Efficacy of three EPA-registered antimicrobials and steam against two human norovirus surrogates on nylon carpets with two backing types

ABSTRACT

Carpet cleaning guidelines currently do not include the use of an antimicrobial, except after a bodily fluid event. To address this gap, we compared the efficacy of three antimicrobials—two hydrogen peroxide-based (H₂O₂) products (A and B) and one chlorine-based product (C)—and a steam treatment against two norovirus surrogates, specifically feline calicivirus (FCV) and Tulane virus (TuV). These tests were performed on nylon carpets with either water-permeable or waterproof backing types. The effect of repeated antimicrobial use on carpet properties was also evaluated. For a carpet with water-permeable backing, products A, B, and C achieved a 0.8, 3.1, and 0.9 log₁₀ PFU/coupon reduction of FCV and 0.3, 2.5, and 0.4 log₁₀ TCID₅₀/coupon reduction of TuV, respectively, following a 30 min contact time. For carpet with waterproof backing, only product B achieved a 5.0 log₁₀ PFU/coupon reduction of FCV and >3.0 log₁₀ TCID₅₀/coupon reduction of TuV, whereas products A and C achieved a 2.4 and 1.6 log₁₀ PFU/coupon reduction of FCV and a 1.2 and 1.2 log₁₀ TCID₅₀/coupon reduction of TuV, respectively. Steam treatment achieved a ≥ 5.2 log₁₀ PFU/coupon reduction of FCV and a > 3.2 log₁₀ TCID₅₀/coupon reduction of TuV in 15 seconds on the carpet with both backing types. The repeated use of products A and B decreased the tensile strength of the carpet backing, while use of product B resulted in cracks on carpet fibers. Overall, steam treatment for 15 seconds was efficacious on both carpet types, but only product B achieved efficacy after a 30-minute exposure on the carpet with waterproof backing.

IMPORTANCE

Carpets are common in long-term care facilities, despite its potential as a vehicle for transmission of agents associated with healthcare-associated infections, including human norovirus (NoV). Presently, our understanding of carpet disinfection is limited; hence, there are no commercial antimicrobials against norovirus available for use on carpets. Our findings showed that steam treatment, which minimally affected the properties of carpet fibers and backing, was more efficacious against human norovirus surrogates on carpets compared to the three chemical antimicrobials tested. Additionally, the two surrogates were more sensitive to chemical antimicrobials on the carpet with waterproof backing compared to carpets with water-permeable backing. These findings can inform development of antimicrobials for use on carpets contaminated with human norovirus.

INTRODUCTION

In the United States, 75% of norovirus (NoV) outbreaks are linked to long-term care facilities (LTCFs) (1). While noroviruses are commonly spread person-to-person, transmission can also occur when one contacts a contaminated surface (2). Unlike hospitals where surfaces are almost exclusively constructed of non-porous materials, LTCFs commonly use porous materials (e.g., carpet and upholstery) to maintain a home-like atmosphere (3). Epidemiological studies have shown carpets to be a fomite for transmission of NoV, particularly if not properly cleaned after a vomiting/diarrhea event (2, 4, 5). The complex nature of carpets (i.e., fiber composition and backing) creates an environment conducive to NoV attachment and survival (6). Additionally, the act of walking on the carpet can hypothetically resuspend viral particles not adequately removed during cleaning, potentially leading to the contamination of other surfaces (2, 6, 7). Unlike non-porous materials, such as stainless steel, scant evidence is available about the use of antimicrobials on carpet.

Current carpet cleaning guidelines do not include use of an antimicrobial, except after a bodily fluid event (i.e., steam cleaning on the contaminated area for r minutes). Buckley et al. (8) reported that a steam treatment achieved a 3-log₁₀ reduction of two NoV surrogates (murine norovirus and feline calicivirus) within 90 seconds. Of note, this contact time was considered relatively lengthy for routine practice (8). Ideally, contact time should be as short as possible. U.S. Environmental Protection Agency (EPA)-registered antimicrobials, especially ready-to-use products, are extensively used in healthcare settings due to their ease of application. Ready-to-use products reduce the need for oversight and training in their usage. Currently, no EPA-registered chemical antimicrobials that have a claim against NoV are commercially available for use on non-launderable porous materials (e.g., carpet and upholstery) (9).

