Human noroviruses (hNoV) are the leading cause of acute gastroenteritis worldwide (Pires et al., 2015), causing an estimated 684 million cases annually. The illness burden caused by hNoV represents a considerable cost to society, with recent estimates suggesting over $60 billion in societal impact globally, primarily in the form of loss of productivity due to illness (Bartsch et al., 2016). Because they can be transmitted readily via contaminated food, hNoV are also the leading cause of foodborne disease globally (Lopman et al., 2012; Pires et al., 2015), causing an estimated 5.5 million cases annually in the United States alone (Scallan et al., 2011). Immunity to hNoV is believed to be short-lived (Simmons et al., 2013) and while several vaccines are currently in development, none are yet widely available (Mattison et al., 2018). Thus, preventing transmission of hNoV relies on practices such as exclusion of sick individuals from workplaces and public settings, and adherence to sanitation and hygiene best practices.

A systematic review of outbreaks attributed to hNoV between January 2003 and July 2017 showed that food handlers contribute significantly to disease burden, with restaurants being the most common setting for hNoV outbreaks (Hardstaff et al., 2018). In particular, ready-to-eat foods, or those subjected to extensive human handling immediately preceding consumption, are common causes of outbreaks (Hall et al., 2014). In a recent study, it was reported that 53% of hNoV foodborne outbreaks are associated with poor personal hygiene of infected food handlers (Hall et al., 2012). In addition, enteric virus contamination of food could occur via the hands of pickers during manual harvesting such as, for instance, soft red fruits (Li et al., 2015) and green bell peppers (León-Félix et al., 2010). Collectively, it is clear that hands of infected food handlers are a major route of hNoV contamination of foods.

Washing hands with soap and water is universally accepted as an important hygiene measure for managing transmission of a variety of pathogens (Huang et al., 2014). In retail food settings in the U.S. (such as restaurants and grocery stores), handwashing with soap and water is considered the “gold standard” for performing hand hygiene. The U.S. Food and Drug Administration’s (FDA) Model Food Code (U.S. Food and Drug Administration, 2017), which serves as a regulatory framework for retail food handling across the country, provides strict guidelines related to handwashing for food handlers. Compliance with these guidelines in food handling environments remains extremely low. For example, a recent study found that 60–80% of foodservice establishments were out-of-compliance for employee handwashing (Verrill et al., 2021). Given the historical low compliance metrics with handwashing, and the fact that handwashing is the only permitted option for performing hand hygiene in retail food settings, there has been an interest in developing and evaluating alternative hand hygiene measures, including hand sanitizers (Allwood et al., 2016; Boyce and Schaffner, 2020). Taken together, there is a clear need for fast, convenient, and effective hand hygiene treatments for hNoV control on human hands in food retail, processing, and harvesting environments.

While not considered a replacement for proper handwashing for food handlers, hand sanitizers have long been recognized as an effective means of performing hand hygiene in many settings. The U.S. Centers for Disease Control and Prevention (CDC) recommends the use of hand sanitizers, especially those containing at least 60% alcohol, as an acceptable method of hand hygiene when soap and water are not readily available (U.S. Centers for Disease Control and Prevention, 2021). Hand sanitizers have several benefits relative to effective hand washing, including speed of use (Hilburn et al., 2003), convenience, skin mildness (Boyce, 2000; Boyce et al., 2000; Mukherjee et al., 2018) and broad-spectrum efficacy (for well-formulated products) (Macinga et al., 2008; Edmonds et al., 2012). Commercial hand sanitizers contain active ingredients intended to destroy or otherwise inactivate pathogens. The most common active ingredients are alcohol (either ethanol or isopropanol) and BAK (Macinga et al., 2008; Edmonds et al., 2012). While the active ingredient(s) is important to the efficacy of a hand sanitizer, overall product formulation is equally important, as illustrated by the fact that products with the same active ingredients have been shown to have vastly variable antimicrobial efficacy against a variety of viral and bacterial pathogens (Macinga et al., 2008).

While hand sanitizers may play an important role in controlling transmission of hNoV, their efficacies against this virus are poorly characterized in vivo. The purpose of this work was to characterize the efficacy of eight commercially available hand sanitizers, as well as a benchmark 60% ethanol solution, against GII.4 hNoV using an in vivo fingerpad method (ASTM E1838-17).

