Increasing Hand Washing Compliance With a Simple Visual Cue

Abstract

We tested the efficacy of a simple, visual cue to increase hand washing with soap and water. Automated towel dispensers in 8 public bathrooms were set to present a towel either with or without activation by users. We set the 2 modes to operate alternately for 10 weeks. Wireless sensors were used to record entry into bathrooms. Towel and soap consumption rates were checked weekly. There were 97 351 hand-washing opportunities across all restrooms. Towel use was 22.6% higher (P = .05) and soap use was 13.3% higher (P = .003) when the dispenser presented the towel without user activation than when activation was required. Results showed that a visual cue can increase hand-washing compliance in public facilities.

Promoting hand hygiene compliance is an ongoing public health effort.¹ Proper hand hygiene is among the most important measures for preventing and controlling microbial pathogen cross-transmission^2–4 and is a cost-effective intervention for the control of many infectious diseases.⁵ The Centers for Disease Control and Prevention’s Cover Your Cough campaign advised people to use their upper sleeve or elbow, rather than a hand, to cover their mouth when they cough.^6,7 The Secretary of Health and Human Services notoriously chastised a reporter during a press conference for not following the guidelines.⁸ More recently, the increased incidence of a vector resistant to alcohol gels, Clostridium difficile, in US hospitals has increased the importance of effective hand-washing behaviors.⁹ Despite efforts to increase public awareness, hand hygiene compliance rates are difficult to influence and remain stubbornly low.¹⁰ In particular, increasing hand-washing rates after the use of public restrooms has been challenging.^11,12 Interventions have been tested in numerous settings (e.g., schools and hospitals); however, no single intervention has produced consistent and lasting improvements in hand-washing rates.^3,13

Education, training, and behavior modification are often cited as important components of hand-washing compliance (HWC) programs.^11,14,15 However, interventions with only 1 component have not yielded significant gains in HWC, or the gains have been transitory, falling back to baseline levels over time.^{13,14,16–18} Moreover, the effectiveness of education-centered interventions is primarily limited to the individuals targeted by the effort.¹⁹ Research examining hand-washing interventions often lacks methodological rigor, limiting causal inferences.^13,20 Therefore, an ideal HWC intervention would be noninvasive and create sustained increases in HWC that could be statistically verified. The intervention would also be easily implemented as part of the existing hand-washing station equipment and provide a positively framed message to engage in the desired behavior.⁴ Modifying towel dispensers to signal consumers is a behavioral cue that may have sustained impact on HWC rates, with concomitant benefits.

VISUAL CUES TO ACTION

Creating physical environments that provide visual cues and minimize barriers to HWC has achieved some increases in hand-washing rates that endure beyond the initial intervention period.²¹ The hand-washing built environment or layout has been reported to be an important determinant of HWC among health care workers.^22–24 The perception of convenience has also been associated with increased HWC.²

The presentation of a towel by an automatic towel dispenser as a visual cue to increase HWC has not been assessed in the literature. However, other types of visual cues have shown success at improving HWC in health care and university environments. Thomas et al. found that conspicuous and proximate positioning of alcohol-based hand sanitizer dispensers resulted in statistically significant increases in product usage.²⁵ Similarly, Birnbach et al. reported that the positioning of a free-standing alcohol-based hand sanitizer dispenser directly in front of a security desk in a hospital lobby significantly improved visitor HWC.²⁶

Johnson et al. examined whether the absence or presence of reminders about hand washing in public restrooms on a university campus influenced HWC.²⁷ Researchers observed a significant increase in HWC among women—but not men—who were exposed to a hand-washing reminder sign.

Use of visual cues to improve HWC has also been investigated in other community environments. Text messages flashing a variety of prompts were used in a highway service station in England.²⁸ The study used 7 types of text messages: knowledge of risk, knowledge activation, norms or affiliation, status or identity, comfort, disgust, and cues. Only the normative message, “Is the person next to you washing with soap?” resulted in significant increases in hand washing with soap for both genders.

Although the effect of automatic towel dispenser settings on HWC behaviors has not been investigated previously, research regarding oculomotor behavior during a hand-washing task suggests the efficacy of this type of visual cue in promoting HWC behaviors. Pelz and Canosa examined the eye movements that occur during the common task of hand washing and posited that they were part of a search process, seeking specific targets associated with the task of hand washing; approximately 95% of the eye fixations were dedicated to completing an immediate hand-washing subtask.²⁹ The authors suggest that the remaining 5% of fixations may help to reduce the conscious effort needed to sequentially transition between the distinct subtasks of hand washing. According to Pelz and Conosa, look-ahead fixations may be a sophisticated perceptual strategy that serves to “ease the cognitive and attentive loads of perception in the real world.”^29(p25) Thus, the presentation of a towel may be effective as a visual cue by reducing cognitive barriers to HWC and focusing attention on completing the subtask. Look-ahead fixations occur to serialize complex activities as required by a task and take place below consciousness.²⁹ This may enable this type of visual cue to unobtrusively improve HWC.