Carpet type (i.e., fiber composition) has been shown to affect disinfection efficacy (6, 8, 10, 11). Buckley et al. (8) explored the effect carpet fibers have on viral reduction; however, as far as we know, no studies have assessed the effect of backing type on antimicrobial efficacy. Additionally, repeated use of antimicrobials could damage carpet fibers. For example, the oxidation activity of chemical antimicrobials could degrade dyes on carpet fibers, leading to color fading (12). Antimicrobials can also lead to fiber breakage and delamination, whereby fibers become susceptible to mechanical movement, resulting in detachment from the backing, compromising shock absorption (13–15). Additionally, antimicrobials can remain on carpets, possibly forming a sticky layer that attracts dust and microorganisms (8). Lastly, antimicrobials have been shown to influence the physical properties of carpet backing, potentially diminishing the durability and shortening the carpet’s lifespan (16). This underscores the importance of assessing the effect of repeated disinfection when testing disinfection efficacy.

In this study, we tested and compared the antimicrobial efficacy of three EPA-registered antimicrobials against steam treatment. We evaluated their effectiveness against two NoV surrogates—feline calicivirus (FCV), which is a requirement for U.S. EPA product registration, and Tulane virus (TuV). These tests were conducted on nylon carpets with two distinct backing types. We hypothesized that backing type affects distribution of the virus and the efficacy of antimicrobials against NoV surrogates. Additionally, the effect of repeated use of an antimicrobial over time on the properties of carpet fibers and backing was determined.

RESULTS

Stability of H₂O₂ on carpets

Two hydrogen peroxide-based antimicrobials, products A and B, were tested. As hydrogen peroxide is generally stabilized in commercial products by acids at low pH (17), two 1% hydrogen peroxide solutions at pH 5.5 (non-stabilized solution) and 3.0 (stabilized solution) were evaluated. After applying to carpet coupons, the concentration of the 1% H₂O₂ solution at pH 5.5 decreased from 1% to 0.69% after incubating for 1 hour, while the concentration of the 1% H₂O₂ solution at pH 3.0 decreased to 0.76%. No significant difference (P > 0.05) was detected in the concentration of H₂O₂ between non-stabilized and stabilized H₂O₂ solutions after a 1 hour contact time (Fig. 1). The initial concentration of H₂O₂ in product A was measured as 0.6%, slightly above specifications listed on the product labels (Table S1), but it remained within the range of chemical manufacturing standards mandated by law (18). However, the initial concentration of H₂O₂ in product B (0.05% peracetic acid) was 3.9% higher than what was reported on the label, probably due to the presence of peracetic acid, as it is indistinguishable from H₂O₂ when measured using a hydrogen peroxide test kit (19). The concentration of H₂O₂ in product A was relatively stable, decreasing from 0.6% to 0.4% in 30 minutes and to 0.3% in 60 minutes. In contrast, the concentration of H₂O₂ in product B decreased to 1.9% after incubating for 15 minutes and then slowly decreased to 1.1% after incubating for 60 minutes.

Fig 1.

Changes of H₂O₂ concentrations in products A (○), B (□) after 1% H₂O₂ solution (pH 5.5,△) and 1% H₂O₂ solution (pH 3.0, ◊) are applied to the carpet. Error bars represent standard deviations from triplicates.

Efficacy of three antimicrobials on carpet coupons

The effect of desiccation and recovery methods was determined by comparing the virus concentrations in the inoculum and unscrubbed control. Overall, titer reduction of a ≤ 0.2 log₁₀ plaque-forming unit (PFU)/coupon of FCV and a ≤ 0.6 log₁₀ median tissue culture infectious dose (TCID₅₀)/coupon of TuV was observed in all the testing trials (Table S2). Product A (H₂O₂-based) achieved a 0.8 log₁₀ PFU/coupon of FCV reduction and a 0.3 log₁₀ median TCID₅₀/coupon reduction of TuV on the carpet with water-permeable backing (WPerB), while it achieved a 2.4 log₁₀ PFU/coupon reduction of FCV and 1.2 log₁₀ TCID₅₀/coupon reduction of TuV on the carpet with waterproof backing (WProB) after 30 minutes (Fig. 2). Product B (H₂O₂-based) had a 3.1 log₁₀ PFU/coupon reduction of FCV and 2.5 log₁₀ TCID₅₀/coupon reduction of TuV on the carpet with WPerB but achieved a 5.0 log₁₀ PFU/coupon reduction of FCV and >3.0 log₁₀ TCID₅₀/coupon reduction of TuV on the carpet with WProB. In contrast, product C (chlorine-based) yielded a 0.9 and 1.6 log₁₀ PFU/coupon reduction of FCV and 0.4 and 1.2 log₁₀ TCID₅₀/coupon reduction of TuV on carpets with WPerB and WProB, respectively.

