H.; Schroeder C. Bacteriophage T3 and Bacteriophage T7 Virus-Host Cell Interactions. Microbiol. Rev. 1981, 45, 9–51. 10.1128/mr.45.1.9-51.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Viscidi R. P.; Shah K. V.. Polyomaviruses. In Infectious Diseases, 3rd edition; 2010; vol 2, pp. 1570–1572.
  30. Ahi Y. S.; Mittal S. K. Components of Adenovirus Genome Packaging. Front. Microbiol. 2016, 7, 1503. 10.3389/FMICB.2016.01503/BIBTEX. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Valegård K.; Murray J. B.; Stonehouse N. J.; van den Worm S.; Stockley P. G.; Liljas L. The Three-Dimensional Structures of Two Complexes between Recombinant MS2 Capsids and RNA Operator Fragments Reveal Sequence-Specific Protein-RNA Interactions. J. Mol. Biol. 1997, 270, 724–738. 10.1006/JMBI.1997.1144. [DOI] [PubMed] [Google Scholar]
  32. Yates M. V.Norovirus. In Microbiology of Waterborne Diseases: Microbiological Aspects and Risks, 2nd edition; Academic Press, 2014; pp. 515–522. [Google Scholar]
  33. Michen B.; Graule T. Isoelectric Points of Viruses. J. Appl. Microbiol. 2010, 109, 388–397. 10.1111/j.1365-2672.2010.04663.x. [DOI] [PubMed] [Google Scholar]
  34. Lion T.; Kosulin K.; Landlinger C.; Rauch M.; Preuner S.; Jugovic D.; Pötschger U.; Lawitschka A.; Peters C.; Fritsch G.; Matthes-Martin S. Monitoring of Adenovirus Load in Stool by Real-Time PCR Permits Early Detection of Impending Invasive Infection in Patients after Allogeneic Stem Cell Transplantation. Leukemia 2010, 24, 706–714. 10.1038/LEU.2010.4. [DOI] [PubMed] [Google Scholar]
  35. Knowlton S. D.; Boles C. L.; Perencevich E. N.; Diekema D. J.; Nonnenmann M. W. Bioaerosol Concentrations Generated from Toilet Flushing in a Hospital-Based Patient Care Setting. Antimicrob. Resist. Infect. Control 2018, 7, 16. 10.1186/S13756-018-0301-9/TABLES/1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Newsom S. W. B. Microbiology Of Hospital Toilets. Lancet 1972, 300, 700–703. 10.1016/S0140-6736(72)92102-2. [DOI] [PubMed] [Google Scholar]
  37. Alvarez M.; Friend J.; Yeo L. Rapid Generation of Protein Aerosols and Nanoparticles via Surface Acoustic Wave Atomization. Nanotechnology 2008, 19, 455103 10.1088/0957-4484/19/45/455103. [DOI] [PubMed] [Google Scholar]
  38. Tolman R. C. The Effect of Droplet Size on Surface Tension. J. Chem. Phy.s 1949, 17, 333. 10.1063/1.1747247. [DOI] [Google Scholar]
  39. Chan M. C. W.; Sung J. J. Y.; Lam R. K. Y.; Chan P. K. S.; Lee N. L. S.; Lai R. W. M.; Leung W. K. Fecal Viral Load and Norovirus-Associated Gastroenteritis. Emerg. Infect. Dis. 2006, 12, 1278. 10.3201/EID1208.060081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Teunis P. F. M.; Moe C. L.; Liu P.; Miller S. E.; Lindesmith L.; Baric R. S.; le Pendu J.; Calderon R. L. Norwalk Virus: How Infectious Is It?. J. Med. Virol. 2008, 80, 1468–1476. 10.1002/JMV.21237. [DOI] [PubMed] [Google Scholar]
  41. Gutekunst R. R.; White R. J.; Edmondson W. P.; Chanock R. M. Immunization with Live Type 4 Adenovirus: Determination of Infectious Virus Dose and Protective Effect of Enteric Infection. Am. J. Epidemiol. 1967, 86, 341–349. 10.1093/oxfordjournals.aje.a120744. [DOI] [PubMed] [Google Scholar]
  42. Yezli S.; Otter J. A. Minimum Infective Dose of the Major Human Respiratory and Enteric Viruses Transmitted Through Food and the Environment. Food Environ. Virol. 2011, 3, 1–30. 10.1007/s12560-011-9056-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. McBride G. B.; Stott R.; Miller W.; Bambic D.; Wuertz S. Discharge-Based QMRA for Estimation of Public Health Risks from Exposure to Stormwater-Borne Pathogens in Recreational Waters in the United States. Water Res. 2013, 47, 5282–5297. 10.1016/J.WATRES.2013.06.001. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ew2c00521_si_001.pdf (149.2KB, pdf)

Articles from ACS Es&t Water are provided here courtesy of American Chemical Society

RESOURCES