Comparison of electric hand dryers and paper towels for hand hygiene: a critical review of the literature

Summary Numerous studies are published on the benefits of electric hand dryers vs paper towels (PT) for drying hands after washing. Data are conflicting and lacking key variables needed to assess infection risks. We provide a rapid scoping review on hand‐drying methods relative to hygiene and health risks. Controlled vocabulary terms and keywords were used to search PubMed (1946–2018) and Embase (1947–2018). Multiple researchers independently screened abstracts for relevance using predetermined criteria and created a quality assessment scoring system for relative study comparisons. Of 293 papers, 23 were included in the final analysis. Five studies did not compare multiple methods; however, 2 generally favoured electric dryers (ED); 7 preferred PT; and 9 had mixed or statistically insignificant results (among these, 3 contained scenarios favourable to ED, 4 had results supporting PT, and the remaining studies had broadly conflicting results). Results were mixed among and within studies and many lacked consistent design or statistical analysis. The breadth of data does not favour one method as being more hygienic. However, some authors extended generalizable recommendations without sufficient scientific evidence. The use of tools in quantitative microbial risk assessment is suggested to evaluate health exposure potentials and risks relative to hand‐drying methods. We found no data to support any human health claims associated with hand‐drying methods. Inconclusive and conflicting results represent data gaps preventing the advancement of hand‐drying policy or practice recommendations.


Introduction
The term hygiene has a complex set of meanings but is generally associated with 'conditions and practices that help to maintain health and prevent the spread of disease' (World Health Organization n.d.). Effective hand hygiene is well recognized as a primary defence against communicable diseases in both clinical and nonclinical environments (Bolon 2016;Zivich et al. 2018). Agencies such as the Centers for Disease Control and Prevention (CDC), World Health Organization (WHO) and US Food and Drug Administration (FDA) have published extensive guidelines on recommended wash times and protocols for hand hygiene practices in medical, food preparation and other environments (Boyce and Pittet 2002;World Health Organization 2009;FDA 2017). CDC guidelines suggest scrubbing hands for at least 20 s with soap and water, which has been shown to reduce approximately 1Á5 log CFU per hand (Jensen et al. 2015). A 30 s wash resulted in a 2Á42 log CFU reduction in another study (Todd et al. 2010).
In addition to handwashing, effective hand drying is important in reducing the level of bacterial contamination present on hands and available for transfer from skin to other surfaces, including skin, food and fomites. Several studies have shown more efficient transfer of bacteria to surfaces from wet compared to dry hands or in highhumidity environments (Patrick et al. 1997;Todd et al. 2010;Lopez et al. 2013;Gerba et al. 2016;Kimmitt and Redway 2016). Patrick et al. (1997) found that hand drying reduced transfer by up to 99Á8%.
Numerous studies have been published with the objective to evaluate the 'best' method for hand drying, with the two most common categories being electric hand dryers and towels. Electric hand dryers include both warm air and jet air designs, while towel drying includes disposable paper and cloth products. Some studies also evaluated evaporation or wiping of hands on clothing as a means of drying hands.
Variables considered include bacterial removal efficacy, environmental contamination potentials, ecological or cost benefits, noise and overall hygiene comparisons. Reports in the literature are mixed in terms of which method is more hygienic and consensus has not been found across studies as to which method is the most preferred. However, publications in both the peer-reviewed and grey literature have made recommendations relative to hygiene and health concerns generally favouring disposable towels over dryers despite a lack of evidence or experiments to appropriately evaluate health outcomes. Most studies have been sponsored by industries with potentially biased interests and have been conducted under a wide range of scenarios with inconsistent study design variables, making it difficult to compare study results.
The objective of this paper is to provide a scoping review of the current scientific peer-reviewed publications comparing the use of paper towels (PT) vs electric dryers (ED) and their impact on environmental health using a standardized set of criteria. Our focus was to categorize, and quantitatively prioritize, the papers relative to scientific rigor in the study design and compare statistically significant results across defined categories of hand-drying methods. We also examined statistically significant results relative to paper conclusions and the impact of these studies on public perception (as determined by media posts and other grey literature) and provided recommendations for future studies.