Fig 2.

Efficacy of three chemical antimicrobials against FCV and TuV on carpets with water-permeable backing (WPerB) and waterproof backing (WProB). The contact time was 30 minutes. The error bars are standard deviations (SDs) from ten replicates. Stars indicate reaching the detection limit (3.0 log₁₀ TCID₅₀/coupon reduction of TuV).

Using the same samples collected for Fig. 2 data, we conducted an analysis of genome copy reduction among treatments using RT-qPCR. On the carpet with WPerB, products A, B, and C achieved a 0.4, 1.7, and 0.5 log₁₀ genome copy (GC)/coupon reduction of FCV, respectively, and 0.5, 0.3, and 0.3 log₁₀ GC/coupon reduction of TuV, respectively (Fig. 3). On the carpet with WProB, products A and B achieved a 2.0 and 1.8 log₁₀ GC/coupon reduction of FCV and 0.5 and 0.3 log₁₀ GC/coupon reduction of TuV, respectively. However, product C did not significantly affect either FCV or TuV genomes with a 0.2 log₁₀ GC/coupon reduction on the carpet with WProB.

Fig 3.

Reduction of FCV and TuV genome copies on carpets with water-permeable backing (WPerB) and waterproof backing (WProB) treated with chemical antimicrobials. The contact time was 30 minutes. The error bars are standard deviations (SDs) from ten replicates.

Efficacy of steam treatment on carpets

The effect of desiccation and recovery methods resulted in a reduction of ≤0.2 log₁₀ PFU/coupon of FCV and nearly no reduction of TuV in all the testing trials (Table S3). Following steam treatment, no viable FCV and TuV were detected on any coupons, except for one of 10 tested replicates, where a 1.9 log₁₀ PFU/coupon of FCV was detected after 15 seconds of steaming on the carpet with WPerB (Fig. 4). The steam treatment resulted in a reduction of 5.2, >5.3, and >5.3 log₁₀ PFU/coupon of FCV and a reduction of >3.2 log₁₀ TCID₅₀/coupon of TuV on the carpet with WPerB following contact times of 15, 30, and 60 seconds, respectively. On the carpet with WProB, the steam treatment achieved a > 5.3 log₁₀ PFU/coupon reduction of FCV and >3.2 log₁₀ TCID₅₀/coupon reduction of TuV across all contact times.

Fig 4.

Efficacy of steam treatment against FCV and TuV on carpets with water-permeable backing (WPerB) and waterproof backing (WProB). The error bars are standard deviations (SDs) from ten replicates. Stars indicate reaching the detection limit (5.3 log₁₀ PFU/coupon reduction for FCV and 3.2 log₁₀ TCID₅₀/coupon reduction for TuV).

Using the same samples collected for Fig. 4 data, we conducted an analysis of genome copy reduction among different steam treatment times using RT-qPCR. The stream treatment achieved a reduction of 0.0, 0.3, and 1.6 log₁₀ GC/coupon of FCV and 0.4, 0.6, and 1.8 log₁₀ GC/coupon of TuV on the carpet with WPerB after 15, 30, and 60 seconds, respectively (Fig. 5). On the carpet with WProB, the steam treatment resulted in a reduction of 0.0, 0.5, and 0.9 log₁₀ GC/coupon of FCV and a reduction of 0.6, 1.1, and 2.0 log₁₀ GC/coupon of TuV after 15, 30, and 60 seconds, respectively.

Fig 5.

Reduction of FCV and TuV genome copies on carpets with water-permeable backing (WPerB) and waterproof backing (WProB) treated by steam. The error bars are standard deviations (SDs) from ten replicates.

