Bacterial Contamination of Public and Household Restrooms, and Implications for the Potential Risk of Norovirus Transmission

Abstract

The transmission of infectious diseases via the use of public restrooms has been previously documented. The goal of this study was to compare bacterial contamination in public vs. household restrooms and, using quantitative microbial risk assessment (QMRA), to assess the probability of infection from fomite contact with selected high-touch sites within the restrooms. Fomite surfaces in four public and four household restrooms were sampled over a period of two months. The public restrooms were in an office building occupied by 80 individuals and were considered moderate usage. The toilet seat, toilet flush handle, countertops, and floor were sampled for heterotrophic, coliform, and Escherichia coli bacteria. The highest numbers of heterotrophic bacteria and coliforms were detected on the countertops, followed by the floor. The greatest numbers of E. coli were recovered from the countertops in the household restroom, but the greatest numbers in the public restroom were recovered from the toilet flush handle. Numbers of heterotrophic bacteria and coliforms were 10 to 100 times greater in household restrooms than in public restrooms. The QMRA suggested that the greatest risk of acquiring a norovirus infection involved the touching of the countertops in household restrooms and the toilet flush handles in public restrooms.

1. Introduction

The transmission of pathogenic bacteria and viruses associated with public restrooms has been documented previously [1,2,3,4], and is believed to result primarily from hand contact with pathogen-contaminated surfaces and the subsequent transfer from the contaminated hands to the mouth, followed by ingestion [1,4,5]. Contamination of surfaces in restrooms may occur through touching with contaminated hands, coughing or sneezing, and through aerosols created during toilet flushing [3,4,5,6,7,8,9,10,11,12], and use of faucet taps during hand washing [13]. Li et al. [3] found that 40–60% of the particles in the toilet during flushing reached the toilet seat. The transmission of respiratory and enteric viruses via inhalation of viruses aerosolized during toilet flushing and other restroom activities has also been postulated [4,5,14,15]. A Quantitative Microbial Risk Assessment (QMRA) model developed by Higham et al. [4] using droplet production during toilet flushing suggested that ventilation plays an important role in risk of transmission of norovirus and SARS-CoV-2. The greatest risk of transmission occurs for viruses which have a relatively low infectious dose (ID₅₀) in humans [16]. Adenoviruses, noroviruses, and other viruses which infect humans have been detected on public restroom surfaces and in restroom air [6,17]. For instance, adenoviruses were detected on 71% of surface samples and 75% of air samples taken in two public restrooms tested in one study [6].

To our knowledge, there exist no published studies which have directly compared the recovery from fomites in public vs. household restrooms of heterotrophic and enteric bacteria in the same study nor the potential risk of disease transmission imparted by such fomite contamination. Previous studies have focused either on public restrooms or household restrooms [13,17,18,19]. Potentially, exposure to pathogens in public restrooms may be greater because of the greater numbers of different individuals using public restrooms compared to household restrooms, greater turbulence generated during flushing, more frequent use, with a greater potential production of pathogen-containing aerosols during flushing [5,10,11]. For this reason, it has also been suggested that public toilets may serve as reservoirs for the transmission of antibiotic- or multidrug-resistant bacteria in a community [19].

There is a perception that public restrooms present a greater risk of infection transmission, and many individuals limit their use of public restrooms for this reason [20,21]. To address this question, this study was designed to compare the relative levels of heterotrophic and enteric bacteria (coliforms and E. coli) on public vs. household restroom fomites and to leverage these results in assessing the probability of acquiring an infection with norovirus through touching highly contaminated restroom fomites.

2. Materials and Methods

2.1. Sampled Restrooms

Restrooms in four households in Tucson, AZ, USA with at least four occupants were sampled four times each over the course of two months. This resulted in each location (surface) within the household restroom being sampled a total of 16 times. The household toilets were typical siphonic toilets with attached water tanks and no gap in the front of the seat. The four public restrooms (two male and two female) that were sampled over a period of five weeks were in an office building in the same city in which ~80 people worked daily during a five-day work week. The public toilets were tankless and used in-line water pressure for flushing. The seats were U-shaped with a gap in the front, as required by uniform United States plumbing codes.