Visual cues can serve as conspicuous signals that gain the attention of individuals in a hand-washing environment and increase awareness that hand hygiene products are available.²⁵ The visible presentation of a towel applies the following physical and cognitive principles of ergonomics that encourage hand-washing behaviors: clear visibility, unobstructed and easy access, and a minimal effort requirement.³⁰ The visible presentation of a towel by an automatic towel dispenser more closely matches the user’s conceptual model of the usability of the dispenser with the appropriate action and facilitates hand washing.³¹ Thus, this type of external cue may help trigger the perception of convenience by individuals in this environment.^32,33 A simple, visual cue that a hand-washing environment is ready should have a positive impact on HWC. In particular, the presentation of a towel by an automatic towel dispenser indicates that a critical element of the hand-washing event is available.³⁴ This type of visual cue has been termed by Norman as a perceived affordance and is defined as “visual feedback that advertises the affordances.”^35(p39) The automatic towel dispenser serves as a perceived affordance because it provides a cue that a paper towel is ready to be dispensed for use in hand washing. Because hand washing is a ritualized behavior consisting of several steps, observing that a towel is already presented by a dispenser may reduce the cognitive effort associated with completing these steps.^16,29

The hand-drying aspect of the hand-washing ritual is particularly important for 3 reasons. First, the most unsatisfactory hand-washing experiences end with the discovery by a user with wet hands that no towels are available to complete the ritual.^30,36 People often dry their hands on their own clothing, eliminating the benefits of hand washing.³⁷ Second, the visual cue that the last task in the ritual is likely to be completed successfully creates a strong signal that the entire exercise will go smoothly.¹⁶ Such a predisposition is likely to lead the individual to complete the ritual as normatively prescribed.³⁸ In the case of hand washing, one normative behavior that is likely to be positively influenced by the presence of towels is the use of soap—which is often neglected.³ Finally, the towel provides a means for exiting the facility while maintaining sanitary hands by using the towel to open the door.³⁸ Because presentation of a towel is less obtrusive than cues used in other interventions (e.g., written warning signs and audible signals triggered by motion), it represents a nonjudgmental cue to action that is closely linked to the intended behavioral action.^39,40 It is a “gain-framed” signal that emphasizes that the hand-washing experience is likely to be positive.⁴¹ This contrasts with a “loss-framed” message that the risks of not complying with hand hygiene guidelines are harmful.⁴² The gain-framed message may be more efficacious at improving HWC and is more consistent with good public health practice.⁴¹ Therefore, we hypothesize that providing a simple, nonintrusive visual cue to wash station readiness will lead to a sustained increase in HWC relative to an environment where no signal related to gain or loss is provided.

We tested the efficacy of a simple, visual cue to promote HWC in public restrooms. The visual cue is to present the towel (i.e., ready for immediate use) from a motion-activated dispenser when the user arrives at the hand-washing station. The visual cue is absent when users have to activate the towel dispenser by waving their hand in front of or below the machine. We conducted an interrupted time-series study to assess the impact of presenting the towel on paper towel and liquid soap consumption in public restrooms located in an academic classroom building of a large university. We also described the human factors that influence HWC in public facilities.

If shown to be effective, this intervention could significantly reduce the disease burden associated with poor HWC simply and cost effectively. For public health officials, the findings can be integrated into routine inspections of public facilities as either a guideline or a regulation. Where towel dispensers are already present, and their use is widespread, merely changing the setting to the displayed option is a relatively effortless and low-cost intervention. For towel dispenser manufacturers, the findings provide valuable information for how machines’ factory settings should be configured upon delivery. The benefit to the public of having a simple, unobtrusive cue for improved health behaviors is an ideal public health intervention in that it does not rely on high-threat messaging or other overt stimuli that diminish in effectiveness with sustained exposure.⁴²

METHODS

We conducted our study in the Bryan School of Business and Economics building at the University of North Carolina at Greensboro, a campus with more than 17 500 students. The study took place over 10 weeks, from February 20, 2012, to April 29, 2012. We used an alternating stimulus design because our goal was to examine 2 treatment alternatives in the same population.⁴² Treating washrooms users as a repeated experiment with alternating conditions is consistent with the environment's usage, which brings the same cohort to the same space on multiple occasions. Therefore, other designs and analyses that treat intervention and control participants as different samples were less desirable.