Effect of repeated disinfection on carpet properties

No color change was visible to the naked eye after repeated use of each antimicrobial (30 cycles). The average of ΔE, color difference, of the carpet with WPerB treated with water, steam, or products A, B, and C were 1.34, 1.58, 1.21, 1.08, and 0.74, respectively, while the ΔE of the carpet with WProB treated with water, steam, or products A, B, and C was 0.54, 0.98, 0.77, 0.74, and 0.54, respectively (Table S4). No significant (P > 0.05) difference was detected in ΔE among the cleaning control and all antimicrobial treatments on either carpet with WPerB or WProB. No significant difference in fiber damage was detected between control and treatment groups, as determined using a 40X magnification with a confocal microscope, except for product B (Fig. 6). Product B caused cracks on both types of carpet fibers.

Fig 6.

Carpet fibers of carpets with water-permeable backing (Color Accent) and waterproof backing (Highlight) under 40× magnification with a confocal microscope: (1) untreated carpet, (2) water-treated carpet, (3) steam-treated carpet, (4) product A-treated carpet, (5) product B-treated carpet, and (6) product C-treated carpet. The yellow rectangles indicate cracks on the fibers, and the green rectangles indicate random residues on fibers due to manufacturing. The scale bars (yellow lines) indicate 50 µm.

The tensile strength of carpets with WPerB decreased from 33.81 MPa to 26.59 and 27.36 MPa after treating with water or steam, respectively, while the tensile strength of the carpet with WProB decreased from 41.24 to 35.64 and 37.62 MPa, respectively (Fig. 7). Following treatment with the three antimicrobials, the tensile strength of the carpet with WPerB decreased to 19.51, 19.45, and 26.14 MPa for products A, B and C, respectively, while the tensile strength of the carpet with WProB decreased to 33.43, 32.20, and 34.81 MPa, respectively. No significant differences were detected among cleaning control, steam, and product C on tensile strength of the carpet with WPerB, but products A and B significantly (P < 0.05) decreased tensile strength. In contrast, there was no significant difference between the cleaning control and all antimicrobial treatments on the carpet with WProB.

Fig 7.

Tensile strengths of water-permeable backing (Color Accent) and waterproof backing (Highlight) affected by various treatments. Error bars are standard deviations from six replicates. Different letters (i.e., a and b) on each column of each carpet type indicate significant difference (P < 0.05) among treatments.

DISCUSSION

The efficacy of three EPA-registered antimicrobials and steam treatment was tested against two human NoV surrogate viruses, FCV and TuV, on nylon carpets with two backing types—water-permeable and waterproof. Our findings showed that only one antimicrobial (product B) achieved efficacy against both surrogates on both carpet types after 30 minutes (impractical contact time), while the steam treatment was efficacious against FCV and TuV in only 15 seconds. The effect of carpet backing on antimicrobial efficacy was significant, with FCV and TuV showing greater sensitivity to chemical antimicrobials on the carpet with WProB compared to the carpet with WPerB. Lastly, we demonstrated that repeated use of antimicrobials could adversely affect carpet properties.

Currently, there are 373 EPA-registered antimicrobials appearing on List G—Antimicrobial Products Registered with EPA for Claims Against Norovirus. All claim efficacy for use on non-porous hard surfaces, but none claim efficacy against NoV on carpets (20). Only one of the three antimicrobials tested in this study achieved a > 3-log₁₀ PFU/coupon reduction of FCV on the carpet with WProB with a contact time of 30 minutes, which is much longer than the product claim of 4 minutes for use on non-porous surfaces, such as stainless steel and plastics. The slow action of antimicrobials on the carpet may be attributed to viruses forming more aggregates within carpet fibers and instability of active ingredients on the carpet compared to non-porous materials. Additionally, viruses may be shielded by the hydrophobic nylon fibers from the antimicrobial (21). Conversely, a steam treatment for 15 seconds was more efficacious against FCV and TuV on both carpets with WPerB and WProB. Similarly, Buckley et al. (8) reported that a 90-s steam treatment resulted in a reduction of 3.68 and 3.80 log₁₀ PFU/coupon of FCV on nylon and wool carpet, respectively. One possible explanation for the high efficacy of steam treatment in our study was the thickness (2.92–3.20 mm) of the carpet. Both types are much thinner than the carpet (6.35–10.6 mm) tested in Buckley and colleagues’ study (14), which allowed steam to penetrate fibers more quickly and thoroughly. Based on our results, we concluded that a 15-s steam treatment was efficacious for thinner carpet fibers, but thicker pile carpets may require longer contact times.

