Impact of Different Toilet Cleaning/Disinfecting Regimens on Reducing the Risk of Exposure to Toilet-Borne Pathogens in American Household Restrooms

Abstract

Restrooms are associated with the transmission of bacterial and viral illnesses. Disinfecting contaminated surfaces is associated with reducing transmission risk. The goal of this study was to determine how cleaning/disinfecting frequency affects restroom pathogen contamination. The Phase 1 intervention included cleaning toilet surfaces (the toilet bowl, water, and rim) using 9.5% w/w hydrochloric acid. The Phase 2 intervention used fomite-specific products to clean/disinfect additional restroom surfaces, including the vanity countertop, sink faucet handle, toilet seat, flush handle, floor, and doorknob. A designated household member was responsible for cleaning/disinfecting surfaces at the beginning of each interval. Fomite sample collection was randomized, and samples were tested for heterotrophic, coliform, and Escherichia coli bacteria after specified intervals: 1, 2, 3, or 7 days. The greatest numbers of bacteria and largest reductions occurred on fomites after three days. A statistically significant difference was found for heterotrophic bacteria (p = 0.009), coliforms (p = 0.10), and E. coli (p = 0.13) with cleaning/disinfecting every three days. A quantitative microbial risk assessment (QMRA) estimated a >98% reduction in risk of infection by norovirus with an every-3-day cleaning/disinfection routine on the most heavily contaminated sites. Results indicate an optimal cleaning frequency of twice weekly for minimizing exposure to pathogens.

1. Introduction

Restrooms have been associated with the transmission of pathogens, as they are a source of enteric bacteria and viruses resulting from the occurrence of pathogens in feces and urine [1]. During continued use, the toilet bowl becomes contaminated, and pathogens can persist on the bowl surfaces and in toilet bowl water even after multiple flushes [2]. Salmonella has been shown to persist for weeks on surfaces under the toilet rim in homes following salmonellosis [3]. Biofilms developing at the water–air interface in toilets may extend the persistence of pathogens [4].

While exposure to pathogens in the toilet bowl may be limited, numerous studies have shown that viruses and bacteria are aerosolized during each flush [1]. Toilet flushing can cause cross-contamination of multiple areas in the restroom, including the toilet seat and lid, the sink, countertops, rug, and floor, and restroom contamination can likely spread throughout the home [5,6,7,8]. Transmission of these pathogens can occur both by contact with contaminated fomites and through exposure to aerosolized (airborne and droplet) pathogens. Closure of the toilet lid during flushing does not prevent contamination of fomites, nor does it control the aerosols generated by flushing [8]. Just lifting the toilet seat can result in a risk of infection from viruses deposited on the toilet seat from flushing [9]. The use of disinfectants or a toilet bowl cleaner has been shown to reduce the degree of bacterial contamination in the bowl [1] and the amounts of viruses aerosolized during flushing [1,8]. Also, contact with the flush handle may contaminate the hands [1].

Even if proper handwashing is practiced after toilet use, the risk of contamination of the hands may still exist from touching the faucet handle, sink, countertops, and door handle [1]. Failure to properly wash hands may lead to contamination of the doorknob, other restroom surfaces, or even other areas of the home. Rusin et al. [10] demonstrated that the use of hypochlorite cleaners in the restroom for cleaning and disinfection could significantly reduce contamination of fomites by coliforms and E. coli. Scott and Bloomfield [11] found that the levels of enteric bacteria in the toilet could be reduced by disinfection of surfaces within the toilet bowl. However, in their studies, the impact of frequency of cleaning/disinfecting was not studied. The goal of the present study was to assess the optimal frequency of cleaning and disinfecting for control of enteric bacteria in the American household restroom.

The research study was designed to determine the following:

• The impact of cleaning/disinfection frequency on restroom levels of enteric bacteria to reduce the risk of pathogen exposure/transmission.
• Phase 1: The impact of the use of a single disinfectant (a hydrochloric acid-based toilet bowl cleaner only).
• Phase 2: The impact of using multiple types of disinfectants (

  Table 1. Cleaning/disinfecting products and site of use in household restrooms.