The numbers of samples collected in the restrooms totaled 40 from the floor, toilet flush handle and top of the toilet seat, and 30 total from the countertops. Samples were collected in the afternoons from both the household and public restrooms. The areas of the sampled surfaces varied from 10 cm² for the toilet flush handles to 385 to 500 cm² for the restroom floors. All bacterial count data were normalized to numbers of colonies recovered per 100 cm² of the sampled surfaces.

2.2. Sample Collection

Samples were collected using Sponge-Sticks (3M, Minneapolis, MN, USA) infused with 10 mL of the Dey–Engley neutralizing broth (D/E Broth, St. Paul, MN, USA) to neutralize the presence of any residual disinfectants that might have been present. The collected samples were immediately placed onto ice packs and stored overnight in a refrigerator.

2.3. Sample Processing and Extraction

The Sponge-Sticks were removed from refrigeration the next morning and mechanically agitated using a Stomacher (Seward, West Sussex, UK) for 120 s at normal speed to facilitate the release of captured bacteria. The Sponge-Sticks were then pressed firmly to release the liquid D/E broth. The resulting volumes of the D/E broth were measured and transferred into sterile conical tubes. An additional unused Sponge-Stick was processed concurrently with the surface samples as a negative assay control.

2.4. Heterotrophic Plate Count (HPC) Assays

Heterotrophic plate count (HPC) bacteria were enumerated in all study samples. The D/E sample extracts were vortexed, and an aliquot was serially diluted 1:10 over several dilutions in 0.1 M phosphate-buffered saline (PBS, MP Biomedicals, Solon, OH, USA). A minimum of three dilutions were spread-plated onto prepared R2A media (Hardy Diagnostics, Santa Maria, CA, USA) in replicates of two. All plates were incubated for seven days at room temperature in the dark. The resulting colony-forming units (CFUs) were then counted and recorded for each sample.

2.5. Colilert Assays for Total Coliforms and Escherichia coli

Total coliforms and E. coli are widely recognized indicators of unsanitary conditions and fecal contamination, respectively. Each was assayed in all samples that were collected from the restrooms using the Colilert assay (Colilert Quanti-Tray system, IDEXX, Westbrook, ME, USA) and quantified by the most probable number (MPN) method. An aliquot of the D/E extract from each sample sponge was added to a sterile mixing vessel. The total volume of the sample was brought up to 100 mL using sterile deionized water. The Colilert reagent was added and mixed until dissolved; then, the mixture was loaded into a Quanti-Tray. Loaded and sealed Quanti-Trays were incubated for 18 to 24 h at 37 °C. Quanti-Tray wells demonstrating a yellow color change were scored as positive for total coliforms; wells demonstrating both the yellow color change and fluorescence were scored as positive for E. coli.

2.6. Quantitative Microbial Risk Assessment (QMRA)

The QMRA model developed by Abney et al. [16] was used to assess the probability of acquiring a norovirus infection from touching contaminated restroom surfaces. This model includes percent transfer of the virus from a contaminated fomite to the hand, and percent transfer from fingers to the mouth followed by ingestion. The risk of infection was based on a one-time exposure event. The QMRA methodology has been described in detail previously [16]. The concentrations of E. coli recovered from the restroom surfaces in this study were used as a surrogate for norovirus concentrations for performing the QMRA. The rationale for this is that the ratios of norovirus numbers to E. coli numbers recovered in the stools of people infected with these enteric pathogens are approximately 1:1 [22] and the exposure to norovirus in a restroom used by an infected person is therefore assumed to be informed by the E. coli recovery data for the purposes of this risk assessment. Human norovirus infectivity likely persists longer on surfaces than does E. coli, so norovirus infection probabilities are possibly greater than those estimated based on the E. coli recovery values [1].

3. Results

Recovery of Bacteria from Restroom Fomites

The greatest numbers of heterotrophic bacteria and coliforms were recovered from the restroom floors and countertops, both in public (male and female restroom data were pooled) and household restrooms (

Table 1. Recoveries of heterotrophic and coliform bacteria from restroom surfaces ¹.

Restroom Surface	Heterotrophic Plate Count Bacteria (CFU per 100 cm2)		Coliforms (MPN per 100 cm2)
	Public	Household	Public	Household
Toilet seat top	1611 ± 1607 2	38,800 ± 31,800	3.94 ± 2.32	8560 ± 10,800
Toilet flush handle	487 ± 532	60,000 ± 73,600	22.2 ± 64.2	<0.01
Countertop	1703 ± 1709	2,260,000 ± 2,860,000	0.784 ± 1.46	155,000 ± 202,000
Floor	16,567 ± 21,355	561,000 ± 797,980	198 ± 362	94,125 ± 47,127

¹ Abbreviations used: CFU, colony-forming units; MPN, most probable number. ² Values shown are the means ± standard deviations.