Antimicrobials also require a longer contact time on carpets (90 s) than on glass (<10 s) to achieve the same level of efficacy (8). Moreover, a longer contact time is believed to enhance the efficacy (22). Our results suggest that a longer contact time is not only impractical but also ineffective for chemical antimicrobials due to the instability of active ingredients on the carpet (Fig. 1). Active ingredients of antimicrobials (e.g., chlorine and peroxides) were not stable both in suspension and on surfaces during prolonged contact time due to degradation (23). Thus, to better understand the performance of antimicrobials, we assessed the stability of two H₂O₂-based antimicrobials (products A and B) during prolonged contact on carpet coupons. Fig. 1 revealed that the high concentration (3.13%) of H₂O₂ in product B was rapidly lost, but the lower concentration (0.5%–1%) of H₂O₂ (product A and H₂O₂ controls) was reduced slowly, highlighting the importance of stability of active ingredients in antimicrobials.

Compared to infectivity loss, the antimicrobial effect on the genomes of FCV and TuV was limited, indicating RT-qPCR as an improper evaluation method for disinfection efficacy tests. Overall, the log₁₀ reduction in genome copies was less than half of the infectivity loss caused by either chemical antimicrobial or steam treatment. This observation agreed with those of previous studies (8, 24), suggesting viral genomes are protected by capsids. Moreover, this study revealed that TuV genomes were relatively more sensitive to steam while exhibiting more resistance to chemical antimicrobials when compared to FCV genomes. One possible explanation is the presence of specific amino acid sequences in the TuV capsid proteins that confer resistance to oxidation (17). However, the capsid integrity of both FCV and TuV was affected by heat, allowing steam to breach the capsid and degrade viral genomes (25).

We also showed that disinfection efficacy against FCV and TuV differed based on the backing type. Capillary action facilitated antimicrobial absorption and sequestration by carpet fibers (26). However, the carpet fibers tufted in clusters on the backing interface created areas that only had limited contact with the antimicrobial (26). In contrast, significantly more FCV and TuV virions did not penetrate the hydrophobic material of the carpet with WProB backing (Fig. S1). This is probably because the carpet with WPerB facilitated the infiltration of the virus inoculum in the form of large droplets, driven through the primary backing layer by gravity, not through capillary action (27). Thus, despite thorough scrubbing to increase contact between antimicrobials and viruses in fibers, antimicrobial had limited contact with viruses in the WPerB, while WProB allowed more contact between the antimicrobial and viruses. This difference in the distribution of antimicrobials and viral particles on carpets with WPerB and WProB might explain variations in the antimicrobial efficacy (Fig. S1). As a result, FCV and TuV were more sensitive to antimicrobial action on the carpet with WProB than on the carpet with WPerB.

Our results showed that steam treatment and two chemical antimicrobials (products A and C) did not cause any changes in fiber appearance, confirming nylon fibers can withstand repeated exposure to antimicrobials (3). However, product B, which contained 3.13% H₂O₂ and 0.05% peracetic acid, resulted in cracks in the nylon fibers after repeated antimicrobial use (30 cycles). This was attributed to the strong oxidation activity of product B (15, 28). The active ingredients of all three antimicrobials are strong oxidizers, so use, particularly repeated use, can result in color loss in textiles (12, 29, 30). Surprisingly, no significant difference in color change was found across all antimicrobials and deionized water, likely due to the developed dyeing system for nylon fibers (31, 32). Additionally, the dye was also resistant to the heat provided during the steam treatment. Nevertheless, our results on ∆E values also showed that the antimicrobial effect on the carpet with WPerB (black color) was greater than that of the carpet with WProB (navy blue color), suggesting different dyes used in the carpet play a role in color change (Table S4). Specific components of the dye remain unknown. Additional research is needed to understand the mechanism of action behind this difference. Lastly, the damage of fibers caused by the repeated antimicrobial use might enhance the attachment of pathogens (i.e., viruses and formation of biofilms), leading to decreased antimicrobial efficacy. However, it is important to note that this hypothesis has yet to be tested.