Study Design

We used 8 bathrooms (4 each for men and women) to compare the effects of the intervention (towel presented) and the control (towel not presented) condition on HWC. In the control condition, the automatic towel dispensers were set to the motion-activated or on demand mode, which required the motion sensor to be activated by the user. All dispensers were set to the control condition for several weeks before winter break and to the study’s measurement start date in the early part of the spring semester. In addition, the towel dispensers in the rest of the building, primarily faculty office spaces, were set to the towel not presented position in the fall semester. Therefore, the building’s users had not been exposed to the towel-presented stimulus for several months before the intervention.

For the intervention condition, the dispensers were set to the towel-presented mode, in which no user action was required to receive a towel. The first week of the study served as the baseline, towel not presented measurement period, and dispensers were set to the towel-presented condition in the even-numbered weeks (weeks 2, 4, 6, 8, and 10 were intervention periods) for the rest of the 10-week study.

We measured restroom use with electronic motion detectors at the doors. We assessed HWC by measuring the volume of towels used in the 14 automatic towel dispensers and the volume of soap dispensed by the 8 soap dispensers in the study bathrooms. The measurement of the volume of hand hygiene products is considered to be an indirect indicator of the frequency of HWC behaviors,^25,43,44 and increased use of product has been found to be an effective proxy for HWC.^2,45,46

Procedures

We placed new towel rolls and soap containers in the 14 automatic towel dispensers (enMotion, Georgia-Pacific, Atlanta, GA) and 8 soap dispensers (Symmetry, Buckeye International Inc, Maryland Heights, MO) after the removal of all partially used towel rolls and soap containers at the end of each week of the study. We also collected any rolls that had been replaced during the week. This occurred after the building was officially closed for the weekend on Sunday afternoons or before 8:00 am on the following Monday. At this time we also set the automatic towel dispensers to either present or not present the towel, depending on the week.

We weighed each of the partially used towel rolls and soap containers with an electronic scale (Zieis 6000-g-capacity scale accurate to 1.0 g, ISO 9001 certified; Zieis, Apple Valley, MN). We subtracted the weights of the unused product from the beginning inventory to determine the volume of soap and towels used for each week of the study. We used the final weekly towel roll and soap bag weights as the numerator for calculating the ratio of people entering and exiting the bathrooms to those who exhibited hand hygiene behaviors by using soap and towels.

Each of the 8 study bathrooms was equipped with a wireless infrared sensor (PC-PIR15-D Wireless Detector, SenSource Inc, Youngstown, OH). These sensors recorded the number of people who moved in and out of each bathroom. The total count of people entering and exiting the bathrooms, divided by 2, indicated the number of hygiene opportunities that occurred each week. We used these values as the denominator for the determination of weekly HWC rates.

RESULTS

We measured more than 97 351 hand-washing opportunities (HWOs) during the 10-week study period. The weekly ratio for towel consumption to restroom users ranged from 59.6 to 80.6 ounces per HWO per 1000 persons. Weekly rates for the intervention (towel presented) and control (towel not presented) conditions are shown in Figure 1. The weekly ratio for soap consumption to restroom users ranged from 10.5 to 1.53 ounces per HWO per 1000 persons. Weekly rates for the intervention and control conditions are shown in Figure 2.

FIGURE 1—

Volume of towels used per hand-washing opportunity per 1000 persons in 10-week study of a visual cue intervention to increase hand-washing compliance: University of North Carolina at Greensboro, 2012.

FIGURE 2—

Volume of soap used per hand-washing opportunity per 1000 persons in 10-week study of a visual cue intervention to increase hand-washing compliance: University of North Carolina at Greensboro, 2012.

The mean weekly ratio for towel consumption to restroom users was 61.8 (SE = 1.7) ounces for the control condition and 75.9 (SE = 1.5) ounces for the intervention condition per HWO per 1000 persons (χ² goodness-of-fit test, P = .05; Figure 3). The mean weekly ratio for soap consumption to restroom users was 11.4 (SE = 0.4) ounces for the control condition and 13.1 (SE = 0.7) ounces for the intervention condition per HWO per 1000 persons (χ² goodness-of-fit test, P = .003).

FIGURE 3—

Mean weekly ratio of towel consumption to restroom users in 10-week study of a visual cue intervention to increase hand-washing compliance: University of North Carolina at Greensboro, 2012.

* Statistically significant difference at P = .05.