Materials and methods
Scoping reviews are a preferred approach to explore emerging research areas, allowing for a broader view of the depth and breadth of available literature (Grant and Booth 2009). A scoping review typically does not utilize a formal quality assessment in the initial screening, allowing for characterization of the quantity and quality of the literature as part of the data extraction process.
In order to determine next steps towards improved exposure and risk assessments relative to hand hygiene practices, we were interested in a rapid (<6 month) assessment of the current body of literature aimed at identifying gaps in existing research (Khalil et al. 2016). This approach allowed us to consider a broad range of publication types and furnish as comprehensive an overview of the topic as time allowed. We followed a previously published methodology consisting of five steps: (i) identifying the research question; (ii) identifying relevant studies; (iii) selecting studies; (iv) charting the data; and (v) collating, summarizing and reporting the results (Arksey and O'Malley 2005; Levac et al. 2010).

Identifying the research question
Data relative to the most effective methods for hand hygiene (i.e. handwashing and drying) are highly varied. Many studies have evaluated the efficacy of various soap products and formulation, water temperature, duration and frequency of washing, as well as drying methods. After reading several conflicting reports from the grey literature (i.e. media reports and internet blogs) (Astor 2018;Dean 2018), with some recommending avoidance of air dryers for certain environments (Hafner 2018), we sought to conduct a scoping review of the peer-reviewed literature to answer the research question: Are hand dryers more hygienic than paper towels? Given the implication that improved hygiene is correlated with health, we also sought to answer the question: Are PT safer than hand dryers relative to human infection risks?

Identifying relevant studies
Controlled terms (e.g. MeSH and Emtree) and keywords were used to search PubMed (1946PubMed ( -2018 and Embase (1947Embase ( -2018 (Table 1). Literature searches were completed on 23 October 2018. No limits were applied to the searches (e.g. publication date, language). Relevant items (articles/reports) identified through Google Scholar were also included for screening. Articles published after the initial search date and up to February 2019 were included if they met the inclusion criteria.
Discovered papers were uploaded into Covidence software (Covidence.org), which provided a standard platform for organizing, screening and de-duping articles. Remaining articles were randomly assigned and reviewed by at least two of our three-member team (KAR, AN and JDS). Two authors (KAR and JDS) independently screened full-text articles for relevance. In cases of disagreement, the third author (AN) discussed the criteria with each reviewer and made a final inclusion decision.
Full-text articles and reports were selected if they met the following inclusion criteria: (i) involved quantitative assessments; (ii) utilized sampling for environmental microbes or tracers; (iii) evaluated one or more hand-drying methods; and (iv) published in peer-reviewed literature. Case studies, reviews, systematic reviews and opinion articles were excluded from the quantitative synthesis but may have been included in the summary of grey literature. Articles were not excluded due to poor quality but rather a process of assessment was utilized during data extraction to detail and compare study design and analytical rigor with reported conclusions to identify gaps and future research needs.

Extraction of results
Results were extracted from individual papers using a standardized questionnaire formatted using the Qualtrics online survey software (Qualtrics.com). Extracted data included: author, year, study design, drying method, tracer, environment, sampling site, results, study duration, major findings, strengths, limitations, country, funding source and other notes/comments.

Search results
The initial search returned 287 studies. Further review led to the identification of 6 more papers of interest, for a total of 293 studies. Of the 229 articles that remained after duplicates were removed, 191 were excluded after title and abstract screening because of irrelevance to the topic. Inclusion criteria, as outlined above, were applied to the full text of 38 articles. Of these, 23 met the full set of criteria and were included in the final review.