). The levels of these bacteria were 10- to >100-fold greater in the household restrooms than in the public restrooms. Greater numbers of these bacteria occurred on the floor in the public restroom than in the household restroom, likely because of greater foot traffic occurring in the former case. While low numbers of coliforms were recovered from the public toilet flush handles, none were recovered from the household toilet flush handles (Table 1).

The numbers of E. coli MPN recovered were greater on the fomite surfaces in the public restrooms (male and female restroom data were pooled), except for the toilet flush handles, for which numbers were greater in the household restrooms (

Table 2. Recoveries of Escherichia coli from restroom surfaces.

Restroom Surface	Escherichia coli (MPN 1 per 100 cm2)
	Public	Household
Toilet seat top	1.89 ± 4.32 2	4.2 ± 6.2
Toilet flush handle	22.2 ± 64.2	<0.01
Countertop	0.5 ± 0.0	1760 ± 1760
Floor	0.29 ± 0.79	255 ± 234

¹ Abbreviation used: MPN, most probable number. ² Values shown are the means ± standard deviations.

). The greatest numbers of E. coli recovered in the household restrooms were from the countertops, while in the public restrooms the greatest numbers were from the flush handles.

The estimated probabilities of acquiring a norovirus infection resulting from touching household vs. public restroom surfaces are shown in

Table 3. Estimated probability of acquiring a norovirus infection from touching restroom surfaces. Probabilities of infection exceeding 0.0001% (10⁻⁴%; 1:10⁻⁶) are of concern based on U.S. EPA. public health guidelines for drinking water [23].

Restroom Type	Toilet Seat	Toilet Flush Handle	Countertop
	(% Risk of Infection)
Public	2.4 × 10−4	6.1 × 10−4	6.3 × 10−5
Household	5.2 × 10−4	<4.0 × 10−7	1.1 × 10−1

. The greatest risk was determined to involve touching the countertops in the household restrooms and the greatest risk in public restrooms involved touching the toilet flush handles. The risk associated with touching the toilet seats did not differ substantially between the public vs. household restrooms.

4. Discussion

Both heterotrophic plate count bacteria and coliforms were recovered at numbers 10- to >100-times greater on all tested sites in the household vs. public restroom. The exception was the toilet flush handles in the household restrooms, on which no heterotrophic bacteria or E. coli were detected. It is possible that this discrepancy reflects the differences in frequency of use or in the design of flush handles in public vs. household toilets. The public restroom flush handles are designed for a hand grip whereas less hand contact with the handles is required for flushing the household toilets.

Greater concentrations of E. coli were recovered on the countertops, toilet seats, and floors in household vs. public restrooms. This may result from several factors, including frequency of use and of cleaning/disinfection, types of disinfection products used, user habits, type and usage of toilet paper, degree of aerosolization of bacteria from toilet flushing, design differences of the toilet seat, the numbers of individuals using the restroom during the day, and the fact that primarily adults use the public restrooms. Of these factors, perhaps the major difference is the frequency of cleaning/disinfecting of the restrooms, which occurred once a day in the office building restrooms evaluated. Most homeowners (71%) clean the toilet once a week or less [24].

While several studies have investigated the presence of bacteria in public and household toilets, few have quantified the levels of bacteria on fomites [1,18,25,26]. Rusin et al. [18], in a study of 15 household restrooms over a period of 10 weeks, found geometric average numbers of heterotrophic bacteria and coliforms per 100 cm² in the range that we observed for the toilet seats, countertops, and floors in the present study. However, these authors detected coliform bacteria on the household toilet flush handles, while we did not recover coliforms from this restroom surface. The latter likely reflects the inherent variability in bacterial sampling outcomes from restroom surfaces observed in this and several previous studies [1,18,27,28].