DISCUSSION

Adequate HWC among public restroom users is among the most effective means of preventing the spread of infectious disease.⁴⁷ HWC comprises disinfecting hands at appropriate moments, complete hand rubbing, and sufficiently long drying time. We explored the impact of a simple intervention on hand hygiene practices in a public university setting—a visual cue to towel availability. We studied measured towel and soap consumption rates per facility use separately. Previous studies reported baseline compliance incidences from 44% to 66% in hospital settings where significant efforts had been made to promote hand hygiene.^28,39–41 Although we did not estimate compliance rates, we observed significant increases in both of the elements that are central to complete HWC—soap and towel use—under the presence of the stimulus. Furthermore, the significant and sustained increase in product use indicated that both hand washing and complete HWC (using both soap and water) were positively affected by the stimulus of presenting a towel without user action.

Unlike other, commonly used interventions—education and signage—the towel-presented stimulus is not an overt suggestion. Active interventions are prone to the Hawthorne effect, whereby the participants consciously recognize that a stimulus is being applied and act accordingly for a brief period.^48,49 Our intervention did not have this limitation and did not appear to suffer from diminished impact over time—at least for the 10 weeks of our study.

Most of the research to date has been conducted in health care facilities, where the risk of disease communication by hand is particularly high and the consequences more severe. An increase in the incidence of C. difficile infections in hospitals is emerging as a major public health threat.^50,51 Because C. difficile is a spore, it is not susceptible to other hand hygiene strategies, such as use of alcohol-based gels. Therefore, effective HWC that addresses the C. difficile threat is increasingly important in public facilities—health care settings in particular. Increasing the use of soap and towels through a simple visual cue may prove efficacious in these settings.

Limitations

We used the ratio of the volume of soap and towels consumed to individuals using the restrooms as an indicator of HWC, rather than direct observation. Measuring product use as a proxy for HWC has some limitations,⁵² such as (1) the use of hand-washing products does not reveal whether hand washing is actually being performed or is being performed correctly⁴⁵; (2) measuring product use can sometimes be inaccurate, because of wasting or spilling of product or changes in volume dispensed⁴⁵; and (3) information is not provided about when and why HWC is not occurring.⁵³ However, the validity of this type of indirect measurement has been established and serves as an unobtrusive method of collecting HWC data.^28,54

We used the weight of the soap consumed as the indicator of soap usage. It is possible that each activation of the soap dispenser produced varying amounts of soap, depending on the mechanical force applied to the dispenser by the user. Thus, the average weight of the soap dispensed with each activation may have varied, possibly decreasing the reliability of this instrument in indicating HWC. To the extent that the varying-pressure phenomenon influenced soap consumption results, it is likely that the large sampling size of encounters per condition and the random nature of the variance mitigated this issue. Nevertheless, the use of counters in soap dispensers that record each time that the dispenser is activated would improve the reliability of soap use as an indicator of hand-washing behaviors.

This was a pilot study to ensure that reliable measurement could be made. The study duration was relatively short, and we used only 8 restrooms. Potential problems with alternating stimuli were sequential confounding, carryover effects, and alternation effects.⁴³ However, all of these potential issues would diminish the statistical power of the analysis. Therefore, the significant findings were more likely to be understated. Future studies should use longer study periods to increase effect sizes and study power. The use of a discontinuous, multiple-baseline design would make the study more robust—although our design yielded significant results. In addition, anecdotal evidence that facility managers may be changing automated dispensers to the setting that requires user action in order to save money merits further investigation.

We did not separate female and male responses to the intervention stimulus. Other research has indicated that men and women respond differently to messages targeting various psychological domains.²⁸ A similar difference may exist for simple, nontext visual cues, which merits further exploration.

Conclusions

Having a towel presented upon arrival in a public restroom is an effective, unobtrusive stimulus to increase HWC. We observed significant increases in both towel and soap consumption when dispensers were set to automatically present towels when restroom users approached hand-washing stations, indicating that this visual cue can improve HWC in public settings. Furthermore, such machines are already in widespread use and merely need be set to present towels without user activation to improve HWC across a significant population. However, in many locations the dispensers are set to require motion activation, thus requiring consumers to make conscious assessments about the wash station’s readiness.

Anecdotal evidence from facility administrators indicates that they frequently set machines to motion activation mode to save money. Although this is a cost-effective strategy from the maintenance department’s perspective, the societal costs of increased disease burden likely outweigh the marginal savings. Therefore, an education campaign directed at public facilities managers and manufacturers of towel dispensers might be warranted. Although facilities managers may be more concerned about cost savings than public health consequences, manufacturers have an incentive to set dispensers to automatically present towels, because increased HWC inevitably leads to increased consumption of the manufacturers’ products: soap and towels.

For consumers, the use of paper towels as part of the hand-washing routine is ideal. Other alternatives, such as blowers, do not provide complete drying, and consumers often avoid them altogether or reduce HWC effectiveness by completing drying in another manner.