Major study findings
The majority of studies selected for final data extraction included a comparison of multiple drying methods to determine which drying method was the 'best'. Table 2 provides detailed information on citations, funding, study design and major findings of extracted papers. In most instances, the chosen best method resulted in comparatively less contamination of the environment (surfaces, air, clothing, hands). Papers targeting economic or ecological impacts of drying methods were excluded from the formal data extraction.
Included articles either evaluated one hand-drying method or compared the use of hand towels (paper or cloth) with ED (jet or warm air). Studies frequently utilized laboratory simulations targeting indigenous microbiomes of the environment and hands, or microbial tracers, to determine which hand-drying method correlated with the least amount of hand, air or surface contamination. Two studies used nonmicrobial aerosol tracers (i.e. acid droplets or paint) to assess travel distances following specific drying behaviours and one study used a risk model approach to estimate probabilities of hand contamination.
A greater number of studies favoured the use of PT over air dryers based on comparisons of bacterial count trends on surfaces or in the air. Some studies found higher counts on PT dispensers. Even hands-free towel dispensers were reportedly contaminated due to frequent contact with dispenser exit surfaces. Levels varied with malfunction frequency and user height (Harrison et al. 2003).
One study recommended a combined approach using towels followed by warm air-drying to minimize bacterial transfer rates and maximize drying efficacy (Patrick et al. 1997). Given that wet hands transfer more bacteria between surfaces, improved efficacy of air dryers was an identified benefit in some studies. Rubbing with PT resulted in fewer bacterial counts likely due to the transfer of indigenous bacteria to the towel. No discussion was given to how this phenomenon relates to health. Little is known about the role of skin microbiota but authors suggested a beneficial function to these commensal organisms (Cogen et al. 2008).
Interpretation of the validity of study conclusions is complicated by variable experimental designs and fluctuating results between and within studies. Commonly referenced hand-drying studies showed distinctly variable results. Some studies concluded that PT were more hygienic than dryers (Blackmore 1989;Ngeow et al. 1989a;Knights et al. 1993), while others found air dryers improved air quality (Ansari et al. 1991;Yamamoto et al. 2005). Several studies found no significant difference between the two (Matthews and Newsom 1987;Gustafson et al. 2000;Taylor et al. 2000).
A more recent study included a multi-site, multi-variable design and found that contamination levels in  bathrooms with PT were generally lower than those with jet dryers (Best et al. 2018). However, many variables were identified that could have confounding effects and data trends were not always significant. Even for statistically significant scenario comparisons, questionable is the health risk associated with low contamination levels (i.e. 1 vs 2 MSSA organisms; 7 vs 21 MRSA organisms) when pathogenic bacteria infectious doses may be 5 or more orders of magnitude higher (Rose and Haas 1999). To our knowledge, high levels of baseline pathogenic bacteria concentrations have not been reported in real-world bathroom or hand-drying comparisons. Best et al. (2018) found significant differences with colony counts of Enterobacter and enterococci where zero colonies compared to 1-2 log concentrations were reported in bathrooms using towels vs dryers, respectively. Both Enterobacter and enterococci are widespread in the environment and used to evaluate water or surface treatment interventions. While each can cause infections, including urinary tract infections, they are more commonly considered harmless commensals in the human gut and do not indicate a direct human health risk when detected in the environment.
For data to be meaningful in terms of a potential health risk associated with drying methods or environmental contamination levels, more research is needed to evaluate hazard prevalence and virulence; exposure routes, frequency and duration; dose-response potentials following exposure; and an overall characterization of risk. Epidemiological studies, particularly those assessing low levels of disease incidence, often require a large sample for an appropriately powered study. Such studies can be cost prohibitive and lack the sensitivity needed to assess low-dose, low-risk events. Previously developed tools in quantitative microbial risk assessment (QMRA) have been used to evaluate health outcomes related to pathogen exposures from environmental surfaces (Beamer et al. 2015;Wilson et al. 2018;Canales et al. 2019). Such computational tools could be applied to characterize surface contamination levels associated with various handwashing and hand-drying methods, assess exposure potentials and forecast risk probabilities. Despite inferring health risks, none of the papers published to date considered a quantitative exposure or risk assessment approach to relate environmental contamination levels with human health outcomes. Furthermore, information is lacking relating seeded tracer concentrations with real-world pathogen exposure levels. Extending risk estimates from tracer bacteria concentrations overestimates risk, as most of these starting concentrations do not represent expected pathogen levels (Ryan et al. 2014;Wilson et al. 2019). Consideration of pathogen transmission and

