Article Text
Statistics from Altmetric.com
- behavior and behavior mechanisms
- delivery of health care
- equipment and supplies
- hospital medicine
- psychological phenomena
Summary box
What are the new findings?
A behavioural nudge can be especially effective and durable if based on sludge reduction, defined as lowering a barrier that discourages people from doing the right thing.
We introduced a simple label to encourage people to report an empty hand sanitizer dispenser located in public places inside a large acute care hospital.
We later checked 100 dispensers inside the hospital and found only 2 empty (notified and refilled promptly).
A similar check of 100 dispensers at a nearby hospital found 11 empty (no readily apparent way to notify for refilling).
How might it impact on healthcare in the future?
Other hospitals could immediately adopt this nudge to strengthen the connection between intentions and actions for hand hygiene.
Crowdsourcing provides a practical method to support hand hygiene by allowing outpatients, families, visitors, staff, volunteers and others to notify hospital services about empty hand sanitizer dispensers that requires refilling.
Background
The COVID-19 pandemic has caused over a 100 million cases and substantial numbers of admissions to acute care hospitals. Hand hygiene is a practical, affordable, acceptable, safe, reliable and effective strategy to mitigate the risks of nosocomial transmission to other patients in hospital for other reasons. Hand hygiene is also critical for ensuring the safety of healthcare workers and maintaining overall hospital staffing. However, hand hygiene is fallible because of many cognitive biases in everyday care.1 Clinicians, for example, have no easy way to determine if they are contaminated, whether their hand hygiene is sufficient, or if they are recalling correctly a repetitive activity.2–4
One strategy for improving hand hygiene is to increase the availability of hand sanitizer dispensers in public places is to structure more convenient opportunities for repeated hand washing. Each dispenser installation requires no electricity, plumbing or batteries and can be retrofitted onto existing infrastructure quickly and cheaply (figure 1). The logistics mostly require finding a wall and a power drill. In turn, dispensers can be positioned in diverse locations so individuals have easy access during idle moments. These locations include elevators, corridors, staircases, cafeterias, computer workstations, administrative offices, parking lots and other public places.
Problem
The COVID-19 epidemic means increased numbers of hand sanitizer dispensers are active inside large hospitals. The intent is to boost hand hygiene since access is a necessary—not sufficient—requirement for hand hygiene.5 A downside is that any single dispenser can become empty. Therefore, hand sanitizer dispensers need to be refilled at unpredictable times depending on variable floor traffic and personal styles (eg, double pumpers). Sunnybrook Health Sciences Centre (Sunnybrook) in Canada, for example, has over 800 hand sanitizer dispensers in public places where busy locations may need a refill two to three times per day, whereas other locations might need a refill only two to three times per month.6
Keeping all dispensers functional is now a tedious, vexatious problem in large hospitals. Routine monitoring of dispensers can mean endless monotony for custodial staff charged with checking and refilling (and a prevailing pandemic means staff are already busy).7 Moreover, the fluctuating rates of consumption make routine refills during off-hours an impractical solution. The scattered geographic locations also surpass any reasonable mapping algorithm for checking dispensers during regular hours.8 Furthermore, partial refills for a partly depleted dispenser is an unworkable approach due to the sealed internal cassette systems that prevent fluid evaporation.
Solution
At Sunnybrook, a solution to help keep dispensers functioning is to label each with the internal telephone number (4555) for the hospital Environmental Services office. This number is active 24/7 for managing garbage removal, room cleaning, accidental spills or unforeseeable mishaps. The label is written ‘If empty call 4555’ to avoid jargon and allow a large font (Figure). On average, the office receives over a dozen such calls during an 8-hour shift from individuals reporting an empty dispenser (outpatients, families, visitors, staff and volunteers). We believe the label is a practical way for a hospital to use crowdsourcing to increase the reliability of hand hygiene.