Product efficacies ranged from less than 0.5 log₁₀ hNoV GEC reduction to 3.3 log₁₀ hNoV GEC reduction (Figure 2). For the 30 s exposure time evaluated in this study, Product A [85% ethanol (vol/vol)], was the most efficacious (p < 0.05), with a log₁₀ hNoV GEC reduction of 3.3 ± 0.3. Product B [70% ethanol (vol/vol)] was the only additional product found to be more efficacious than the 60% ethanol benchmark (p < 0.05) with a log₁₀ hNoV GEC reduction of 2.4 ± 0.4. The performance of Products C through F, with log₁₀ hNoV GEC reductions in the range of 1.7–2.0, differed significantly from Product A (p < 0.05) but not when compared to Product B or the 60% ethanol control (p > 0.05). The performance of Product G, with a log₁₀ hNoV GEC reduction of 1.3 ± 0.6 was not significantly different from that of the 60% control (p > 0.05), but this product had a lower performance when compared to Products A and B (p < 0.05). Comparatively, Product H performed significantly worse than all other products included in the study, including the 60% ethanol control (p < 0.05), with a log₁₀ hNoV GEC reduction of 0.3 ± 0.2.

Three individuals participated in the evaluation of all nine treatments included in this study. Results for Volunteers #2 and #3 largely followed the trends of the overall data (Figure 2). However, results for Volunteer #1 were often skewed from the general trends. For example, Product B produced a log₁₀ hNoV GEC reduction of 2.0 on Volunteer #1 as compared to a mean log₁₀ hNoV GEC reduction of 2.4 ± 0.4. Conversely, Product F produced a 2.5 log₁₀ hNoV GEC reduction on Volunteer #1, with a mean log₁₀ hNoV GEC reduction of 1.7 ± 0.5.

Hand hygiene is an important means by which to reduce transmission of hNoV, the leading cause of acute viral gastroenteritis and also foodborne disease (Scallan et al., 2011; Hall et al., 2013). While proper hand washing with soap and water is considered the gold standard hand hygiene intervention, it has been shown that compliance with handwashing requirements in certain food handling settings, such as restaurants, is low (Allwood et al., 2016; Boyce and Schaffner, 2020). Although hand sanitizers are not a substitute for handwashing by food handlers in the retail food sector, they may be used by food handlers after performing a handwash, which has been shown to significantly improve the overall efficacy of the hand hygiene treatment (Edmonds et al., 2016). Additionally, they may be used in the lobby, entrance, dining area, or restrooms (which can serve as a virus reservoir) of a retail food environment to help disrupt the environmental transmission of hNoV caused by infected patrons and guests.

The most common active ingredient in commercial hand sanitizers is ethanol, which has been shown to be very effective against bacteria (Davis et al., 2006) and enveloped viruses (Kampf, 2018). For example, a concentration of 42.6% ethanol (wt/wt) for a contact time of 30 s has been shown to be effective against SARS coronavirus-2, MERS coronavirus, and influenza A viruses (Kampf, 2018). On fingerpads, alcohol-based hand sanitizers (ABHSs) have been shown to be effective against bacteria including Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli, Serratia marcescens, and Acinetobacter baumannii (Rotter, 1984; Ayliffe et al., 1988; Cardoso et al., 1999; Paulson et al., 1999). While ethanol as an active ingredient may be effective against certain bacteria and viruses at relatively low concentrations, it is worth mentioning that the U.S. FDA, which regulates the manufacture and sale of hand sanitizers as over-the-counter (OTC) drugs, requires ethanol-based hand sanitizers to be formulated with no less than 60% ethanol (vol/vol) (U.S. Food and Drug Administration, 2020).

Another common active ingredient in hand sanitizers are quaternary ammonium compounds (QACs). While QACs are effective against bacteria such as S. aureus, P. aeruginosa, and E. coli, among others (Ayliffe et al., 1988; Bondurant et al., 2020; Aodah et al., 2021), their efficacy against non-enveloped viruses such as poliovirus, murine norovirus, and hNoV appear be relatively poor (Feliciano et al., 2012; Tuladhar et al., 2012; Tung et al., 2013; Ha et al., 2016; Lee et al., 2021). The relative inefficacy of Product H (Figure 2), the only product in our study formulated with a QAC as an active ingredient, provides further support of the inability of this active ingredient to inactivate hNoV.