Strengths and weaknesses
Of the 23 studies evaluated, all had a varying degree of scientific rigor, strengths and weaknesses (Table 3). These strengths and weaknesses were considered when assessing the validity of study conclusions. Performing a statistical analysis on the data generated from the experiment was considered a positive study attribute. Of the 23 scientifically reviewed studies, 22 were experimentally based (non-computer simulations) and 17 of those conducted a statistical analysis. Studies that employed a statistical analysis plan sometimes favoured PT (PT) or electric (jet and warm air) dryers (ED) (Taylor et al. 2000;Yamamoto et al. 2005), or in some cases neither (Matthews and Newsom 1987;Gustafson et al. 2000;Yamamoto et al. 2005;Margas et al. 2013;Kimmitt and Redway 2016;Wilcox et al. 2017). In cases where statistics were not utilized, conclusions were the result of trend analysis, which can be misleading, especially in cases of large data variability or low contamination levels with a questionable relationship to significant health risks. In addition to a lack of statistical analysis, results often lacked information on data distributions and summary statistics. Detailed summary statistics are helpful in analysing study power and data outliers that could be contributing to erroneous conclusions. Without this information, it is difficult to verify statistical methods, compare results across studies or utilize data to inform a QMRA model.
Sample size was also assessed and ranged from a few participants to >100. Larger sample sizes often led to application of statistical testing and thus were considered a stronger attribute compared to studies with smaller sample sizes. A large sample size is especially imperative in studies with large data variability in subjects, test organisms or environmental characteristics. Subject participation varied throughout the studies. In some cases, participants were required to dry their hands using only one method, while other studies reduced uncertainty between subjects by evaluating the same volunteers across multiple methods. In cases where participants utilized multiple drying methods, more weight was given to studies that randomized the drying method order to reduce participant bias.
Types of target organisms were also considered either a study strength or weakness. Targeting multiple organisms (i.e. HPC and faecal coliforms), instead of single tracer types, was considered a strength. Targeting one organism can result in bias, especially for naturally occurring skin bacteria that could be over/under-represented based on handwashing practice and efficacy. The use of microbial tracers was preferred over indigenous skin microbes as a more consistent representation of baseline concentration. A few studies looked at chemical tracers, such as paint, to evaluate the spread of contamination. While providing useful information on the direct contamination around dryers, these methods do not necessarily correlate with microbial transfer, pathogen survival, exposure potentials or health outcomes.
Tracer application to hands also affected data variability. Many studies relied on natural contamination of the hands and environments. Other studies inoculated hands at either exaggerated concentrations or levels indicative of real-world conditions. The hand/microbe interface is complex and impacted by friction, soap/hand sanitizer use, human cell-microbe interaction or adhesion factors, and the presence of organic matter. Some studies seeded gloved hands to minimize cell-microbe interactions. Such unrealistic conditions were considered a study weakness as they create uncertainty relative to the effect of surface interactions on pathogen transfer.
Study location was another source of uncertainty, ranging from biosafety hoods to laboratory washrooms to fully functioning bathrooms. Studies completed in biosafety hoods and laboratory washrooms did not take into consideration real-world effects of human traffic and behaviours. This should be considered when evaluating conclusions or comparing results across studies. Studies utilizing fully functioning bathrooms often did not provide a detailed description of the bathroom layout, which makes it difficult to compare results to other studies given room influences on microbe dispersion due to airflow, room size, and placement of walls and hand-drying dispensers.
Included studies ranged from the 1980s to present day. In this~40-year time span, air dryer and PT dispenser technology evolved. For example, automated towel dispensers reduced hand contact with potentially contaminated handle surfaces but did not eliminate this exposure route completely due to frequent malfunctions (Harrison et al. 2003). Jet dryer technology reduced dry times to approximately 10 s compared to 30 s with warm air dryers but was associated with increased concentrations and dispersion rates of indigenous microbes and tracers in several studies.
Optimized laboratory recovery methods No statistical analysis; unrealistic conditions; large seed concentrations Patrick et al. (1997) Small sample size (n = 7) but large number of replicates (n ≥ 36) Towels autoclaved prior to use; no statistical analysis Gustafson et al. (2000) Advanced statistical analysis; large sample size (n = 99); Randomized study design; high seed concentrations but implemented a realistic handwashing scenario No evaluation of real-world microbial contaminants Taylor et al.
Statistical analysis used with a 5% confidence level for significance All studies agreed on the importance of effective hand drying but consumer preference ranged across methods and consideration of hygiene perceptions, speed and drying efficacy with 63% of 2008 survey respondents preferring PT to warm air dryers (28%) or cloth towels (10%) (Todd et al. 2010). Given that the survey, and much of the historical research on hand-drying hygiene and efficacy, was completed prior to some of the later hand-drying innovations (i.e. automated towel dispensers, jet dryers or HEPA-filtered dryer air) perceptions and data analysis today may be different.
Variable applications and technologies complicated comparisons between studies in our analysis. In some studies, detailed information regarding dryer or towel dispenser technology was included, while others only list general descriptions. Few studies evaluated newer drying innovations such as hands-free PT dispensers or HEPA filtered air dryers on tracer or microbial concentrations, which could have an impact on the overall conclusions of technology comparisons. Table 4 is a semi-quantitative characterization of major study findings and some highly referenced grey literature, sorted relative to descending rigor. Light grey boxes indicate qualitative and quantitative study strengths (i.e. large sample size, statistical significance, controlled variables, advanced statistics) and were assigned a value of two. Minus signs (dark grey boxes) indicate weaknesses relative to variable consistency, methodology, realistic conditions, data quality and other parameters (i.e. result inconsistencies) and were assigned a value of zero. White boxes were assigned a value of one relative to neutral assessments of defined qualitative and quantitative parameters (i.e. determined to be neither a strength nor weakness). An evaluation of the scenario indicates whether study conclusions generally favoured ED, PT or both equally. A scenario category of 'N/A' was assigned for studies that did not compare hand dryers to PT. The studies with the highest relative rigor scores (11 and 10) favoured ED or neither method; however, studies with a relative score equal to or above a six (n = 11) favoured PT (n = 4), ED (n = 2), mixed results (n = 3) or did not compare multiple methods (n = 2).