Several fundamentals of an effective nudge are exemplified by a simple label for reporting an empty dispenser.9 First, the label is readily incorporated onto existing equipment with trivial costs and high durability. Second, the nudge imposes no costs under normal circumstances when the dispenser is functioning yet provides a reminder that hand hygiene is considered seriously. Third, the critical monitoring occurs naturally through regular activity with no extra supervision. Finally, the necessary triggering stems from the end user and may potentially help to convert negative aggravation into positive action (phone number can also accessed by the public on a cellular phone).10
Theory
By design, this nudge differs from past strategies to improve hand hygiene through posters or other reminders to promote behaviour.11 Studies suggest more exhortation may not always lead to more hand hygiene since people already have baseline awareness.12 13 In contrast, this nudge is designed to target sludge, defined as elements of a process that discourage people from doing the right thing.14 A classic non-medical example of sludge is an awkward system hindering people from cancelling a magazine subscription they no longer need. The reduction of sludge can be a distinctly effective, acceptable and lasting way to advance positive behaviour.15
Sludge reduction differs from formal education because the nudge tells people something they already could know. In particular, a moment of reflection could readily cause a hospital visitor, staff member or other rational agent to call the hospital switchboard (dial ‘0’) and report an empty dispenser. The label, therefore, does not add essential new information; instead, the label makes the unwanted anomaly apparent and actionable. In addition, the label signals that the hospital recognises anomalies, welcomes such calls and is sincere about hand hygiene. Together, this suggests the nudge works partially by increasing salience and by offering a hint when it is needed most.16
A downside of nudges based on sludge reduction is that each can seem a serendipitous discovery. No easy guidelines offer a systematic approach to serendipity, yet the EAST taxonomy from the United Kingdom Behavioural Insights Team is popular.17 The elements mean easy (few obstacles), attractive (appealing design), social (encourage commitment to others) and timely (prompting people when receptive). In our case, the nudge is easy (single number), attractive (clear simple label), social (affable phone receptionist) and timely (moment chosen by user). In the future, the EAST approach may help identify other nudges to improve hand hygiene.18
A simple label for reporting an empty dispenser highlights general features of effective nudges. The nudge involves almost no memory, repeated attention or doubts over achieving the goal. The telephone call raises no hard thinking about complex tradeoffs, long-term consequences, or ethical quandaries. The interaction offers nearly immediate confirmation that action has begun. The self-identity affirmation of providing a benevolent act can convey meaningful intrinsic positive feelings for self-image. Like all nudges, the strategy addresses a discrepancy between intentions and actions rather than challenging a person’s innate priorities, preferences, beliefs or self-esteem.
Much of behavioural economics is rooted in Prospect Theory.19 In particular, losses tend to loom larger than the corresponding gains so that, in our situation, the frustration of encountering an empty dispenser exceeds the satisfaction of contacting a full dispenser. The negative emotion, in turn, might trigger the affective heuristic that leads to devaluing the benefits of hand hygiene and the credibility of authorities recommending adherence.20 These natural reactions may be fleeting, discounted and unremembered when judged later by self-report. The nudge, in contrast, may help prevent structural barriers (eg, empty dispenser) from worsening cognitive barriers (eg, fallible motivation).
Theories beyond behavioural economics additionally support the potential of sludge reduction as nudges to change behaviour. In particular, the Stages of Change model of behaviour suggests action flows from decisional balance that reflects both pros (eg, positive reasons) and cons (eg, negative drawbacks).21 This means optimising hand hygiene requires both increasing motivation and decreasing disenchantment, thereby justifying diverse strategies so one nudge does not preclude further improvements. An irony of nudges based on sludge reduction, however, is that successful implementation means a problem disappears and the gains may be unappreciated.22
Limitations
Of course, it is easier to suggest an intervention improves care than to prove an intervention is effective.23 In particular, we have no data testing whether the increased availability of hand sanitizers throughout a hospital has reduced nosocomial infections or enhanced hand hygiene beyond what might have already changed with a pandemic.24 As a field test for reliability, however, we sent a medical student to check 100 dispensers at Sunnybrook and found only 2 empty (each phoned and refilled in an hour). In comparison, a similar check of 100 dispensers at a different hospital within 10 km of Sunnybrook found 11 empty (no readily apparent way to notify for refilling).