  Product 1	Site of Application	Cleaning/Disinfecting Product Active Ingredient
  Toilet bowl cleaner	Toilet bowl water, toilet rim, and bowl surface above the water level	Hydrochloric acid (9.5% w/w)
  Disinfecting wipes	Toilet seat bottom and top, toilet lid	Alky dimethyl benzyl ammonium chloride (0.26%)
  Disinfectant spray	Doorknob and toilet flush handle	Alkyl dimethyl benzyl ammonium saccharinate (0.10%), ethanol (58%)
  Bathroom foamer	Vanity countertop and sink faucet handle	Citric acid (2.5%)
  Multipurpose surface cleaner	Floor	Alkyl dimethyl benzyl ammonium chloride (1.18%)

  ¹ All products were manufactured by Reckitt Benckiser LLC, Montvale, NJ, USA.

  ) on multiple surfaces (bundle cleaning/disinfection approach).

2. Materials and Methods

2.1. Selection of Households

A total of four households were selected for this study. To qualify, each household had to include at least four members (range 4 to 7), including at least two children under 18 years of age. Households 1 and 3 consisted of two working adults and two school-age children under 18. The stay-at-home adults in households 1 and 3 were females who cooked, cleaned, and cared for the children (i.e., a stay-at-home mom and a grandmother, respectively). Households 2 and 4 consisted of one stay-at-home adult, two working adults, and two or three school-age children under 18. Stay-at-home adults in households 2 and 4 consisted of a college-enrolled male and a grandfather, respectively. Household restrooms contained an American-style toilet located less than two feet from the small vanity/counter with a sink. Occupants could not be heavy disinfectant users prior to the study and had to be willing to use products containing disinfectants on hard surfaces per the use directions on the product label. The households were all located in Tucson, AZ, USA. In Phase 1, households were asked to follow their normal cleaning/disinfecting practices, except they were given a product that contained 9.5% w/w hydrochloric acid and a toilet bowl brush to clean the toilet bowl, toilet water, and the toilet rim at frequencies of once a day, once every 2 days, once every 3 days, or once a week. At the end of each cleaning interval, restroom samples were collected by a member of the laboratory staff from the University of Arizona. In Phase 2 of the study, the households were provided with a selection of cleaning products and asked to use them at specific sites in the restroom and at the same frequencies as during Phase 1. The dates of sampling were randomized during the week so that sampling did not occur on the same days of the week. The cleaning/disinfection products used are listed in Table 1.

2.2. Bacterial Sampling and Testing

All homes were assessed for heterotrophic plate count (HPC) bacteria, E. coli_, and coliforms. A 3M sponge stick with Dey–Engley neutralizing broth (D/E; St. Paul, MN, USA) was used to collect samples from restroom fomites for bacterial enumeration. The sponge sticks were placed on ice in an ice chest for transport to the laboratory. Sponge sticks were mechanically agitated for 120 min and then manually pressed with a seam press roller to facilitate the release of the bacteria and the liquid D/E broth extract. The resulting volume of D/E broth was measured and transferred to a sterile conical tube using a sterile 10 mL pipette. An additional, unused sponge stick was processed concurrently with the samples as a negative assay control. Fomite areas sampled within the restroom included the vanity countertop (585 cm²), sink faucet handle (10 cm²), doorknob (25 cm²), toilet bowl water (10 mL), toilet flush handle (10 cm²), toilet seat bottom (300 cm²), toilet seat top (300 cm²), toilet lid (200 cm²), floor (a total of 300 cm², with 100 cm² sampled in front of and on both sides of the toilet), inside surface of the bowl above the water line (about 558 cm²), and the toilet rim (377 cm²). The volume of water in the household toilet bowls was ~2.2 gallons (8.327 L), and the 10 mL sample of bowl water was obtained during each interval using a sterile 25 mL pipette. The samples collected were transferred to sterile conical tubes and refrigerated until assayed. Total numbers of heterotrophic bacteria were determined using the colony-forming unit (CFU) assay on Trypticase soy agar. Coliforms and E. coli were determined by the most probable number (MPN) test using the Colilert Quanti-tray system (IDEXX, Westbrook, ME, USA) following incubation for 24 h at 37 °C [12]. The results have been presented as the arithmetic average for samples taken from all of the households.

2.3. Statistical Analysis

An unpaired one-tailed T-test was performed on quantified bacterial data to determine the statistical difference between the group mean bacterial (heterotrophic, coliforms, or E. coli) numbers on fomite surfaces after the use of the toilet bowl cleaner only (Phase 1) vs. the use of multiple bundled disinfectants used on multiple surfaces (Phase 2). Statistical significance was set at p < 0.05 for heterotrophic bacteria and p ≤ 0.1 for coliform and E. coli bacteria due to the lower occurrence of coliforms in the indoor environment.