Electric dryers favoured
Of the 23 papers of determined relevance, 7 had results supporting the use of ED over PT. Reasons for this conclusion ranged from observations of less bacterial counts in the environment or on external device surfaces (Matthews and Newsom 1987; Ansari et al. 1991;Gustafson et al. 2000;Taylor et al. 2000;Yamamoto et al. 2005;Mutters and Warnes 2019) or less concentration of liquid droplets on surfaces near towel or dryer devices (Margas et al. 2013). However, most of the studies favouring ED over PT showed mixed results in terms of statistically significant data depending on which variables were evaluated. Often, within the same study, conditions favoured ED under certain scenarios and PT in others (Margas et al. 2013). Others evaluated results statistically but found no significant difference between hand-drying methods (Gustafson et al. 2000;Taylor et al. 2000;Margas et al. 2013). Four studies favouring ED were funded in part by private sources (Gustafson et al. 2000;Taylor et al. 2000;Margas et al. 2013;Mutters and Warnes 2019), the remaining three were funded by what appeared to be independent or public sources (Matthews and Newsom 1987;Ansari et al. 1991;Yamamoto et al. 2005).

Paper towels favoured
Thirteen papers generally favoured PT use over ED and seven of these studies showed statistically significant differences (Blackmore 1989;Yamamoto et al. 2005;Snelling et al. 2011;Best et al. 2014Best et al. , 2018Kimmitt and Redway 2016;Pitt et al. 2018). Five of the 13 studies were funded by The European Tissue Symposium (Best et al. 2014(Best et al. , 2018Best and Redway 2015;Kimmitt and Redway 2016;Wilcox et al. 2017) and 4 were conducted by the same research collaborators. Five additional studies that   (Blackmore 1989;Hanna et al. 1996;Montville et al. 2002;Snelling et al. 2011;Margas et al. 2013). Two others were funded what appeared to be independent or public sources (Ngeow et al. 1989b;Yamamoto et al. 2005). One study reportedly received no funding (Pitt et al. 2018).
Paper towels equal to electric dryers Eight papers found, at least in part, that PT were equally favoured as ED (Matthews and Newsom 1987;Gustafson et al. 2000;Taylor et al. 2000;Snelling et al. 2011;Margas et al. 2013;Kimmitt and Redway 2016;Wilcox et al. 2017;Best et al. 2018). At least three of these studies conducted statistical analysis but found no difference between variables regardless of drying method used. Specifically, Taylor et al. (2000) found no significant difference in microbial aerosols created from ED compared to PT. Others also found that PT and ED were not significantly different in terms of contamination potentials from microbial aerosols under a variety of test conditions (Gustafson et al. 2000;Margas et al. 2013;Kimmitt and Redway 2016;Wilcox et al. 2017;Best et al. 2018).