Evaluating sanitizer efficacy against hNoV is complicated by access to relevant virus strains and the general absence of an affordable and simple in vitro cultivation system. For these reasons, most previous studies characterizing the anti-hNoV efficacy of hand sanitizers include only in vitro suspension assays, or in vivo fingerpad studies using cultivable surrogates such as feline calicivirus (FCV), murine norovirus (MNV) or bacteriophages (MS2) (Macinga et al., 2008). Many of these surrogates behave differently as compared to hNoV when exposed to different chemicals or product formulations. For example, FCV is a respiratory pathogen and is less acid tolerant than is hNoV; MNV is more sensitive to alcohol than is hNoV; and bacteriophage MS2 is highly resistant to ethanol (Cromeans et al., 2014). Those few studies done with hNoV show a general trend of poor efficacy for both alcohol-based products (Liu et al., 2009, 2011; Tuladhar et al., 2015) as well as QACs (Girard et al., 2010; Tung et al., 2013). However, with careful formulation, a few recent alcohol-based products have shown comparatively better efficacy against hNoV compared to earlier formulations (Liu et al., 2011; Escudero-Abarca et al., 2020).

The purpose of this study was to compare the efficacy of seven commercial alcohol-based hand sanitizers, with varying concentrations of ethanol (labeled in the range of 62–85%), one commercial hand sanitizer containing the QAC BAK (0.1% BAK), and a 60% ethanol solution used as a benchmark control, on human fingerpads against hNoV using the in vivo assay ASTM E1838-17. Previous studies have utilized earlier versions of the ASTM E1838 fingerpad method (ASTM E1838-10 or earlier) to evaluate the efficacy of hand sanitizers (Macinga et al., 2008; Liu et al., 2009, 2011). The major differences between the two protocols are (1) the addition of friction (rubbing) during the sanitizer application step for the newer protocol; and (2) the use of an alternative elution method (Petri dish vs. cryovial) (Figure 1). The justification for relevance of the newer (2017) method is that the addition of the rubbing step is more representative of sanitizer use in real world settings and increases the degree of exposure of the product to the inoculum.

The inactivation of hNoV in suspension by simple ethanol solutions has been studied extensively, with concentrations up to 90% (vol/vol) failing to show significant reductions in hNoV, usually less than 0.5 log₁₀ hNoV GEC (Tung et al., 2013), although there are strain-to-strain differences in product efficacy (Park et al., 2016). In our study, all alcohol-based hand sanitizers demonstrated over 0.5 log₁₀ hNoV GEC reduction, most approaching 2 log₁₀ hNoV GEC reduction. In this case, alcohol content alone did not necessarily dictate efficacy against hNoV. Converting wt/wt to vol/vol for comparative purposes (Table 1), for example, Product F, which had an ethanol content of 85% (vol/vol), showed a 1.7 ± 0.5 log₁₀ hNoV GEC reduction. This was statistically less of a reduction as compared to Product A, despite both products having similar ethanol content. Additionally, the efficacy of Product F was not statistically different than Products B, C, D, E, and G, despite having a higher ethanol content [85% (vol/vol) vs. 68–70% (vol/vol), respectively], These data are illustrative of the concept that hand sanitizer efficacy is a function of both active ingredient concentration and product formulation. Indeed, many factors related to product formulation (e.g., inactive ingredients such as skin conditioners and thickeners) impact the overall efficacy of the final product (Macinga et al., 2008; Edmonds et al., 2012). Various compounds, including mixtures of alcohols, metals such as copper and silver, or citric acid, may produce increases in the antiviral activity of ethanol, or act with ethanol in a complimentary or synergistic manner (Macinga et al., 2008; Escudero-Abarca et al., 2020).