Several additional contextual points merit mention because removing sludge does not mean hand hygiene improves automatically. Many other gaps arise due to education, motivation or inconvenience.25 The chain of manufacturing, shipping, receiving and distribution must be intact. The Environmental Services team must have sufficient capacity to meet demands and the number of extraneous calls needs to be infrequent. To our knowledge, Medline offers no past study on the problem of empty hand hygiene dispensers elsewhere and we have no historical data about the frequency of empty dispensers at Sunnybrook before the pandemic.
A randomised trial to address these limitations and provide more data on effectiveness would not be easy. The trial would need a large sample size because of the multifactorial links between improvements in structures (eg, sanitizer dispensers), processes (eg, hand hygiene) and outcomes (eg, nosocomial infections).26 Enforcing randomisation would be hard due to the information spillover within a hospital.11 Hidden confounding might still persist since participants would be aware of the intervention and vulnerable to subconscious bias (akin to the non-blinded design of most nudges).27 These limitations mean substantive data on effectiveness are not likely to be available anytime soon.
Implications
The COVID-19 epidemic can help explain why this nudge was not previously considered (to our knowledge). Manufacturers are now offering dispensers as loss leaders to generate revenues from subsequent sanitizer sales and philanthropists are now donating funds to support hand hygiene as a tangible way to aid front-line healthcare workers in a crisis. This means activity has increased substantially due to an intuitive principle: encourage behaviour by making it easier. The added nudge, therefore, adapts to a hospital context where the ‘average’ dispenser needs a refill once per week (capacity 1200 cc) and total sanitizer liquid flow now exceeds 1000 L/ week.
The tremendous amount of hand-hygiene means even a small nudge is important.28 Consider, for example, 1000 L of sanitizer liquid consumption implies a million hand washes per week (1 squirt=1 cc). If each occasion requires 10 s, this totals to over 2500 hours of activity (1,000,000×10/60/60) with a cash value of $50 000 assuming an average salary is $20/hour (2500×20). The total procurement cost of sanitizer liquid is about $10 000 ($10/L) suggesting the time cost of hand hygiene (staff plus visitors combined) is five times greater than the marginal cost of materials. The nudge, therefore, helps address cost-effectiveness by reducing wasted time.
The opportunity for improved hand hygiene has also attracted market forces for additional design features following suggestions by others and ourselves. Dispensers at Sunnybrook, for example, now have a front cover manufactured that sacrifices product branding and includes a message window instead (figure 2). The message window then allows the appropriate telephone number of the local hospital to appear based on a card clipped inside the dispenser. This new manufacturing design suggests the nudge for hand hygiene is anticipated to extend widely and has a reasonable business case for sustained commercial profit.
Summary
We believe these three words and single number may contribute to improved hand hygiene. At Sunnybrook, the words are ‘If empty call 4555’. Other healthcare organisations might wish to create a similar label to affix to the hand sanitizer dispensers inside their own facilities. More broadly, this intervention helps illustrate the potential value of a simple nudge for improving daily medical care and that a crisis is an opportunity to make constructive changes that might otherwise be neglected.29 The COVID-19 epidemic, therefore, has the potential to introduce durable healthcare gains that increase patient safety and reduce clinician frustration.
Acknowledgments
We thank the following individuals for helpful comments on earlier drafts of this article: Lou Fernandez, Fizza Manzoor, Kelvin Ng, Joseph Pali, Sheharyar Raza, Eldar Shafir and Michael Young.
References
Footnotes
Contributors Both authors contributed to the design, analysis and interpretation of the findings. The first author (DAR) wrote the first draft. Both authors approved the final submitted version.
Funding This project was supported by the Canada Research Chair in Medical Decision Sciences and the Canadian Institutes of Health Research.
Competing interests None declared.
Patient consent for publication Not required.
Author note Crowdsourcing provides a practical method to support hand hygiene by allowing outpatients, families, visitors, staff, volunteers and other to notify hospital support services about empty hand sanitizer dispensers that requires refilling.