Miscellaneous study results
Five papers failed to evaluate PT and ED in the same study (Patrick et al. 1997;Harrison et al. 2003;Gendron et al. 2012;Jensen et al. 2015;Huesca-Espitia et al. 2018).
These studies provided useful background information relative to other variables of potential interest. For example, Gendron et al. (2012) focused on the potential for PT to add bacteria to the environment. Sources include indigenous organisms found on the product post-manufacturing but pre-use, particularly with recycled pulp products, and further contamination potentials during transport to dispensers (Gendron et al. 2012).

Impact of study rigor
A common observation in our analysis is that most of the studies, regardless of conclusion, lacked sufficient rigor to form defensible conclusions. Some utilized small sample sizes (Kimmitt and Redway 2016;Wilcox et al. 2017) or unrealistic conditions, such as evaluating how paint spreads (Best et al. 2014), which may have no relationship to bacterial transmission. Others did not consider either handwashing or soap use before drying (Snelling et al. 2011;Best and Redway 2015;Jensen et al. 2015), which may exaggerate results given that washing is an expected drying precursor for contaminated hands. Furthermore, highly variable results within and between scenarios was a common effect (Taylor et al. 2000;Yamamoto et al. 2005;Margas et al. 2013;Best and Redway 2015;Jensen et al. 2015). Several authors made conclusions on the safety of different hand-drying methods that were not supported by the data (Ngeow et al. 1989a;Gould 1994;Kimmitt and Redway 2016). Regardless of scenario, most studies lacked objective statistical significance testing or failed to define test criteria (Ngeow et al. 1989a;Best et al. 2014;Kimmitt and Redway 2016).

Grey literature summary
In 2018, there was a resurgence of news interest on the topic of PT vs hand dryers due to newer published research (Best et al. 2014(Best et al. , 2018Best and Redway 2015;Kimmitt and Redway 2016;Huesca-Espitia et al. 2018).
In addition, several grey literature reports went 'viral' on the Internet, such as the photo of a Petri dish placed in a jet air dryer as part of a small class experiment (Astor 2018). Grey literature includes materials (i.e. research, white papers, media and industry reports, government documents, surveys, theses and dissertations, data summaries, etc.) produced by individuals or organizations outside the traditional peer review process. While the classification of information as grey literature does not necessarily mean the data is not scientifically valid, there may be less oversight and review of the content. Here we included a review of the grey literature using the top Google results when searching 'PT air dryers'. We also included sources disqualified during the scoping review, such as papers that did not contain original research or were not peer-reviewed (Gould 1994;Lee 1994;Huang et al. 2012;Haynes 2014). To extract information from the grey literature, we initially reviewed known popular sources, including Dr. Oz, MythBusters, the New York Times, Washington Post, Snopes.com and Oprah.com. Most of these reports feature headlines suggesting there may be a more hygienic way to dry hands ( Table 5). The majority of these outlets favoured PT based on cited research from peer-reviewed or other grey literature sources but often stated that data was inconclusive relative to safety concerns. Some media outlets performed their own small-scale experiments, or simply cited personal preference.
Some of the data in the grey literature misconstrued research results. For example, multiple online articles about the Huesca-Espitia et al. (2018) study state that PT are a better drying option and that dryers blow faecal matter on hands, which is not completely reflective of the authors' conclusions. The Huesca-Espitia et al. (2018) study did not evaluate PT or run any comparative tests between dryers and towels. The paper does voice concern over dryers potentially blowing pathogens around bathroom environments but also reports very low counts of microbes and was favourable towards warm air dryers with HEPA filters that reduced counts by fourfold. Few of the grey literature reports provide citations or links to publications referenced. Many of the findings were overgeneralized and included unsubstantiated claims, Internet rumours and inaccurate information reported as fact. According to Snopes.com, a popular fact-checking website, an email was circulated in 2009 claiming schools and universities were required to replace all hand dryers with towel dispensers in response to the H1N1 outbreak ('FACT CHECK: Do Hot Air Hand Dryers in Restrooms Spread Disease?' 2018). In its analysis, Snopes rates the claim as 'unproven'.
One of the concerns with the grey literature reviewed in our study is the potential bias in the reporting. Some of the top-ranking articles in our Google search were white papers authored or sponsored by paper companies (Vitali n.d.;Lee 1994). Additionally, media reports frequently used sensationalized headlines. While such headlines may increase traffic, they sometimes overgeneralize or exaggerate study results. Consumers may only read headlines which can influence public opinion towards biased or erroneous conclusions (Ecker et al. 2014). Ecker et al. also found that readers are more likely to recall details of the article relative to expectations set by inferences in the headline and not by what the article actually says.