Product A statistically outperformed all other products tested (p < 0.05), demonstrating a greater than 3 log₁₀ hNoV GEC reduction. While the exact mechanism is unknown, it is likely (as mentioned above) due to the total product formulation boosting the effectiveness of the ethanol as the active ingredient. Product A was the only product specifically designed with an alkaline pH (8.8–10.3). Product B was specifically designed with an acidic pH (3.8–5.2) and the remaining products included in this study fell in a more neutral pH range (5.0–9.0) (Table 1). It has been previously shown that alcohol-based hand sanitizer efficacy against hNoV can be enhanced by adjusting the product’s pH to either acidic or alkaline conditions (Sato et al., 2020), presumably due to the pH extremes altering the capsid morphology in a way that exposes amino acid residues, allowing them to be more vulnerable to active ingredients than at pH neutral conditions. Even with relatively high efficacy, this product did not completely eliminate detectable hNoV from the fingerpads of volunteers (maximum assay resolution 5.0 log₁₀ hNoV GEC reduction). Whether this is an artifact of using RNase-RT-qPCR as the virus quantification method, or is truly associated with incomplete virus inactivation, remains unknown. While it would have been interesting to additionally evaluate a “low level” of initial contamination of fingerpads to determine if inactivation trends remained similar to those observed for the “high level” of fingerpad contamination evaluated in this study, the loss of assay resolution using a lower starting inoculum would have made evaluating these treatments much more difficult, and made it more challenging to measure statistically significant differences. While a successful hNoV cell culture model has been described recently (Ettayebi et al., 2016; Costantini et al., 2018), its routine use is limited by high cost, complexity, sensitivity to cytotoxicity, a limited number of cultivable strains, and the inability to produce quantitative results in the form of log₁₀ reduction in infectious virus. Nonetheless, this model was recently used to validate the efficacy of Product A against hNoV (Escudero-Abarca et al., 2020), providing data complementary to that produced by RNase-RT-qPCR. In that study, hNoV infectivity was abolished in the cell culture model following exposure to Product A in suspension for 60 s. When evaluated by RNase-RT-qPCR, the same study demonstrated a log₁₀ reduction in hNoV GEC of 2.3–3.2 when exposed to Product A (with and without additional soil load) for 30–60 s in suspension. This was compared to the 0.6–0.9 log₁₀ reduction in hNoV GEC following exposure to the 60% ethanol benchmark under the same treatment conditions (Escudero-Abarca et al., 2020). Interestingly, the 3.3 log₁₀ reduction of hNoV on fingerpads by Product A in the current study is almost identical to the 3.2 log₁₀ reduction observed in the previous study for a 30 s exposure in suspension without additional soil (Escudero-Abarca et al., 2020), demonstrating that the product seemingly does retains efficacy on human fingerpads.

A unique aspect of this study was that three of the ten volunteers participated in the evaluation of all nine hand sanitizer treatments, making it possible to compare efficacy of all the products amongst these three volunteers (Figure 2). For some products, log₁₀ hNoV GEC reduction was quite similar between the volunteers, for others, there was up to a 2 log₁₀ difference in calculated hNoV GEC reduction between volunteers. In addition, the overall trends in product efficacy largely stayed consistent for volunteers #2 and #3, however, for volunteer #1, results seemed to skew from the trends observed for the group as a whole. In other words, the products did not consistently perform better on one volunteer over another, suggesting that differences in skin properties such as skin hydration level, skin pH and/or skin microbiome (Mukherjee et al., 2018) may influence the efficacy of different product formulations. Similar to data observed using in vivo fingerpad methods and bacteria, our results support the fact that subjects are a large source of variability when the same methods are applied to evaluating anti-hNoV activity in vivo (Rotter, 1984; Suchomel et al., 2018). In order to control for this effect, studies should be appropriately sized and, ideally, be cross-over designs where all subjects are evaluated with all treatments, and their individual results can be compared. However, it is also recognized that these criteria may be difficult to meet using a population of human volunteers.

In conclusion, the results of this study support the long-held belief that not all hand sanitizers have anti-hNoV efficacy, and those that do may not completely eliminate the virus from fingerpads. It is not understood how the reductions of hNoV by hand sanitizer on fingerpads compares to reductions by handwashing, the current recommended practice. Further studies investigating the individual and combined effects of handwashing and hand sanitizer use on the removal and inactivation of hNoV on human hands are warranted, particularly studies that capitalize on hNoV cultivation as new culture methods are refined. There remains a need for more hand sanitizer formulations having greater activity against hNoV, a virus that is comparatively recalcitrant relative to other pathogens of concern in community, healthcare, and food preparation environments.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the North Carolina State University Institutional Review Board. The patients/participants provided their written informed consent to participate in this study.

BE-A performed laboratory-based experiments and data collection. RG, L-AJ, RL, and CM performed data analysis. BE-A prepared the first draft of the manuscript, while RG, L-AJ, RL, CM, KG, and JA reviewed and revised prior to submission. All authors contributed to project conception and design of experiments. All authors contributed to the article and approved the submitted version.

RL, CM, KG, and JA were full time employees of GOJO Industries as scientists. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.