2.4. Quantitative Microbial Risk Assessment (QMRA)

A QMRA approach was used to assess the effect of cleaning/disinfecting on the reduction in risk of norovirus transmission. The products used on the toilet seat and countertop were disinfectants registered with the United States Environmental Protection Agency for efficacy against norovirus. Norovirus was selected for this QMRA because it survives for prolonged periods on environmental surfaces and has a low infectious dose in humans [1]. The goal of the QMRA was to estimate the risk (probability of infection) from exposure to norovirus. The model described by Abney et al. [13] was used to determine risk before and after the two cleaning phases. The model includes the percent transfer of the virus from the fomite to the hand, the percent transfer of the virus from the hand to the mouth, and ingestion and the resulting probability of infection. The two exposure scenarios selected were those exhibiting the greatest reductions in surface levels of E. coli, namely:

• Scenario 1. Hand contact is made during the lifting of the toilet seat. This results in the contamination of the hand, which is then brought to the mouth, resulting in the ingestion of the pathogen.
• Scenario 2. Hand contact is made with the restroom vanity countertop. This results in the contamination of the hand, which is then brought to the mouth, resulting in the ingestion of the pathogen.

For each of the above scenarios, the probability of infection from a single restroom visit was determined. The QMRA model parameters used were previously detailed [13]. We assumed that concentrations of norovirus on fomite surfaces are not influenced by multiple restroom uses over time. Therefore, we did not assume cumulative increases in fomite surface virus concentrations over time. Details of the analytical methods can be found in Abney et al. [13].

To estimate exposure before and after the use of the two cleaning phases, the concentrations of E. coli detected on the surfaces were used as an approximation of norovirus concentrations. This assumption was based on the finding that the levels of E. coli detected in the feces of an infected person are similar to the number of norovirus particles detected in the feces of an infected person [14]. E. coli was only detected frequently or in high enough concentrations to perform a QMRA on the bottom of the toilet seat and on the vanity countertop. Concentrations of E. coli on other restroom surfaces were not used in the QMRA exposure assessment due to the infrequency of positive E. coli detections from these surfaces.

We assumed that the starting norovirus concentrations were zero at the beginning of each restroom visit, as we lacked the data to properly estimate surface contamination that might be transferred from outside of the restroom prior to each restroom visit. Any norovirus present on hands would likely be lost due to hand hygiene interventions outside of the restroom or transfer of norovirus from the hand to other surfaces.

3. Results

3.1. Detection of Bacteria on Restroom Fomites

The numbers of heterotrophic bacteria detected on various restroom fomites have been displayed in Figure 1 and Figure 2. The greatest numbers of heterotrophic bacteria on all sample locations were detected on day 3 after cleaning/disinfecting, as were the greatest reductions in fomite cross-contamination in Phase 2 following the application of the bundled cleaning/disinfection products. There was a statistically significant difference between the Phase 1 cleaning (toilet bowl cleaner use only) and the Phase 2 cleaning (bundling of multiple products for multiple surfaces) indicated for the every-3-day cleaning cycle with heterotrophic bacteria, p = 0.009 (Figure 1 and Figure 2). The same trends were observed for the detection of coliform bacteria and E. coli on restroom fomites (

Table 2. Average coliform and Escherichia coli counts ± S.D. ¹ per 100 cm² surface area on restroom fomites from sampling from all days.

Restroom Site	Escherichia coli MPN		Coliform MPN
	Phase 1	Phase 2	Phase 1	Phase 2
Toilet rim	4670 ± 9341	1.6 ± 3.3	5213 ± 9276	49.6 ± 59.7
Toilet seat bottom	35.8 ± 71.5	0	29.9 ± 277	0
Toilet seat top	4.2 ± 8.3	21.2 ± 25.6	8562 ± 2854	20.2 ± 22.5
Toilet lid	0	0	26.3 ± 8.8	22.2 ± 44.5
Toilet flush handle	0	250 ± 500	0	2020 ± 3223
Vanity countertop	1757 ± 2470	270 ± 423	118,328 ± 39,443	31,300 ± 36,500
Sink faucet handle	0	966 ± 1856	9,760,450 ± 3,253,483 2	31,271 ± 49,172
Exit doorknob	0	0	36,300 ± 12,096	167 ± 236

¹ Abbreviations used: MPN, most probable number; Phase 1 (use of toilet bowl cleaner only); Phase 2 (bundling of multiple products for multiple surfaces); S.D., standard deviations. ² Represents the relatively high degree of variability in the sampling data.

and

Table 3. Escherichia coli MPN ¹ detected on surfaces per 100 cm² ± S.D. and percent reduction on touch surfaces sampled during the every-3-day (twice a week) cleaning and used for quantitative microbial risk assessment.