Conclusion
The objective of our study was to provide a scoping review of the current scientific peer-reviewed publications comparing the use of PT vs ED and their impact on environmental and human health using standardized criteria. Our focus was to categorize the papers relative to scientific rigor in the study design and compare statistically significant results across defined categories of hand-drying methods. We also examined statistically significant results relative to paper conclusions and the impact of these studies on public perception (as determined by media posts and other grey literature). Study categories were scored quantitatively as positive/+, neutral or negative/À, using a scoring system of 2, 1 and 0 respectively. While it is important to review each study holistically, providing a study rigor score to each reviewed paper adds context to our assessment, which should be read in full for a broader perspective.
The most prevalent conclusion from our scoping review is that the breadth of scientific data does not support a uniform conclusion as to which drying method is safer or more hygienic. Some studies showed a statistically significant reduction of contaminants within certain scenarios but not all-whether studies involved large sample sizes or were small pilot studies, the differences between drying methods were not consistently clear. Many studies reported no statistical significance and used either low concentrations representing low-risk scenarios or high concentrations that are not representative of pathogen levels on hands or bathroom surfaces.
Overall, the large discrepancy in results is likely due to the wide range of experimental conditions implemented. Variables include: baseline concentrations, target organisms, population and sample sizes, drying times, number of towels used, dispenser designs, rubbing behaviours, microbial assay methods, Hawthorne effects, behavioural reproducibility, handwashing product types and efficacy, skin integrity, residual moisture levels, ambient air movement and more.
Methodological differences compounded these uncertainties. Mixed results favouring one drying method over another between, and even within studies, was common. None of the studies in the published literature related • 'There seems to be no particular microbiological danger to use of hot air hand dryers'. (Matthews and Newsom 1987) • 'Hot air dryers are generally not recommended for use in health care settings because such dryers are relatively slow and noisy and their hygiene performance is questionable' (Blackmore 1989) • '. . .associated infection risks. . .make [air dryers] unsuitable for use in critical patient care areas' (Ngeow et al. 1989a).
• 'Risk reduction strategies should aim to achieve a level of risk as low as practically possible' (Harrison et al. 2003).
• 'The health and safety aspects of jet air dryers where hygiene is paramount should still be carefully examined by the scientific community' (Redway and Fawdar 2008).
• '. . .a diverse community of culturable bacteria contaminates unused paper towels. . .some of these bacterial strains may be toxin producers. . .This may be significant for some clinical and industrial settings as well as for immunocompromised individuals. . .' (Gendron et al. 2012).
• 'From a hygiene standpoint, paper towels are superior to air dryers' (Huang et al. 2012).
• 'These results suggest that air dryers may be unsuitable for use in healthcare settings as they may facilitate microbial cross-contamination via airborne dissemination to the environment or bathroom visitors' (Best et al. 2014).
• '. . .in locations where hygiene and cross-infection considerations are paramount, such as healthcare settings and the food industry, the choice of hand-drying method should be considered carefully' (Kimmitt and Redway 2016).
• 'Electric hand dryers are not suited to clinical settings. . . Infection control building guidance needs to be amended and strengthened' (Best et al. 2018).
Some studies stated concerns relative to the use of air dryers in healthcare settings. It is important to note, however, that the most effective, promoted and commonly practiced hand hygiene method in hospitals and other healthcare settings involves the use of an alcoholbased hand rub (Vermeil et al. 2019). In healthcare settings, alcohol-based hand rubs require less time, are more effective than washing with soap and water, are more accessible than sinks and naturally exclude the use of hand-drying protocols (Boyce and Pittet 2002;World Health Organization 2009).
Increased bacterial counts in the environment do not necessarily indicate increased health risks. To date, no study has shown a correlation between hand-drying methods and health outcomes using empirically derived or QMRA model simulations. Thus, conclusions as to the safety of PT vs air dryer use relative to health outcomes are unsubstantiated and premature. Consideration of infective doses, transmission pathways, handwashing and drying efficacy, transfer rates and other exposure variables is critical before a recommendation can be made based on health outcomes.
Not all surfaces are equal contributors to exposure pathways. While floors may be more contaminated in air dryer bathrooms, their relevance in terms of an exposure or risk scenario has not been evaluated. The most direct exposure potential is from contaminated hands and thus drying methods resulting in fewer microbes on the hand are likely to be associated with lower risk potentials. Therefore, contaminated dryer or PT dispenser buttons may represent the highest risk in practice. Conversely, new technologies such as automated dryer and towel dispenser designs may largely eliminate this transmission route.
Our scoping review identified broad analysis of the research question: 'Are hand dryers more hygienic than PT?' but with collectively inconclusive results. Studies were conducted by a variety of private and independent research groups spanning at least nine countries. Fifteen of the 23 studies were funded by private, or a combination of private and public, sponsors. Although independent funding is preferred, federal or state funding agencies are unlikely to sponsor such special interest projects-particularly when current research does not indicate a significant health concern.
The second question we sought to answer is 'Are PT safer than hand dryers relative to human infection risks?' We found no data to support any human health claims relative to hand dryers vs PT use. Inconclusive results in the current body of research and a lack of data to support health claims represent data gaps preventing the advancement of policy or practice recommendations relative to hand-drying protocols. Of notable importance is the need to evaluate risks from hand-drying activities in consideration of handwashing scenarios, given that the greatest uncertainty in hand contamination is associated with the handwashing method, and not the drying method.