In addition to assessing the effect of repeated antimicrobial use on carpet fibers, we investigated the effect of repeated antimicrobial use on the physical properties of carpet fibers and backing type. Consistent with previous research (33), the tensile strength of the carpet with WPerB was influenced by repeated water-wash and steam treatment, whereas the tensile strength of the carpet with WProB due to its hydrophobicity was mildly affected. Interestingly, products A and B significantly (P < 0.05) decreased the tensile strength of the carpet with WPerB compared to other treatments. This can be explained by the presence of hydroxyl radicals in these two antimicrobials, which accelerated oxidation of latex used in WPerB, ultimately leading to changes in its physical properties (13, 34). In contrast, the carpet with WProB exhibited higher strength due to its multilayers of secondary backing. Furthermore, the carpet with WProB also had greater resistance to prolonged exposure to antimicrobials than the carpet with WPerB. The observation suggests potential improvements in the antioxidation property of polymers used in the carpet with WProB (3). However, the specific materials used to manufacture the carpet with WProB remained undisclosed, preventing a direct comparison to the carpet with WPerB. In contrast to previous reports, our results suggested appearance change analysis alone might not be sufficient to evaluate the overall effect of antimicrobial use on carpets (3, 8, 35). Hence, a comprehensive evaluation to include both fibers and backings is essential for evaluating the effect of repeated antimicrobial use on carpets.

Our study has several limitations. While repeated disinfection affected the tensile strength of the backing, particularly noticeable in the carpet with WPerB, further study on the interactions between virus particles and the carpet fiber or backing may shed light on more efficient methods for eluting virus particles, thereby enhancing disinfection efficacy. The effect of other carpet characteristics was not investigated as it was outside the scope of this study. For example, fiber material (i.e., wool, polyester, and nylon) and construction of fiber (i.e., looped and pile cut) have been reported to affect the antimicrobial efficacy (8). Furthermore, these fiber characteristics may also interact with carpet backing in a complex manner (3). Therefore, our findings cannot be extrapolated to other carpet types due to these variations.

In conclusion, we showed that steam treatment is efficacious against NoV surrogates on carpets with the least effect on the properties of carpet fibers and backing. Despite H₂O₂-based antimicrobial product B effectively reducing NoV surrogates on the carpet, it is important to be aware of the excessively long contact time and potential damage to carpets by repeated antimicrobial application. Hence, steam treatment is recommended for carpet disinfection rather than using a chemical antimicrobial, such as after a bodily fluid event. Additionally, the effect of carpet backing on disinfection efficacy should be considered while selecting appropriate antimicrobials. These findings can inform the development of effective strategies for the disinfection of carpets contaminated with NoV.

MATERIALS AND METHODS

Virus propagation and assays

Crandell–Rees Feline Kidney (CRFK) cells (ATCC CCL-94) were cultured in Eagle’s modified essential medium (MEM; Gibco Life Technologies, NY, USA) containing 5% low-endotoxin heat-inactivated fetal bovine serum (FBS) (Seradigm, VWR International, Radnor, PA, USA), 100 U/L penicillin (HyClone, GE, MA, USA), and 100 mg/L streptomycin (HyClone). Ninety percent (90%) of confluent monolayers of CRFK cells Head2 were infected with the FCV strain F9 (ATCC VR-782; American Type Culture Collection, VA, USA) at a multiplicity of infection (MOI) of 0.01 and held at 37°C for 2 days. FCV was then harvested from cell lysates by three freeze–thaw cycles followed by centrifugation for 10 min at 5,000 × g and 4°C. FCV stocks at ca. 10⁸ PFU/L were aliquoted and stored at −80°C. Infectious FCV was quantified by standard plaque assay, as previously described (8). To test cell line permissiveness and contamination, FCV and PBS served as a positive and negative control, respectively. CRFK cells were passaged less than 30 times.

LLC-MK2 cells (ATCC CCL-7) were cultured in Opti-MEM I reduced serum medium (Gibco Life Technologies) supplemented with 2% low-endotoxin heat-inactivated FBS, 100 U/L penicillin, and 100 mg/L streptomycin. Ninety percent (90%) confluent monolayers of LLC-MK2 cells were infected with TuV, kindly provided by Dr. Jason Jiang (Cincinnati Children’s Hospital, OH, USA), at an MOI of 0.1 and held at 37°C for 2 days. TuV was harvested from cell lysates similar to FCV harvesting. Infectious TuV titer was quantified by the TCID₅₀ assay as described, with modifications (36). Twenty microliters (20 mL) of serially diluted viruses was added in each quantification well per column and eight wells used for each dilution. TuV stocks at ca. 10⁷ TCID₅₀ l⁻¹ were aliquoted and stored at −80°C. LLC-MK2 cells were passaged fewer than 30 times.