Restroom Site	MPN for Phase 1	MPN for Phase 2	Reduction in MPN
Vanity countertop	6009 ± 10,107	0	≥99.98% (3.7 log10)
Toilet seat bottom	47.3 ± 81.9	0	≥98.0 (1.7 log10)

¹ Abbreviations used: MPN, most probable number; Phase 1 (use of toilet bowl cleaner only); Phase 2 (bundling of multiple products for multiple surfaces); S.D., standard deviations.

and Supplemental Materials Tables S2 and S3). However, the p-values for coliforms and E. coli (see Supplementary Materials and Table 2) were greater than those obtained for heterotrophic bacteria and statistically significant at the 0.1 level. T-test significance in the case of coliforms and E. coli was set at a higher value due to the generally lower occurrence of coliforms in an indoor environment, especially at low humidity. Additionally, only three homes were evaluated for coliforms and E. coli due to sample contamination for household 4; the small sample size may have also impacted the results. The statistical differences for coliforms (p = 0.10) and E. coli (p = 0.13) were significant only for the every-3-day (twice a week) cleaning and sampling sites of the toilet rim, toilet seat bottom and top, and vanity countertop (p = 0.10).

Table 3 shows the amount of E. coli (MPN/100 cm² surface area) detected on the vanity countertop and the toilet seat bottom for the every-3-day (twice a week) cleaning. These were the touch sites where the Phase 2 intervention resulted in the greatest reductions in E. coli MPN relative to the Phase 1 values, per Table 2.

3.2. Quantitative Microbial Risk Assessment (QMRA) of Infection Probability for Norovirus

Similarities in the pathogen concentration in the feces of patients infected with E. coli vs. norovirus [14] have been leveraged such that the E.coli quantitation data on restroom surfaces shown in Table 2 and Table 3 informed QMRA modeling to determine the relative risk of acquiring a norovirus infection following use of a restroom cleaned using the Phase 1 vs. Phase 2 cleaning interventions.

Table 4. QMRA ¹ determined risk (% probability) of acquiring a norovirus infection from a single contact with restroom sites contaminated with norovirus during the every-3-day (twice a week) cleaning.

Restroom Site Touched	Risk for Phase 1	Risk for Phase 2	Reduction in Risk
Vanity countertop	1.14 × 10−1	≤3.30 × 10−4	≥99.7%
Toilet seat bottom	5.8 × 10−3	≤3.30 × 10−4	≥94.3%

¹ Abbreviations used: Phase 1 (use of toilet bowl cleaner only); Phase 2 (bundling of multiple products for multiple surfaces per Table 1); QMRA, quantitative microbial risk assessment.

shows the probability of acquiring a norovirus infection from use of restrooms cleaned using the two cleaning interventions (Phase 1: use of toilet bowl cleaner only vs. Phase 2: bundling of multiple products on multiple surfaces per Table 1). The modeling demonstrated that the risk of acquiring a norovirus infection from touching the vanity countertop would be reduced by >94.3% and the risk from lifting the toilet seat would be reduced by >99.7% by using the Phase 2 cleaning approach.

4. Discussion

The greatest number of heterotrophic bacteria and coliforms were detected on the toilet bowl, bowl rim and seat, floor, vanity countertop, and sink faucet handle. The greatest number of heterotrophic bacteria were detected on the toilet bowl rim, and the greatest number of coliform bacteria were recovered from the sink faucet handle. The greatest number of E. coli were detected on the toilet bowl rim.