Future study recommendations
Future studies are needed to examine the relationship between contamination that occurs due to hand-drying methods and human health outcomes. This can be accomplished by utilizing a real-world scenario while still controlling for certain variables. Ideally, the study should be done in multiple functioning bathrooms that have different characteristics, such as single stall and multiple stall bathrooms. A microbial tracer should be applied to participants' ungloved hands at a concentration that would be representative of what is expected on hands after a handwashing event. The hands of the participants and surfaces that participants would encounter after drying their hands should be targeted for sampling as well as air. Samples should be collected after single and multiple drying events to see the effects over a variety of usage conditions. Results from the study can then be used to inform mathematical models to evaluate human health outcomes expected for each type of drying event. Sensitivity analysis should be performed to rank which exposure variables drive risk outcomes.
In summary, needed are long-term observational studies and/or health impact related evaluations/simulations. Gaps in the research include: • Evaluation of contamination potentials with current technologies and standardized methodology.
• Consistent variable assessment representing real-world exposure potentials (i.e. consideration of baseline contamination values, handwashing and drying efficacy, pathogen dose response, and correlation between indigenous microbiome or tracer concentrations to predicted pathogen levels).
• An exposure assessment and sensitivity analysis considering surfaces of increased probability to impact exposure.
• A risk assessment relating relevant exposure scenarios to health outcomes.
• A relative risk assessment comparing different contamination levels from other bathroom sources (i.e. toilet aerosols or human behaviours) and organism infectivity with acceptable risk targets.
Although numerous studies have been published evaluating the 'best' method for hand drying, 'best' has been defined in a variety of ways relative to bacterial removal efficacy, environmental contamination potentials, ecological or cost benefits, noise and more. No study to date has examined the 'best' drying method relative to health outcomes.
As reported repeatedly throughout this study, the wide range of data variability makes it difficult to draw finite conclusions about any particular hand-drying method.
We agree with the conclusions of Wilcox et al. (2017) who state that studies to determine the risks associated with hand-drying methods have not been conducted but are warranted to determine whether adverse infection consequences occur.

Funding support
This study was supported, in part, by an unrestricted grant from Excel Dryer, Inc. The funding source played no role in study design, data acquisition, analysis or decision to report these data.

Conflicts of Interest
K.A. Reynolds has received research funds from Excel Dryer, Inc.