RNA extraction and qRT-PCR

Viral RNA extraction was performed as previously described (8). Briefly, the RNAs of FCV and TuV were extracted from 0.15 mL of samples using an ENZA viral RNA kit (Omega Bio-Tek, GA, USA) as per the manufacturer’s guidelines. After extraction, RNA was stored at −80°C before further analysis. Quantitative reverse transcription PCR (qRT-PCR) was performed for FCV and TuV separately to determine the loss of viral genome copies using the Platinum SYBR Green PCR kit (Invitrogen, CA, USA). The forward and reverse primer sequences for FCV qRT-PCR analysis were GCCATTCAGCATGTGGTAGTAACC and GCACATCATATGCGGCTCTG, respectively, and for TuV analysis were TTCACCCGACCAACCCTG and ACGCCCCAACGCACCTA, respectively (8, 37). The standard curves were prepared individually for FCV and TuV by 7-step tenfold dilutions of virus stocks. Melting curves were measured for the quality control of qRT-PCR assay.

Selection of antimicrobials and preparation of carpet coupons

Two H₂O₂-based (products A and B) and one chlorine-based (product C) antimicrobials were tested (Table S1) (17), as was a steam treatment (Ladybug 2300, Advap, WA, USA). Efficacy tests were performed on two similar nylon loop pile carpets—Color Accent carpet (Shaw Inc., GA, USA) and Highlight carpet (Shaw Inc., GA, USA)—using the Carpet and Rug Institute Test Method 112 (38). Color Accent carpet had a WPerB, whereas Highlight carpet had a WProB (Table 1). Although both types of carpets had anti-soil coatings, each was confirmed free of virucidal activities against FCV and TuV. Carpet samples were cut into 5 × 5 cm² coupons (Interface Inc., GA, USA) and then dusted by a gloved hand to remove loose fibers. Before testing, carpet coupons were autoclaved on a 20-min dry cycle and cooled at room temperature overnight.

TABLE 1.

Characteristics of selected carpets

Specifications	Carpet with water-permeable backing (WPerB)	Carpet with waterproof backing (WProB)
Fiber (construction method)	Nylon 6 (loop)	Nylon 6 (loop)
Fiber finishing	Solution-dyed, soil-resistant coating	Solution-dyed, soil-resistant coating
Average density	0.348 g/cm3	0.317 g/cm3
Finished pile thickness	2.92 mm	3.20 mm
Primary backing	Synthetic	Synthetic
Secondary backing	Stalok	Ecoworx performance broadloom
Backing to moisture	Water-permeable	Waterproof

Determination of H₂O₂ stability on carpets

Preliminary results showed the three antimicrobials did not achieve a 3-log₁₀ reduction of FCV and TuV within 10 minutes, so longer contact times were tested. The concentration of hydrogen peroxide (H₂O₂) was monitored for products A and B to determine the optimal contact time to achieve maximum efficacy of these two products. Diluted 1% H₂O₂ solutions at pH 3.0 and 5.5 were also tested to evaluate the effect of stabilization of H₂O₂. Each carpet coupon was treated with 6 mL of each antimicrobial product. Each coupon was scrubbed clockwise and counterclockwise for 30 seconds with an antimicrobial-saturated (approximately 1 mL) surgical scrub brush (BD E-Z Scrub, BD, NJ, USA), as previously described (8). After 1-, 5-, 10-, 15-, 30-, and 60-min contact times, coupons were immediately transferred into a flask with 100 mL water to elute H₂O₂. The concentration of H₂O₂ was measured using a hydrogen peroxide test kit (HYP-1, Hach, IL, USA) within 10 minutes after samples were collected as per the user manual.