The bundling of multiple disinfectants for multiple surfaces (Phase 2) reduced the number of heterotrophic bacteria, coliforms, and E. coli during cleaning/disinfecting on days 1, 2, and 3, but not during day 7. The Phase 2 intervention resulted in a cleaner toilet bowl as compared to the Phase 1 intervention, in which only the toilet bowl cleaner was used. This might have resulted from a reduction in resuspension/re-aerosolization of bacteria from cleaned or disinfected surfaces in the restroom, i.e., due to breaking the chain of infection transmission [15]. When restroom surfaces are clean (e.g., contain less dust and bacteria), activities conducted while using the restroom result in fewer bacteria being resuspended/re-aerosolized and subsequently settling from the air onto surfaces [16]. Therefore, a cleaning cycle performed every three days using bundled products reduces/interrupts bacterial contamination of restroom surfaces and subsequent gravitational deposition into the toilet bowl [15].

It is not clear why cleaning every three days (twice a week) has the greatest impact (i.e., relative to the other possible cleaning frequencies evaluated); however, lower bacterial quantities may be related to the development of biofilms on toilet bowl surfaces. A study by Pitts et al. [17] found that biofilm growth occurs exponentially, reaching maximum numbers around 6 days following inoculation, with 90% of the bacterial growth occurring within the first 2–3 days. Our study reflected near-peak levels of bacteria on the restroom surfaces in 3 days. It appears that at three days, the biofilm is most susceptible to removal in the transitional stage and more resistant to removal at one week because it has reached a “mature” growth stage (stationary phase). Research shows that enteric bacteria are more resistant to disinfectants during periods of low nutrient availability [18,19]. Additionally, bacteria appear to be more sensitive to disinfectants during the initial lag and exponential growth phases [20].

To determine the risk of acquiring a norovirus infection, a QMRA model was used for the two restroom fomite surfaces on which E. coli were detected in the largest numbers (every-3-day cleaning) and for which the Phase 2 intervention (bundling of multiple products for multiple surfaces) had the greatest impact (i.e., toilet seat and vanity countertop). The risk of acquiring a norovirus infection was high for touching both surfaces, being much greater than the 1:10⁻⁶ probability that is often used as a guideline [1]. This suggests that these specific surfaces within a household restroom present the greatest risk of infection from hand contact when a person in the household is infected with norovirus. Infectious norovirus is highly contagious; a patient may shed 10–1000 billion norovirus particles per gram of feces, and ingestion of as low as 1–10 infectious norovirus particles can lead to an infection [1]. A greater than 90% reduction in risk was achieved with the tested products for the vanity countertop and the toilet seat. QMRA modeling indicated that the risk of norovirus infection after touching these sites in the restroom cleaned every 3 days (twice a week) was reduced. Since norovirus can likely survive longer than E. coli on surfaces, the modeled infection risk associated with touching these fomites may be underestimated [1].

Based on the evaluation of the restroom cleaning approaches and the QMRA analysis, the optimal frequency of cleaning/disinfecting of restroom surfaces appears to be every 3 days (approximately twice a week) for minimizing exposure to potential pathogens and maintaining a hygienically clean restroom. To achieve the maximum reduction in pathogen exposure risk from contaminated fomites in the restroom, regular disinfection of all restroom surfaces (e.g., continuous presence of disinfectants in the bowl and/or a disinfectant that leaves residual microbicidal activity on fomites) [21,22] is suggested.

The limitations (uncertainty) of this study include the differences in restroom use between households (frequency of use and age of the individuals), frequency of touching various surfaces in the restroom, and variation in cleaning style. These factors influence the large variability of bacterial numbers detected on surfaces in the households assessed. This has also been noted in other studies on bacterial numbers detected in household restrooms [1,3,10]. In addition, the ratio of E. coli to norovirus numbers used in the QMRA may not always reflect the actual ratio.

5. Conclusions

The bacterial reduction was significantly greater during the Phase 2 cleaning intervention (bundling of multiple products for multiple surfaces) vs. Phase 1 cleaning (use of toilet bowl cleaner only) (p = 0.009) when cleaning on an every-3-day cycle (approximately twice a week). The restroom touch surfaces conferring the greatest risk of transmitting a norovirus infection were found to be the vanity countertop and bottom of the toilet seat. The bundling of multiple products to clean/disinfect surfaces can reduce the predicted risk of acquiring a norovirus infection by greater than 90% when there is contact with the vanity countertop or when lifting the toilet seat. Thus, cleaning restroom surfaces every 3 days (approximately twice a week) offers the greatest reduction in exposure to pathogens and risk of infection from norovirus.