Conducting studies on the detection of viruses (as opposed to bacteria) on fomites in the built environment is both costly and time consuming. In addition, infected occupants must be present during the study of viral contamination of public vs. household restrooms. This is more feasible with public restrooms because of the larger numbers of individuals using the restrooms, but more challenging in the household restrooms as there are fewer users. The present study explored an alternative approach to overcome this limitation: that of leveraging the E. coli numbers in human feces for estimating the concentrations of norovirus in feces in a QMRA model used to estimate the risk of acquiring a norovirus infection through touching surfaces in household vs. public restrooms. Norovirus outbreaks occurring as a result of public toilet use have been documented [1]. Studies of norovirus infections acquired through use of household restrooms are much more challenging, since these are not usually investigated by public health departments. Assuming the E. coli numbers are representative of norovirus numbers [22], the greatest risk of acquiring a norovirus infection in both the household and public restrooms appears to result from touching the countertops, with a somewhat greater risk in household restrooms. Touching the toilet flush handles appears to present the greatest risk of infection in public restrooms. These risks are likely underestimated because norovirus can survive on fomites longer than enteric bacteria [1]. However, the approach used in this study provides insights into differences in potential risks of acquiring norovirus infections through use of public vs. household restrooms.

A limitation of this study is the differences in the actual frequency of use of the toilets. The average household toilet is used about five times per day [29,30], and the same for the average office toilet (men’s or women’s) [30]. In the 80-person office building evaluated in this study, we would, therefore, estimate the toilets (not urinals) to be flushed ~320 times per day. This would result in approximately 35 uses per toilet per day. The households evaluated in this study averaged four individuals per household with two toilets, resulting in an estimated 10 uses per day. Thus, usage in the household toilets likely occurred with a lower frequency than the public restroom toilets evaluated in this study. Of course, the size of the public facility could affect these results. The public restrooms in this study were cleaned once per day. The cleaning of the household restrooms in this study varied from once a day to once a week. The greater numbers of bacteria recovered from countertops in the household restrooms may also reflect the use of personal care products, and placement (storage after use) of such products on this surface, compared to little or no products placed on the countertops in public restrooms. In addition, the studied households had at least two children, who may not yet have developed the hygiene habits typically observed by adults (e.g., adequate hand washing). Ibrahim et al. [19] suggested that public restroom design, socioeconomic factors, and cultural habits in a region will influence cleanliness and potential risk of microbial transmission. Use of toilet paper vs. water sprays for cleansing may also represent an important factor.

The results of this study suggest that controlling fomite contamination in household restrooms may be more challenging than in public restrooms used primarily by adults. A lower frequency of cleaning/disinfecting of fomites and less stringent grooming and personal hygiene habits likely combine to make limiting exposure to potential pathogens more difficult to achieve in the household restrooms. The use of disinfectants that impart a residual microbicidal activity onto the surface may be one approach for reducing this exposure during routine cleaning/disinfecting [31]. More recently, the spraying of an air sanitizer after toilet flushing has been shown to reduce viruses in the air and on surfaces by reducing the cross-contamination of fomites due to gravitational settling of pathogens that originate from flushing the toilets used by infected patients [32]. It has been reported that rural home toilet seats are more likely to serve as potential reservoirs and disseminators of antibiotic-resistant bacteria (AMRs) than public restrooms, highlighting the need for more frequent decontamination of household toilets and other restroom fomites, particularly when an occupant of the household is infected [28,33]. In contrast, a study of public restrooms in a shopping mall reported that toilet seats were highly contaminated with multi-drug resistant (MDR) bacteria, indicating the need to monitor and take measures to reduce pathogen transmission [19]. It has also been suggested that improving the ventilation or waiting at least 60 s between occupants in the restroom may decrease the risk of acquiring a norovirus or SARS-CoV-2 infection due to previous use by an infected person [4].

5. Conclusions

We have demonstrated that fomites in public restrooms used primarily by adults are less contaminated by bacteria, including bacteria of fecal origin, than household restrooms in the United States. This likely results from differences in the frequency of the cleaning and disinfection of the fomites, and differences in how these facilities are used (e.g., used primarily by adults vs. adults and children, and differences in how countertops are used for personal care product storage purposes). The greatest risk of acquiring a norovirus infection in the household restroom appears to result from touching the countertops, while in the public restrooms the greatest risk was from touching the flush handle. The toilet seat was found to be similarly contaminated in public vs. household restrooms. Despite the fact that public restrooms are cleaned frequently, the toilet flush handles appear to remain highly contaminated and as such, this touch surface represents a greater exposure risk in public than in household restrooms.