Quantitative carpet test

As no standard test method is available for carpets, the efficacy of treatment with three antimicrobials and steam was evaluated as previously described, with modifications (8). Briefly, all coupons were inoculated with a mixture of FCV and TuV at ca. 7 log₁₀ PFU or TCID₅₀ per coupon, respectively, and then dried within 1 hour at room temperature and a humidity level between 30% and 50%. Next, 6 mL of one of three antimicrobials was applied and then held for 30 minutes, or a steam cleaner head wrapped with sterile terry cloth was put over coupons for 15 , 30 , and 60 seconds. Two sets of controls, unscrubbed and scrubbed, were also used. Unscrubbed controls were enumerated immediately after drying, whereas scrubbed control samples were either scrubbed with a phosphate-buffered saline-saturated surgical scrubber (BD E-Z Scrub, BD, NJ, USA) or by using a cool steam cleaner head. After defined contact times, the treatment and control samples were neutralized using 100 mL of appropriate neutralizer plus 0.02% Tween-80. Inocula on coupons were recovered by ultrasonicating for 1 minute at 40 kHz and stomaching at 200 rpm in a stomacher (Model 400, Seward, NY, USA) for 3 minutes and titrated using a plaque assay for FCV and TCID₅₀ assay for TuV. The viral genome reduction following treatment (i.e., three antimicrobials or a steam treatment) was measured using the RT-qPCR method described above.

Distribution of viruses in carpet backing

To explore the distribution of viruses in the carpets, all coupons were inoculated with a mixture of FCV and TuV at ca. 7 log10 PFU or TCID50 per coupon, respectively, and dried within 1 hour at room temperature and a humidity level between 30% and 50%. Next, the fibers of coupons were shaved off with sterile scalpels only to have the backing left. The inocula in carpet backings were then recovered with 100 mL of PBS plus 0.2% Tween-80 and titrated as described above.

Effect of repeated disinfection on carpet properties

To simulate repeated disinfection, carpet coupons (5 × 12.5 cm²) with two backing types, WPerB and WProB, were immersed in 500 mL of tested antimicrobials at room temperature for 15 hours or treated with steam for 450 seconds, equal to 30 cycles of disinfection treatments. The carpet immersed in water was used as a cleaning control to simulate water-wash of carpet. After treatment, coupons were rinsed with DI water until no residue was observed and then dried completely at room temperature for at least 4 hours before determination of color and tensile strength.

A colorimeter (CR-400, Konica Minolta, NJ, USA) was used to determine coupon color, in accordance with ASTM E2828-20 with modifications (39). Colors were read in three positions on each treatment and control coupon. ∆E values, which indicates difference in total color and brightness between untreated and the cleaning controls or treated carpet coupons, were calculated as follows:

$${\displaystyle \mathrm{\Delta }\mathrm{E}=\sqrt{({\mathrm{L}}_1-{\mathrm{L}}_0{)}^2+({\mathrm{a}}_1-{\mathrm{a}}_0{)}^2+({\mathrm{b}}_1-{\mathrm{b}}_0{)}^2}}$$

where L₀, a₀, and b₀ were color parameters of the untreated carpet and L₁, a₁, and b₁ were color parameters of either cleaning controls or treated coupons. Additionally, fiber damage was observed under a confocal microscope (LEXT OLS4100, Olympus, PA, USA), as reported previously (40). Briefly, fibers of each untreated, cleaning control, and treated coupons were observed under a 40X lens. Images of untreated, cleaning control, and treated coupons were captured without modifications.

To evaluate backing damage due to repeated disinfection, the tensile strength of treated carpet coupons was determined using the ASTM D5034-21 method (41). Briefly, the test involved clamping a coupon (2.5 × 12.5 cm²), either untreated or repeatedly disinfected, in an Instron Universal Testing Machine (Model 4201, Instron Corp, MA, USA). The clamps, initially positioned 5.1 cm apart, were moved at a speed of 30.5 cm/min to stretch the coupon until it broke. The recorded breaking force (MPa) represented the tensile strength.

Statistical analysis

Log₁₀ reductions were calculated by log₁₀ (N₀/N_d), where N_d is the average microbial population from treatment samples and N₀ is the average microbial population from each control sample. Statistical analysis was performed using a one-way multiple-comparison t-test to determine the relationship between contact time and log₁₀ reduction. All results were expressed as mean ± standard deviation. Statistical significance was defined as a P-value of <0.05 to establish a conservative estimate. Statistical analyses were conducted using GraphPad Prism 6.01 (GraphPad Software, Inc., CA, USA).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00384-24.

Figure S1. aem.00384-24-s0001.tiff.

Distribution of FCV and TuV in carpet.

Supplemental tables. aem.00384-24-s0002.docx.

Tables S1 to S4.

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.